Handbook of human factors in litigation

HANDBOOK OF Human Factors in Litigation HANDBOOK OF Human Factors in Litigation EDITED BY Y.Ian Noy Waldemar Karwows...

Author: Waldemar Karwowski

298 downloads 2881 Views 5MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

HANDBOOK OF Human Factors in Litigation

HANDBOOK OF Human Factors in Litigation EDITED BY

Y.Ian Noy Waldemar Karwowski

CRC PRESS Boca Raton London New York Washington, D.C.

This edition published in the Taylor & Francis e-Library, 2006. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to http://www.ebookstore.tandf.co.uk/.” Library of Congress Cataloging-in-Publication Data Handbook of human factors in litigation/edited by Y.Ian Noy and Waldemar Karwowski. p. cm. Includes bibliographical references and index. ISBN 0-415-28870-3 (alk. paper) 1. Evidence, Expert—United States. 2. Forensic engineering—United States. 3. Human engineering—United States. I. Noy, Ian.Y. II. Karwowski, Waldemar, 1953– KF8968.25.H36 2004 347.73′67–dc22 2004056121 This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming, and recording, or by any information storage or retrieval system, without prior permission in writing from the publisher. All rights reserved. Authorization to photocopy items for internal or personal use, or the personal or internal use of specific clients, may be granted by CRC Press, provided that $1.50 per page photocopied is paid directly to Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923 USA. The fee code for users of the Transactional Reporting Service is ISBN 0-415-28870-3 (Print Edition)/05/$0.00+$ 1.50. The fee is subject to change without notice. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. The consent of CRC Press does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from CRC Press for such copying. Direct all inquiries to CRC Press, 2000 N.W. Corporate Blvd., Boca Raton, Florida 33431. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. Visit the CRC Press Web site at www.crcpress.com © 2005 by CRC Press No claim to original U.S. Government works ISBN 0-203-49029-0 Master e-book ISBN

ISBN 0-203-57160-6 (Adobe e-Reader Format) International Standard Book Number 0-415-28870-3 (Print Edition) Library of Congress Card Number 2004056121

Preface Forensic ergonomics and human factors is the application of ergonomics theory, principles, and data to determine causes of system failures, often involving personal injury, considered in the context of litigation. The International Ergonomics Association defines ergonomics (or human factors) as the scientific discipline concerned with the understanding of interactions among humans and other elements of a system, and the profession that applies theory, principles, data, and methods to design in order to optimize human well-being and overall system performance. Derived from the Greek ergon (work) and nomos (laws) to denote the science of work, ergonomics is a systems-oriented discipline which now extends across all aspects of human activity. Ergonomics promotes a holistic approach in which considerations of physical, cognitive, social, organizational, environmental, and other relevant factors are taken into account. Whereas the vast majority of ergonomics and human factors practitioners are involved in the design of human-machine systems to prevent occurrences of human error, forensic practitioners seek to answer the question “What went wrong?” Most of forensic practice falls within the domain of civil litigation, though there are important applications in criminal cases, for example, that relate to the fallibility of eyewitness reports or human actions that may constitute criminal negligence. Analyses of human causes of system failures can uncover valuable information that can help prevent recurrence through engineering or administrative controls. In addition, expert testimony during the litigation process assists the court in determining liability by interjecting scientific evidence concerning human performance and behavior. In the aggregate, this improves the fairness and objectivity of the judicial system in protecting individuals and society. As a result of ergonomics-based evidence, liability attributable to inappropriate designs or business practices has become a significant consideration in corporate decisions influencing the nature and extent of product research and development, workplace and infrastructure interventions, and the manner of service provision. The need to minimize exposure to liability provides strong motivation to designers, service providers, and employers to elevate the safety of their products, workplace, or service to a level higher than might otherwise be the case. Forensic human factors is a professional endeavor that is neither taught nor systematically practiced, even in the U.S., where it is most prevalent. While the profession is anchored in the science of human factors, forensic practice today relies on individually acquired skills and experience. There is little documented material available to help practitioners develop expertise in a systematic manner. When we decided to create this handbook, we intended to produce a reference that was sufficiently comprehensive in scope and having sufficient depth to be a valuable tool for forensic experts, students, and legal practitioners alike. This handbook was not intended to teach the science of ergonomics and human factors—there are several excellent resources

available to satisfy readers interested in the science; rather, it was intended to provide guidance on the preparation, analyses, and presentation of forensic evidence as well as current scientific knowledge in key applications. The Handbook of Human Factors in Litigation is a resource for forensic and legal professionals and includes chapters relating to the litigation process, forensic approaches and methods, and important scientific data in the major application areas, including case studies. It is a compilation of the current state-of-the-art knowledge and practice, written by authors with vast experience in the field, and it reflects their individual expertise and views. The handbook contains a total of 38 chapters organized under six main headlines, including: Professional Issues (7 chapters), Human Performance in the Legal Context (5 chapters), Driving Environments (6 chapters), Physical and Cognitive Factors (6 chapters), Product Liability and Warnings (9 chapters), Human Factors Applications (4 chapters), and a Guide to Human Factors Terminology. Readers will note that some chapters deal with common topics, but from a different perspective. We tried to minimize overlap, but in this kind of endeavor some overlap among chapters is unavoidable. Meeting our initial goals proved a greater challenge than we had anticipated. Forensic application of ergonomics and human factors is an ever-evolving field and the number of qualified forensic practitioners is very small indeed. Although current emphasis is on consumer product deficiencies, slips and falls, and motor vehicle crashes, new applications are emerging. The handbook represents the current state-of-practice. It is the first text of its kind and we are confident that it will serve as an important resource and pave the way for the establishment of a more formal body of relevant knowledge in the area. We are indebted to the contributors to this handbook, who are individually renowned scholars and experts. We are also indebted to Taylor & Francis for encouraging us to edit this volume, and would like to acknowledge editorial assistance provided by Cindi Carelli and Jessica Vakili of CRC Press. Our goal was twofold: to assist legal professionals in recognizing the role and contribution of ergonomics and human factors in litigation, and to assist forensic experts in performing their work in an objective and effective manner. In realizing this handbook, it is our fervent hope that it will contribute to the efficacy and equitability of the justice system, and thereby lead to the safer design and use of technology. Y.Ian Noy Waldemar Karwowski

The Editors

Y.Ian Noy, Ph.D., P.Eng., CPE, is President of Systems Ergonomics, Inc., a consulting firm specializing in industrial and forensic ergonomics. He holds a doctorate degree in Industrial Engineering from the University of Toronto, specializing in human factors. He is a Board Certified Professional Ergonomist (BCPE) with more than 30 years of professional experience in a variety of private and public sector applications. Forensic experience includes expert testimony in civil litigation arising from motor vehicle collisions and slips and falls. He is a recipient of several awards, including the U.S. National Highway Traffic Safety Administration’s Engineering Excellence Award. His career began in 1973 as a behavioral scientist at the Canadian Defence and Civil Institute of Environmental Medicine (DCIEM) working on a variety of military human— machine systems. In 1982 he joined Transport Canada to undertake traffic safety research to support the development of vehicle safety standards and other collision interventions. He was quickly promoted to Chief of the Ergonomics Division and, in 2001, to the position of Director, Standards Research and Development, where he is responsible for motor vehicle safety regulations, crashworthiness, ergonomics and vehicle systems research, and motor carrier safety. Dr. Noy’s applied research experience spans applications in the air, on the ground, and underwater, including military R&D. He has published more than 100 scientific and technical reports, as well as conference and journal articles. He has prepared and presented lectures in human factors on a variety of topics, such as traffic safety, human factors in intelligent transport systems, human-machine interface design and evaluation, human performance and behavior, and the role of ergonomics in civil litigation. He serves on the editorial board of several journals, including Applied Ergonomics, and is Associate Editor of the International Encyclopedia of Ergonomics and Human Factors. In addition, he edited the book The Ergonomics and Safety of Intelligent Driver Interfaces (Lawrence Erlbaum & Associates, 1997).

Dr. Noy is Past President of the International Ergonomics Association (IEA). He is a Fellow of the Human Factors and Ergonomics Society (HFES), a Past President and Honorary Fellow of the Association of Canadian Ergonomists/Association Canadienne d’Ergonomie (ACE), and a member of the Association of Professional Engineers of the Province of Ontario (PEO). He is also a member of the Transportation Research Board Committee on Simulation and the Measurement of Driving. Dr. Noy was the chairman of the 12th Congress of the IEA held in Toronto in 1994. In 1998, he was leader of the People to People Ambassador Programs’ Ergonomics Delegation to the People’s Republic of China.

Waldemar Karwowski, Sc.D., Ph.D., P.E., is Professor of Industrial Engineering and Director of the Center for Industrial Ergonomics at the University of Louisville, Louisville, Kentucky. He holds an M.S. (1978) in Production Engineering and Management from the Technical University of Wroclaw, Poland, and a Ph.D. (1982) in Industrial Engineering from Texas Tech University. He was awarded the Sc.D. (dr hab.) degree in Management Science by the Institute for Organization and Management in Industry (ORGMASZ), Warsaw, Poland (June 2004). He is also a Board Certified Professional Ergonomist (BCPE). His research, teaching, and consulting activities focus on human system integration and safety aspects of advanced manufacturing enterprises, human-computer interaction, prevention of work-related musculoskeletal disorders, workplace and equipment design, and theoretical aspects of ergonomics science. Dr. Karwowski is the author or co-author of more than 300 scientific publications (including more than 100 peer-reviewed archival journal papers) in the areas of work systems design, organization, and management; macroergonomics; human-system integration and safety of advanced manufacturing; industrial ergonomics; neuro-fuzzy modeling in human factors; fuzzy systems; and forensics. He has edited or co-edited 35 books, including the International Encyclopedia of Ergonomics and Human Factors, Taylor & Francis, London (2001). He was a winner of the Best Reference Award 2002 from the Engineering Libraries Division, American Society of Engineering Education, and the Outstanding Academic Title 2002 from Choice magazine. Dr. Karwowski serves as editor of the Human Factors and Ergonomics in Manufacturing international journal published by John Wiley & Sons, New York, and is the editor-in-chief of Theoretical Issue in Ergonomics Science (TIES) (Taylor & Francis, London). Dr. Karwowski serves as co-editor of the International Journal of Occupational Safety and Ergonomics and

consulting editor of the Ergonomics Journal. He is also a member of editorial boards for several peer-reviewed journals, including Human Factors, Applied Ergonomics, International Journal of Human-Computer Interaction, Universal Access to the Information Society: An International Interdisciplinary Journal, Occupational Ergonomics, and Industrial Engineering Research: An International Journal of RE Theory and Application (Hong Kong). Dr. Karwowski served as Secretary-General (1997–2000) and President (2000–2003) of the International Ergonomics Association (IEA). He was elected an Honorary Academician of the International Academy of Human Problems in Aviation and Astronautics (Moscow, Russia, 2003), and was named the Alumni Scholar for Research (2004–2006) by the J.B. Speed School of Engineering of the University of Louisville. He also received the University of Louisville Presidential Award for Outstanding Scholarship, Research and Creative Activity in the Category of Basic and Applied Science (1995), Presidential Award for Outstanding International Service (2000), and the W.Jastrzebowski Medal for Lifetime Achievements from the Polish Ergonomics Society (1995). Dr. Karwowski is Fellow of the International Ergonomics Association (IEA), the Human Factors and Ergonomics Society (HFES, USA), the Institute of Industrial Engineers (IIE, USA), and the Ergonomics Society (United Kingdom). He is past President of the International Foundation for Industrial Ergonomics and Safety Research, as well as past Chair of the U.S. TAG to the ISO TC159: Ergonomics/SC3 Anthropometry and Biomechanics. He served as Fulbright Scholar and Visiting Professor at Tampere University of Technology, Finland (1990–1991), and was named an Outstanding Young Engineer of the Year by the Institute of Industrial Engineers (1989). He can be reached via e-mail at [email protected].

Contributors Gerson J.Alexander Positive Guidance Applications Rockville, Maryland William B.Askren Human Factors Services Dayton, Ohio Gary M.Bakken Analytica Systems International, Inc. Tuscon, Arizona Richard D.Blomberg Dunlap and Associates, Inc. Stamford, Connecticut Francisco J.Bricio The Bricio Law Firm Gary, North Carolina Daniel A.Bronstein Michigan State University East Lansing, Michigan C.Shawn Burke University of Central Florida Orlando, Florida H.Harvey Cohen Error Analysis, Inc. La Mesa, California Deborah Davis University of Nevada Reno, Nevada Patrick G.Dempsey Liberty Mutual Research Institute for Safety Hopkinton, Massachusetts Sidney W.A.Dekker Lund University Lund, Sweden Jason Devereux University of Surrey Guilford, Surrey, U.K. Robert Dewar Western Ergonomics, Inc. Calgary, Alberta, Canada

William C.Follette University of Nevada Reno, Nevada Magdalen Galley Ergonomics Consultant Leicestershire, England Peter A.Hancock University of Central Florida Orlando, Florida Martin G.Helander Nanyang Technological University Singapore Max Hely Safety Science Associates Pty., Ltd. Sydney, Australia Allen K.Hess Auburn University at Montgomery Montgomery, Alabama Henry M.Hobschied Products Liability Consultant and Legal Editor Shawano, Wisconsin Donald P.Horst Private Consultant Sunnyvale, California John M.Howard Crossroads Machine, Inc. Dayton, Ohio Roger C.Jensen Montana Tech Butte, Montana Daniel A.Johnson Daniel A.Johnson, Inc. Olympia, Washington Murray Kaiserman Health Canada Ottawa, Ontario, Canada Waldemar Karwowski University of Louisville Louisville, Kentucky Vincent Kelly University of Surrey Guildford, Surrey, U.K. Markus Kemmelmeier University of Nevada Reno, Nevada

Martin I.Kurke Kurke Associates Springfield, Virginia Frank J.Landy SHL Litigation Support Boulder, Colorado Cindy A.LaRue Error Analysis, Inc. La Mesa, California Kenneth R.Laughery Rice University Houston, Texas Elizabeth E Loftus University of California Irvine, California Celine McKeown Link Ergonomics Nottingham, England Dick Moll University of Wisconsin Madison, Wisconsin Rudolf G.Mortimer Human Factors Engineering Urbana, Illinois Jeffrey W.Muttart Accident Dynamics Research Uncasville, Connecticut Jone McFadden Papinchock SHL Litigation Support Boulder, Colorado Christopher P.Nemeth University of Chicago Chicago, Illinois Thorny Nilsson University of Prince Edward Island Charlottetown, Prince Edward Island, Canada Richard A.Olsen Human Performance: Limited Santa Clara, California Tal Oron-Gilad University of Central Florida Orlando, Florida Heather A.Priest University of Central Florida Orlando, Florida

Patricia Robinson Coronado Consulting Services, LLC Sonoita, Arizona Eduardo Salas University of Central Florida Orlando, Florida Stuart M.Statler Safety Strategies Arlington, Virginia Donald I.Tepas University of Connecticut Storrs, Connecticut David R.Thom Collision and Injury Dynamics, Inc. El Segundo, California David A.Thompson Stanford University Palo Alto, California Alison G.Vredenburgh Vredenburgh & Associates, Inc. Carlsbad, California Evelyn Vingilis University of Western Ontario London, Ontario, Canada Ron Wardell University of Calgary Calgary, Alberta, Canada Katherine A.Wilson University of Central Florida Orlando, Florida Michael S.Wogalter North Carolina State University Raleigh, North Carolina Dennis Wylie D.Wylie Associates Santa Barbara, California Ilene B.Zackowitz Vredenburgh & Associates, Inc. Carlsbad, California

Contents I Professional Issues 1 The Discipline of Human Factors Engineering and Ergonomics 3 Martin G.Helander 2 Preparing and Presenting Evidence in Court 33 Daniel A.Bronstein 3 Presenting Behavioral Science Data as Legal Evidence: Legal Standards 43 that the Ergonomic and Human Factors Expert Needs to Know Allen K.Hess 56 4 Practical Ethics for the Expert Witness in Ergonomics and Human Factors Forensic Cases Allen K.Hess 71 5 A Road Map for the Practice of Forensic Human Factors and Ergonomics William B.Askren and John M.Howard 91 6 Can Training for Safe Practices Reduce the Risk of Organizational Liability? Katherine A.Wilson , Heather A.Priest , Eduardo Salas , and C.Shawn Burke 131 7 The Influence of Daubert on Expert Witness Testimony—The Human Factors Context Jone McFadden Papinchock and Frank J.Landy II Human Performance in the Legal Context 8 Reconstructing Situated Performance in Human Error Investigations Sidney W.A.Dekker 9 Causation Issues in Workers’ Compensation Roger C.Jensen and Francisco J.Bricio 10 Legal Issues in Work-Related Musculoskeletal Disorders: a European Perspective from the U.K Vincent Kelly and Jason Devereux 11 Age and Functioning in the Legal System: Victims, Witnesses, and Jurors Deborah Davis and Elizabeth F.Loftus

148 174 187

215

12 Memory for Conversation on Trial Deborah Davis , Markus Kemmelmeier , and William C.Follette

281

III Driving Environments 13 Human Factors in Traffic Crashes Rudolf G.Mortimer , Richard D.Blomberg , Gerson J.Alexander , and Evelyn Vingilis 14 Estimating Driver Response Times Jeffrey W.Muttart 15 Pedestrian Injury Issues in Litigation Richard A.Olsen 16 Pedestrian Accidents in Traffic Robert Dewar 17 Commercial Motor Vehicle Collisions Dennis Wylie 18 Human Factors Issues in Motorcycle Collisions Peter A.Hancock , Tal Oron-Gilad , and David R.Thom

320

403 436 466 487 512

IV Physical and Cognitive Factors 19 Perceptual-Cognitive and Biomechanical Factors in Pedestrian Falls H.Harvey Cohen and Cindy A.LaRue 20 Measurement in Pedestrian Falls Daniel A.Johnson 21 Balcony Falls Magdalen Galley 22 Preplacement Strength and Capacity Assessment for Manual Materials Handling Jobs Patrick G.Dempsey 23 Identifying the Real Issues in Work-Related Upper Limb Disorders Celine McKeown 24 Exercise Injuries: Human Factors in Fitness Facilities Max Hely

539 567 602 622

629 656

V Product Liability and Warnings 25 Preventing “Accidental” Injury: Accountability for Safer Products by Anticipating Product Risks and User Behaviors Stuart M.Statler 26 Human Factors Issues to Be Considered by Product Liability Experts Alison G.Vredenburgh and Ilene B.Zackowitz 27 Products Liability Law: What Engineering Experts Need to Know Dick Moll , Patricia A.Robinson , and Henry M.Hobscheid 28 Human-Centric Approach to Forensic Analysis for System Liability Gary M.Bakken 29 Product Liability for the Human Factors Practitioner Ron Wardell 30 The Warning Expert Kenneth R.Laughery and Michael S.Wogalter 31 Effectiveness of Consumer Product Warnings: Design and Forensic Considerations Michael S.Wogalter and Kenneth R.Laughery 32 Legibility of Warnings in Color Thorny Nilsson and Murray Kaiserman 33 A Human Factors View of Product Liability and Malpractice Litigation Martin I.Kurke

682

700 715 727 755 763 780

794 819

VI Human Factors Applications 34 The Impact of Shiftwork on Manufacturing and Transportation Workers Donald I.epas 35 Preschoolers, Adolescents, and Seniors: Age-Related Factors Pertaining to Forensic Human Factors Analyses Ilene B.Zackowitz and Alison G.Vredenburgh 36 Sexual Harassment: A Forensic Human Factors Perspective Alison G.Vredenburgh and Ilene B.Zackowitz 37 Health Care Forensics Christopher P.Nemeth

833

851

865 877

VII Human Factors Terminology 38 A Guide to Forensic Human Factors Terminology 902 David A.Thompson , H.Harvey Cohen , Donald P.Horst , Daniel A.Johnson , and Richard A.Olsen Index

960

I Professional Issues

1 The Discipline of Human Factors Engineering and Ergonomics Martin G.Helander Nanyang Technological University 0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

1.1 Introduction Human factors and ergonomics (HFE) are two branches of a scientific discipline concerned with: (1) interaction and design of systems that people use at work and in leisure and (2) appropriate work procedures and practices, tools, and computer systems that encompass the work system. The purpose of the design activities is to match systems, jobs, products, and environments to the physical and mental abilities and limitations of people. HFE professionals are involved in these activities. They are also involved in training or educating operators in the use of the system. Ideally, systems should be well designed so that they are intuitive to use and do not require special training or education. However, this turns out to be a very challenging design task as gauged from various usability problems experienced by operators. In Europe, ergonomics started with industrial applications in the 1950s. Information from work physiology, biomechanics, and anthropometry was applied in the design of workstations and industrial processes. The aim was to enhance the well-being of workers and improve manufacturing productivity. In the United States, human factors and its sister disciplines, human factors engineering and engineering psychology, developed from military applications. Knowledge from experimental psychology and systems engineering was used to improve systems performance and system quality. Despite the historical differences between human factors and ergonomics in the type of knowledge used and in the goals for design, the two approaches are merging. In the Western world, physical workload is no longer common. In manufacturing, hard physical labor has been taken over by materials-handling aids, mechanical processes, and automation. Legislation has also put limits to the amount of workload to which employees can be exposed. At the same time, the introduction of automation and computers in the workplace has transformed many factories, as well as offices. A modern factory can now be demanding because of its high mental demands—rather than physical demands. Similarly, during the last 15 years in the U.S., there has been a great interest in physical workload, including biomechanics of work postures and correct lifting

Handbook of Human factors in litigation

4

techniques. Many companies today employ HFE staff to help with improving the work environment from safety as well as productivity perspectives. Human factors and ergonomics (HFE) have since drawn closer. One indication is the change of name of the Human Factors Society in the U.S. to the Human Factors and Ergonomics Society. As such, the terms human factors and ergonomics are used synonymously in this chapter. 1.2 History of Human Factors and Ergonomics The word ergonomics is derived from the Greek ergo (work) and nomos (rules, law). It was first used by Wojciech Jastrzebowski in a Polish newspaper in 1857 (Karwowski, 1991). One may argue that ergonomics is nothing new. Hand tools, for example, have been used since the beginning of human kind. Various types of hand tools have been developed since the Stone Age, and the interest in ergonomic design can be traced back in history (Childe, 1944; Braidwood, 1951). HFE in the 18th and 19th Centuries In the 18th century Ramazzini (1713) published a book, The Diseases of Workers, in which he documented links between many occupational hazards and the type of work performed. He described, for example, the development of cumulative trauma disorder and believed that these events were caused by repetitive motions of the hand, constrained body posture, and excessive mental stress. La Mettrie’s controversial book L’homme Machine was published in 1748, at the beginning of the Industrial Revolution. Two things can be learned from La Mettrie’s writings. First, the comparison of human capabilities and machine capabilities was a sensitive issue already in the 18th century. Second, by considering how machines operate, one can also learn much about human behavior: “to be a machine, to feel, to think, to know how to distinguish good from bad, as well as blue from yellow, in a word, to be born with an intelligence and a sure moral instinct” (La Mettrie, 1748). Some of these issues remain debated in ergonomics today. For example, an analysis of the constraints in utilizing industrial robots helps us understand how industrial tasks should be designed to better fit humans (Helander, 1995). Rosenbrock (1983) pointed out that, during the Industrial Revolution, efforts were made to apply the concepts of a “human-centered design” to new machines, such as the spinning Jenny and the spinning mule. The concern then was to allocate interesting tasks to the human operator, leaving the machine to perform the more boring and repetitive tasks. Taylorism At the beginning of the 20th century, Frederick Taylor introduced the “scientific” study of work. The tasks of industrial workers were analyzed and opportunities were sought to enhance productivity by simplifying the worker’s movement patterns. Many of Taylor’s studies analyzed the work of bricklayers; he proposed many useful recommendations

The discipline of human factors engineering and ergonomics

5

concerning how many bricks a mason can lift at the same time, and at what height the bricks should be stored to minimize physical work. In line with Taylor’s tradition, Frank and Lillian Gilbreth developed the time and motion study. Ordinary jobs were divided into several microelements called “therbligs” (Konz and Johnson, 1990). The motions of the workers were classified in elements such as “reach,” “grasp,” and “move.” The motion times were then calculated and used for redesigning jobs as well as for calculating workers’ salaries. Today, there are often objections against the time and motion study, which is seen as a tool for exploiting workers. Nonetheless, these methods remain useful for measuring and predicting work performance time. The time and motion study, if used for the right purpose, is a valuable tool. Accident Proneness In the beginning of the 20th century, research in industrial psychology emphasized how one should select operators who are suitable for certain tasks. This was based on studies of human skills. The research on accident proneness is typical of this era. Accident proneness implies that certain individuals are more likely to have accidents than others because they have enduring personality characteristics that make them unsafe. The research in this time period centered on understanding how accident-prone individuals differed from “normal” people, so that one could exclude them from participating in activities in which they were likely to incur accidents. For example, one could revoke a driver’s license. This approach dominated accident research for about 40 years, but it was not fruitful. The results of many research studies have in retrospect proven to be useless because accident proneness and many personality features are not stable—they fluctuate with age and experience (Shaw and Sichel, 1975). In the 21st century, this approach is considered totally misleading. In fact, automobile drivers who have been involved in accidents do not necessarily incur future accidents. Likewise, factory workers who experience accidents in the workplace cannot be seen as accident prone. The tendency to blame individuals for accidents is seemingly counterproductive. With greater awareness of ergonomics, it is now realized that human errors are mostly caused by poor design, and the modern approach is to design environments and artifacts so that they fit the limitations and capabilities of the users. What is needed is a greater focus on environmental and machine designs to make environments and machines safer. For similar reasons, safety training is only partially valid. Many accidents involve unexpected scenarios for which it is difficult to prepare by training. For example, injuries due to manual materials handling are often due to a combination of unexpected events, and asking workers to train in “correct lifting techniques” does not help (Kroemer et al., 1994). International Proliferation of HFE Since the 1950s, HFE has proliferated on the other large continents: Asia, Africa, South America, and Australia (Luczak, 1997). In many developing countries, in particular Southeast Asia, new problems are resulting from rapid industrialization; these have led to the situation in which workers have difficulties in adapting to new technologies. The


6

transition from a rural, agrarian life to urban industries comes at a cost. Workers are paying a price in terms of a tremendous increase of industrial injuries and stress. Many of these problems remain hidden or unreported, and official statistics are rarely available to illuminate the true state of affairs. As Vanwonterghem (1994) claimed, the industrialization has come about too quickly. Thus, developing countries have a great need for ergonomics.

TABLE 1.1 Five Most Important Early and Current Applications of Ergonomics in 25 Ergonomics Societies Importance

Early applications

Current applications

1 Anthropometry Safety 2 Work physiology Industrial engineering 3 Industrial engineering Biomechanics 4 Biomechanics Macroergonomics 5 Psychology Human-computer interaction Source: Brown, O. Jr., Noy, I., and Robertson, M. (1995). Special survey of IEA Federated Societies. Santa Monica, CA: International Ergonomics Association, c/o Human Factors and Ergonomics Society.

In developing countries in which ergonomics is adopted, the emphasis has been on job design including biomechanics, heat stress, and work physiology (Khalid, 2003). However, with the introduction of computers, there is now a sudden shift in focus to the problems of usability of complex systems. Usability is a universal problem that cuts across cultures. In the transition to the new, computerized world, many developing countries are bypassing several stages of development and find themselves immediately immersed in the computerized global environment. What took the Western World 200 years to accomplish has taken the developing countries 30-odd years. Associated with this development are new ergonomics problems in dealing with the globalization of communication, integration of resources, and global management. The Asian tigers are well positioned to take a lead in this area but, at present, lack the necessary infrastructure in terms of experience and trained personnel. Tremendous economic potential lies in designing usable systems for global communication and customized markets. Technology transfer from the Western world may no longer be critical. A survey among ergonomists in 25 countries clearly demonstrates the change in emphasis (Brown et al., 1995). Although biomechanics remains important, a need is emerging for cognitive ergonomics and macroergonomics, as summarized in Table 1.1. It is interesting to note that this trend was valid not only for industrialized countries but also for industrially developing countries. Educational Background of HFE Professionals Ergonomists come from a variety of professional fields. The mixed background is well demonstrated in the composition of the membership of professional societies, which typically consists of engineers, psychologists, and individuals from the medical


7

profession. HFE professionals have received academic training in engineering, ergonomics, medicine, physiology, physiotherapy, psychology, and safety. These educational programs accounted for about 70% of all members. Of the current U.S. membership, 29% are academics, 27% practice in industry, 15% do research, 10% are private consultants, 8% work for the government, and 11% work in other occupations. Certification and licensing of ergonomists are important issues to the practicing professional. Several organizations uphold rules for professional certification, including the Board for Certification of Professional Ergonomists (BCPE) in the U.S. and the Centre for Registration of European Ergonomists (CREE) in Europe. The International Ergonomics Association upholds a set of recommendations that can be used by member countries in certifying HFE professionals. 1.3 International Ergonomics Association (IEA) Although the profession has been strongly developed in many individual countries, there is also an increasing internationalization of HFE. This is managed by IEA, an organization for HFE organizations around the world. Its origin can be traced to a seminar on fitting the job to the worker, which took place in The Netherlands in 1957. The first Congress of IEA was held in Stockholm in 1961 (IEA, 2004). For many years, the focus of IEA was on the welfare and productivity of workers from the point of view of the biological sciences (Smith, 1988). Due to rapid technological development, this perspective has since changed radically. Today, IEA considers also nonwork activities including play, leisure, home, and travel. The new focus is on mental demands and organizational prerequisites. Definitions of Ergonomics Over the years, there have been many definitions of ergonomics. Licht and colleagues (1991) identified 130 definitions of human factors and ergonomics. These keep changing, thereby emphasizing the more recent advances. IEA (2004) defined the discipline of ergonomics in the following way: Ergonomics (or human factors) is the scientific discipline concerned with the understanding of interactions among humans and other elements of a system, and the profession that applies theory, principles, data and methods to design in order to optimize human well-being and overall system performance…. Ergonomists contribute to the design and evaluation of tasks, jobs, products, environments and systems in order to make them compatible with the needs, abilities and limitations of people. The following definition is inspired by Chapanis (1995) and Helander (1997): Ergonomics and human factors use knowledge of human abilities and limitations to the design of systems, organizations, jobs, machines, tools, and consumer products for safe, efficient, comfortable and satisfying human use.


8

Note that design is the main purpose. As such, ergonomics differs from the sciences that support the ergonomics discipline. In anthropology, cognitive science, psychology, sociology, and medical sciences, the primary purpose is to understand and model human behavior, rather than to utilize knowledge for design. Ergonomics may have more in common with engineering, which is design oriented (Moray, 1994). This makes the practice of ergonomics/human factors an applied science. Ergonomics may thus be thought of as a technology, not merely a science. Today, most ergonomics education takes place within engineering programs, which is indicative of the importance of design. 1.4 Development of HFE since 1950 In exploring the field of ergonomics, there may be reasons to look at its evolution in Europe and the U.S., starting around 1950. In the U.S., human factors emerged as a discipline after World War II. Many design problems were encountered in the use of sophisticated war equipment such as airplanes, radar and sonar stations, and tanks. Sometimes these caused human errors with grave consequences. For example, during the Korean War, reputedly more U.S. pilots were killed during training than in war activities. This raised the issues of appropriate design of controls and displays in cockpits: how can information be displayed better in the cockpit, and how can controls be integrated with the task so that they are easier to handle? Improvements were implemented, such as controls that combined several functions, and displays were laid out so that the relationship between controls and displays was clear; control-display compatibility became an important design concept. As a result of these improvements and new pilot training programs, the number of fatalities in pilot training decreased to a fraction (5%) of what they had been. Ever since, much of the research in human factors in the U.S. has been sponsored by the Department of Defense. Consequently, the information in textbooks on human factors is heavily influenced by results from military research. This is not necessarily a disadvantage because most models, theories, and findings are applicable to the design of civilian systems as well; the human operator remains the same. Other U.S. federal agencies have sponsored research on many civilian applications, including: • Federal Highway Administration: design of highways and road signs • National Aeronautics and Space Administration: physiological impact; human factors design of space stations • National Highway Traffic Safety Administration: design of cars to improve crash worthiness, lighting system, and controls • Department of the Interior: ergonomics in mining • National Bureau of Standards (now called National Institute of Standards and Technology): consumer product safety • National Institute of Occupational Health: ergonomic injuries at work; industrial safety; work stress • Nuclear Regulatory Commission: design requirements for nuclear power plants • Federal Aviation Administration: aviation safety; air traffic control


9

Eastman Kodak in Rochester, New York, was the first company in the U.S. to implement a substantial industrial ergonomics program, circa 1965 (Chengalur et al., 2004). In Europe, ergonomics had a different history due to various influences. For example, in Great Britain, the coal mining industry sparked tremendous ergonomics interest, especially in the area of safety and occupational hazards. Ergonomics is also well established in Belgium, France, Germany, Holland, Italy, Poland, the Scandinavian countries, and Spain. In most European countries, the driving factors for ergonomics have been worker safety, health, and comfort. Particularly in the Scandinavian countries and Germany, labor unions have taken a strong interest, and they often dictate the type of production machinery to be purchased. Today, ergonomics in industry has the dual purpose of promoting productivity and improved conditions of work. Several recent studies have pointed to significant improvements in productivity as a result of ergonomic measures taken in industry. With rather modest investments, benefit-cost ratios of 50:1 have been obtained (Helander, 1995). 1.5 Research and Development in Human Factors and Ergonomics— a Systems Approach Ergonomics research can be applied as well as basic. Good research should be supported by theories of human behavior and human functionality. Often theories in the behavioral and cognitive sciences are used. However, other sciences, such as engineering, medicine, biomechanics, and computer science, also contribute toward understanding human behavior and functionality. In formulating research topics and design solutions, not only basic human characteristics such as perception, memory, and manual response but also a host of other related issues should be considered. For example, realizing that human response is different between individuals, one can try to measure some of the cognitive and biomechanical characteristics and relate these to the performance of users. For example, in human computer interaction, good cognitive abilities such as visual memory and logical reasoning are very important. Users with low visual memory capacity do perform as well as users with high visual memory (Egan, 1988). In other words, HFE applies knowledge from related disciplines in its basic research to derive good design solutions. It can be said that HFE engages in basic and applied research to seek support for design solutions. Therefore, it is a science as well as a craft (see Long and Dowell, 1989). “Science” denotes its empirical and systematic approach, while “craft” refers to the art and creativity of its application. A Systems Description of HFE In this subsection, the science of HFE is described by taking a systems approach (see Christensen, 1962). Most ergonomics problems may be explained in a systems context and, for HFE, one may use an environment-operator-machine system, as illustrated in Figure 1.1. The operator is the central focus in HFE and should be described in an organizational context; this is the purpose of Figure 1.1, which illustrates only the most important operator concepts. In reality, human perception, information processing, and


10

response are much more complex, with many feedback loops and variables that are not detailed in Figure 1.1 (for an overview, see Wickens and Hollands, 2000).

FIGURE 1.1 Ergonomics systems model for measurement of safety and productivity. In HFE, a classification of independent and dependent variables is often used to analyze a problem. The independent variables are associated with design parameters of the environment and the machine (such as alternative task allocation, controls, and displays). The dependent variables are associated with the operator subsystem and measure the effect of a proposed design on the operator. These are detailed in Figure 1.1 and include measures of negative as well as positive outcome and satisfaction. The operator (or user) is central to HFE, so the operator or user subsystem will be discussed first. The operator first perceives the environment—mainly through the visual and auditory senses—and then considers the information, makes a decision, and finally produces a control response. Perception is guided by the operator’s attention. From the millions of bits of information available in the visual field, the operator will, by instinct (or deliberately), select the information most relevant to the task. Some attentional processes are automatic and subconscious (preattentive) and thereby, executed instantaneously (Neisser, 1976). Some processes become automatic with training, while some are deliberate with slow strategies that require more time to analyze. For new or unusual tasks, decision making can be time consuming. The operator will need to interpret the information, the alternatives for action, and to what extent those actions are relevant to achieve the goals of the task. For routine tasks, decisions are more or less automatic and can be accomplished more quickly. In this context, researchers rightly question whether “decision making” is an appropriate term. Because there is not a formal decision, “situated action” may be more appropriate to describe the automaticity in response. The perception and following actions flow—just as for an experienced car driver (for example, see Klein, 1998).


11

The purpose of the operator’s response is to convey information through manual response such as control of a machine (e.g., computer) or a tool (e.g., hammer) or an artifact (e.g., football). It can also be a verbal response such as computer voice control of a machine or verbal message to a co-worker. Modulating Variables Several modulating variables affect task performance. These include: operator needs, attitudes, competence, expertise, motivation, age, gender, body size, and strength. These variables describe the fact that different people react or behave differently. For example, an experienced, competent operator will perceive a task differently than a novice operator will. He will focus on different details of importance, filter irrelevant information, and “chunk” the information into larger units so that it is possible to make faster and more efficient decisions. For example, a football professional who watches a football game will see many details of the game and can predict what will happen. Another example of a modulating variable is body size. Small and large individuals require differently sized workstations. One purpose of anthropometry is to design a workspace to fit operators with different body sizes. Effects of Stress Stress is an important variable that affects perception and decision making as well as response selection. High psychological stress levels usually occur when the time to perform a task is limited, or when too much information must be processed. Under stressful conditions, the bandwidth of attention may narrow, and operators develop “tunnel vision.” It becomes difficult for the operator to consider information “outside” the tunnel. Therefore, the probability of operator error increases. High stress levels lead to increased physiological arousal and can be monitored or measured by using various physiological measures, such as heart rate, EEC, blink rate, and excretion of stress hormones (catecholamines such as nor adrenalin). For example, after a long day of stressful work, the level of noradrenalin in the blood will be higher. The Environment The subsystem Environment is used to conceptualize the task as well as the context in which the task is performed. It could be a steel-worker monitoring an oven; in this instance, the organization of work determines the task allocation: some tasks may be allocated to fellow workers, supervisors, or computers. Task allocation is a central problem in ergonomics: how can one best allocate work tasks among machines and operators so as to realize company goals and individual goals? Task allocation affects how information is communicated between employees and computers, and it also affects systems performance. How can one achieve job satisfaction and productivity at the same time? The operator receives various forms of feedback from his actions, for example, from task performance, co-workers, and management. To enhance task performance, communication, and job satisfaction, feedback must be informative. This means that


12

individuals must receive feedback on how well or poorly they are doing. An operator may perform well and his colleague may perform poorly. Both must be informed and receive feedback on their performance. The ambient environment describes the influence of environmental variables on the operator. For example, a steel-worker is exposed to high levels of noise and heat. This increases physiological arousal and stress, thereby affecting task performance, safety, and satisfaction. The importance of the organizational environment has come to fore in the last few years, resulting in a branch of HFE termed macroergonomics (Hendrick and Kleiner, 2002). In the organizational work environment, company policies with respect to communication patterns, decentralization of responsibilities, and task allocation can have an impact on ergonomics design in terms of appropriate design measures. One consideration is to decide first who should do what and how people should communicate. Following these decisions, design of tasks, displays, and controls for the team can be undertaken. Although macroergonomics is important, there is still not enough research to date. One notable exception is the sociotechnical research by the Tavistock Group in the U.K. in the 1950s, which addressed the design of mining equipment and the effect of task allocation. DeGreene (1973) was quick to comment that human factors research has ignored the motivational, morale, and social stress problems created in systems design. Perhaps the dominant influence of military HFE research in the U.S. has ignored many of the organizational concerns, given that the military organization is well established relative to other organizations in industry (Helander, 1997). For the purpose of completeness, it should be mentioned that, although organizational considerations are important in the work context, they are less important for design of leisure systems and consumer products. This is because such systems and products are used by individuals who typically do not need to consider collaboration and task delegation. The Machine Subsystem The Machine subsystem is broadly conceptualized in Figure 1.1. The term “machine” is in a sense misleading, but it is used here to symbolize any artifact. The machine could be a computer, VCR, or football. The term “controls” in Figure 1.1 denotes machine controls used by the operator. The control of the machine may be taken over by automation and computers through delegation of tasks to autonomous processes. As a result of machine control, a changing state is displayed. The changing state can be seen or heard: a pocket calculator will show the results of a calculation; the melting iron in a steel plant will change temperature and color; a computer will produce a sound; and the toaster will pop out the bread. All of these are examples of displays. They convey visual or auditory information and can be designed to optimize systems performance. It is important to note that the system in Figure 1.1 has feedback. Machine information is fed back to the environment subsystem and becomes integrated with the task. Ergonomics is concerned with dynamic systems. It is always necessary to complete the loop. In this regard, ergonomics is different from other disciplines; in experimental psychology, for example, it is not necessary to study dynamic systems. The system in


13

Figure 1.1 will be used to discuss three major systems goals in ergonomics: safety, productivity, and operator satisfaction. 1.6 The Goals of Ergonomics Ergonomics is rarely a goal by itself; safety, operator satisfaction, and productivity are common goals. Rather, ergonomics is a design methodology that is used to arrive at safety, productivity, and satisfaction. The Goal of Safety The safety status of a system can be assessed by comparing the performance requirements of the environment with the performance limitations of the operator (Figure 1.1). A task will impose a demand for operator attention, and this demand varies in different context and over time. For example, a car driver must look constantly at the traffic when the traffic situation is demanding and adjust the attention level when the situation becomes less demanding. The same driver may become sleepy after driving for a long time and have a low level of attention, but a driver of a race car would not be sleepy—the attention level is always high, thus keeping the driver constantly alert. If the task demands for attention are greater than the available attention, there is an increased risk for accidents or errors. Thus, it is important to understand how the limitations imposed by operator perception, decision making, and control action can be taken into consideration in design, so as to create systems with low and stable performance requirements. Injuries and accidents are relatively rare in the workplace, and therefore it may be difficult to analyze safety. Rather than waiting for accidents to happen, it may be necessary to predict safety problems by analyzing other indicators (or dependent variables) such as operator errors, subjective assessments, and physiological response variables. These measures are indicated in Figure 1.1 under the heading of “measures of negative outcome.” When a system is redesigned to make it safer, several different options may be available, such as: • Tasks between workers and machines/computers can be reallocated; workers may be moved from a hazardous area and automation may take over their jobs. • Work processes and workstation can be redesigned to improve work posture, comfort, and convenience. • The exposure to ambient stressors can be reduced, including noise and heat stress. • Organizational factors can be improved, such as allocation of responsibility and autonomy as well as policies for communication among workers and supervisors. • Design features of a machine can be improved, including changes of controls and displays.


14

The Goal of Productivity In any work context, it is important to improve productivity. This can sometimes be accomplished through thoughtful human factors design. The use of Taylorism has already been mentioned; through measurement of the time involved in assembling a product, Taylorism can split up tasks in many short elements, each performed by a dedicated operator. This can create very short job cycles (of less than a minute); in the long term, there may be poor job satisfaction and biomechanical injuries due to repetitive work. There are better ways to enhance productivity; one can design a system that improves performance affordances. For example, one can design a machine so that handling of the machine controls becomes intuitive, such as high control-response compatibility. One can also design display information using principles of ecological display design, which makes it easy to interpret the information and act quickly (for example, see Vincente, 1999). This means that, through efficient design of the system, the operator can excel in exercising his skills. Such system design makes it possible to perceive quickly, make fast decisions, and exercise efficient control. In general, operators like to excel because the work flows smoothly under full control. The operator will perceive the flow and will think that the situation is enjoyable (Czikszentmihalyi, 1990). In Figure 1.1, several measures of positive outcome are indicated. One can measure productivity, quality, and time on task. One can also ask the operator how well the system works, which is a subjective assessment. These measures are common dependent variables used to measure the productivity of a system. The Trade-Off between Productivity and Safety Ergonomic improvements may focus on reducing operator errors as well as increasing efficiency or speed of operation. It is, however, difficult to improve safety and productivity simultaneously. In general, the greater the speed (of vehicles, production machinery, etc.) is, the less time will be available for the operator to react. A shorter time for operator reaction will compromise safety but increase productivity. Operators therefore have a choice between increased speed and increased accuracy. This is referred to as the speed-accuracy trade-off or SATO (Wickens and Hollands, 2000). Industrial managers often encourage employees to increase speed and accuracy (work more quickly and with better quality). This is contrary to the concept of SATO and difficult or impossible to achieve. In some work environments, such as a process plant or nuclear power plant, safety and production of electricity are two self-evident goals, and together they determine the design of the plant. In this case, safety and productivity must be decoupled through a thoughtful design. It cannot be the case that productivity overrules safety—the consequences could be a disaster. However, safety and quality are coupled. If the number of operator errors can be reduced, safety as well as production quality may be improved. An emphasis on quality and safety is therefore more appropriate than the traditional approach on quantity and quality. The goal, then, is to design a safer environment. This can be done most efficiently if the operator’s reactions and actions can be predicted by


15

the designer. The operator’s perception, decision making, and actions are therefore always central to ergonomics design. The Goal of Operator Satisfaction The last systems goal is operator satisfaction and this will be addressed in a broad sense: from work satisfaction to satisfaction of use. Various aspects of dissatisfaction, such as job dissatisfaction or consumer dissatisfaction, will also be considered. The main point in Figure 1.1 is that satisfaction or dissatisfaction may be predicted by comparing operator needs and attitudes with the performance requirements of the environment and the performance affordances of the machine. Satisfaction and dissatisfaction are mediated through the operators’ needs and attitudes, which can be very different among operators. Some can be fully satisfied by a particular job, while others are dissatisfied. Needs and attitudes also vary substantially between countries and cultures. What are con-sidered workers’ rights in Sweden (e.g., to have a window in the office) are less important in other countries. In Sweden, a lack of window would cause great dissatisfaction because office workers have “acquired” a personal need, but office workers in U.S. or Asia may not think twice. In terms of productivity and safety, the trade-off is obvious: increased productivity, which can lead to an increase in error rate and safety problems. Similarly, a lower error rate and greater safety typically lead to a lower speed of production. In job satisfaction or dissatisfaction, no similar trade-off appears to exist. One would think that a satisfied worker would produce more and a dissatisfied worker would produce less and that a satisfied worker would be safer and a dissatisfied worker may not be so safe. However, extensive research on these issues has demonstrated that no connection is present. Rather, job satisfaction often follows from effective job performance. 1.7 The Measurement Problem In Figure 1.1, several dependent variables such as operator errors, satisfaction, and productivity are defined. These can be used to assess the state of the system and evaluate proposed design solutions. In this section, independent variables are defined. Independent variables are design variables; most of them describe design features that the HFE expert may want to implement or test out. For example, an HFE professional may want to design a warning sign to enhance safety. The design question could be: how many words should be used? If the message is short, operators may not understand; if it is too long, they may not read it, or if the lettering is too small, it may be difficult to see. To investigate this problem, one can set up an experiment in a laboratory in which test subjects can read different warning signs. These types of design questions are easier to research in a laboratory experiment than in the real world (Chapanis, 1967). A laboratory experiment offers better control; one can try out different messages and different sizes of letterings and measure the effect on perception and understanding. Unfortunately, a laboratory study misses out on “ecological validity.” That is, what was found in the laboratory may not be true or applicable in the real world. For example, consider a construction site. A safety sign developed in a laboratory may turn out to be barely


16

visible in the cluttered environment of a construction site; also, environmental stressors (e.g., noise and heat) may distract the operator. Therefore, the sign has little effect and utility. Another problem relates to a conceptual gap between some of the dependent variables and the design variables. Consider as an example the measurement of stress in traffic. Assume that we use galvanic skin response (GSR) to measure stress. It can be argued that a high GSR is bad because it is likely that some drivers will be overly stressed and perform worse. A moderate level of GSR is good because it means that the driver is alert. The question is then: how can one distinguish a moderate level of measurement from a high level? For GSR, there are no absolute levels—the base line shifts constantly. It is different with heart rate because one can simply measure the number of beats per minute. Should one then use heart rate instead to measure stress in traffic? Probably not, because an elevation in heart rate may be due to the physical work in turning the steering wheel rather than the mental stress. Many similar questions exist: will slow, rather than quick, decision making improve the quality of decisions and reduce errors? Will quick movements of the steering wheel in a car, rather than slow movements, improve safety? Do moderate levels of stress lead to greater quality in manufacturing? Such questions are sometimes very difficult to prove in research. It is usually accepted that improved visibility, fast decision making, highfrequency steering wheel movements, and moderate stress levels are beneficial. However, in many cases, the correct answer depends on the situation; there are trade-offs and exceptions. An experienced HFE professional will know how to evaluate different situations. Methods, Measurements, and Procedures in Human Factors and Ergonomics To collect data, to evaluate situations, and to design systems, many new methods and procedures have been developed in the human factors arena. Although most of these methods are unique to HFE, some are also used in industrial psychology and in systems engineering. Table 1.2 provides a list of methods without any explanation of how they can be used. The main purpose here is to illustrate that the number of methods is very large. Detailed information is given in Chapanis (1996), Wilson and Corlett (2002), and Salvendy (1997). Human factors experts will know the method to select.

TABLE 1.2 One Hundred Methods to Collect and Analyze Data in Human Factors and Ergonomics Accident analysis Activity analysis Anthropometric analysis/design Biomechanical analysis Body rhythms and shift work design Checklist analysis Climate analysis Cognitive abilities testing Cognitive systems design Cognitive task analysis

Multimedia design Natural language interface design Operational sequence analysis Operator performance assessment Organizational analysis Organizational design Performance measures (time and error) Performance ratings Physical work load assessment Predetermined time analysis


17

Production systems analysis Cognitive walkthrough Psychomotor proficiency testing Cognitive work analysis Psychophysics and scaling Comfort rating Psychophysiological measurements Communication analysis Questionnaire design Cost/benefit analysis Rapid prototyping Critical incident technique Repetitive motion injury assessment Decision support system design Scenario-based design Decision/action analysis Screen design Design reviews Direct observation/activity analysis Selection testing Simulation design/evaluation Discomfort rating Standards and guidelines for design Environmental sampling Stress measurement Error analysis Survey design Error classification Systems analysis Experimental design Systems safety Failure mode and effects analysis Task analysis Fatigue measurement Task performance measures Fault tree analysis Human error rate prediction (therp) Function allocation Thermal stress measurement Goals-means task analysis Time and motion study GOMS analysis Time lapse photography Hazard analysis Training needs assessment Human reliability assessment Usability analysis Human-computer interaction Usability engineering Illumination measurement Usability testing Information analysis User log books Information visualization User population definition Injury analysis Verbal protocol analysis Intervention studies Vibration measurement Interview technique Video recording Job motivation assessment Virtual environment design Job satisfaction measurement Visibility/legibility analysis Kansei engineering Vision/hearing testing Link analysis Visual performance assessment Macroergonomics Manual materials handling assessment Walkthrough analysis Work condition evaluation Mental model assessment Working posture analysis Mental workload assessment Workload analysis Menu design Workspace design Mockup design/analysis


18

FIGURE 1.2 Procedure for design and redesign of a system. 1.8 Analyzing Design Activity Many HFE specialists are involved in design; it is therefore of interest to analyze the steps that lead to a new design. A procedure for design activity is given in Figure 1.2. Systems goals for design are formulated based on requirements by users, markets, or organizations. User and market requirements specify the types of products that should be manufactured. User and organizational requirements may specify how a job should be designed. There may also be legal requirements, for example, that a robotics workplace should be safe, or company requirements that certain existing machines must be used for production. The latter requirements are usually handled as constraints in design, rather than goals (Suh, 1990). Constraints are different from goals; goals define the design space whereas constraints delimit the design space by invalidating certain designs. Based on the systems goals, functional requirements are specified, and the task is then to design a new system, artifact, or job that will satisfy functional requirements. The two main activities in design work are synthesis and analysis. The synthesis stage is when designers use their knowledge and experience to come up with a design solution that will satisfy the functional requirements. Thus, it is a creative task. Studies of designers’ decision making have produced some interesting results. Goel and Pirolli (1992) investigated engineers’ design decisions for design of new products and found that only 2% of the decisions were logical, in the sense that B follows from A. The remaining 98% were decisions based on associations and experiences. This means that designers will first try to apply what they already know and then see if it works. This trial-anderror-based procedure resembles case-based reasoning; that is, take a previous design solution or experience, modify it slightly so that it can fit the present scenario, then evaluate the results. This is a rather erratic process and has been validated by other researchers. Guindon (1990) found that software programmers tend to jump back and forth between different tasks following their associations. The situation was the following: the programmer would be writing code for solving a problem; he would then (out of the blue) think about


19

a good solution for another problem and, as a result, would leave the first (still unsolved) problem and focus on the second. Guindon termed this behavior “situated cognition”; programmers deal with the problem with which it is practical to deal at any given time. It would take more effort to write down the solution to the second problem, so the first problem is delayed until later. There is no predetermined top-down order. Seemingly, such irrational process, driven by the designer’s associations and evaluations of trade-offs, may be the basis for creativity. New combinations are generated and evaluated immediately by the designer and then abandoned or retained. The associative and negotiable nature of creative cognition is probably why artificial intelligence (AI) has failed. Design-based AI cannot consider associations other than those based on logic and has a poor sense of evaluation (Gero, 1990). To be successful in design, the ergonomist must have knowledge and experience that can produce good design solutions. Moray (1994) argued that ergonomists must have an interdisciplinary background. For example, to develop good design solutions in a manufacturing plant, the ergonomist should also be an engineer. Assume that a production machine has caused repetitive motion injuries for the machine tenders, and the company is considering buying a replacement machine. An occupational nurse, who is an expert at diagnosing ergonomic injuries, could not predict if the ergonomic as well as the industrial production problems would be solved by a new machine (Helander, 1995). To deal with such design problems, an ergonomist needs broad background knowledge. To suggest appropriate design solutions, a combination of HFE knowledge and domain knowledge is required. Another example is from human-computer interaction. In a review of usability studies of human-computer interfaces, Landauer (1995) claimed that, on average, the interface had 40 usability bugs (the range of variation for different interfaces was 17 to 140). During a test session with the software, a domain expert can, on average, identify 20% of the bugs. An HFE (or HCI) expert will identify 40%, and an individual who is a domain and HFE expert will identify 60% of the bugs. In short, a combination of human factors and domain expertise is therefore desirable in problemsolution analysis. It takes much experience to come up with a good design solution (called design synthesis). On the other hand, it is fairly easy to evaluate a new design because analytical methods can be proceduralized and computerized. Laboratory studies or studies of realworld performance can be conducted. Typically, the types of dependent variables referred to in Figure 1.1 are used for such evaluation: human performance data, physiological data, and subjective assessments. Based on the outcome of the analysis, an improved design may be proposed, the purpose of which is to improve operator performance and safety. The system is then implemented. Implementation gives an opportunity to collect real data in the real setting for identifying usability problems as well as new customer requirements. A designer will encounter situations in which the customer requirements have not been identified. This is particularly common in software design in which a company is marketing a new computer program. The market for the program has not been tested and the company is waiting for the sales figures and customer comments on the software. For the second version of the software, the company is armed with better estimates of customer requirements and can reformulate the functional requirements and improve the design solution. To compensate for the time delay, the company can perform usability


20

testing of partial or rapid prototypes, but realistic data will only be available when the product is available on the market. Then the feedback loop in Figure 1.2 is closed. 1.9 Standards in Human Factors and Ergonomics In forensic activity, it is crucial to keep track of the various HFE standards that apply to design. About 150 standards documents have been issued by the International Standards Organization (ISO), as listed in Appendix 1.1. These standards were selected based on the criteria that they have the words “ergonomics” or “human factors” in the title or in the text. Most of these standards deal with HFE design; the remainder reference ergonomics design problems and recommend other relevant ISO standards. International Standards Organization ISO standards are important in all countries around the world and can readily be referred to in legal matters; information about ISO standards is available at http://www.iso.org./ ISO standards often take precedence over local country standards, such as those issued by ANSI in the U.S. or by other countries. They are important for good reasons: they are developed by international teams of experts and they have been approved by several countries. The ISO standards regulate HFE design in several different areas, including: • General ergonomics, e.g., anthropometry • Principles for human-centered design • Design of signals and controls and displays • Design of systems for speech communication • Ergonomic principles related to mental workload • Measurement of heat stress and cold stress • Ergonomic design for the safety of machinery • Ergonomic design of control centers and evaluation of control rooms • Evaluation of static working postures • Principles of visual ergonomics—indoor lighting • Ergonomic requirements for office work with visual display terminals • Hardware requirements of visual display, colors, keyboard, and input devices • Software requirements—usability, dialogue principles, dialogues of menu, command, form filling, and direct manipulation • Ergonomic requirements for work with visual displays based on flat panels • Ergonomics in manual materials handling, including lifting and carrying There are also many important ISO standards in safety, only a few of which are referred to in Appendix 1.1. Standards Development in the U.S. In the U.S., interest in HFE standards has been great, particularly in the military community. The best known is “Design Criteria Standard—Human Engineering,” referred to as MIL-STD 1472F. This document is used in the procurement of military


21

equipment. The Handbook for Human Engineering Design Guidelines (MIL-HDBK759C) is also important. These and other military standards documents can be downloaded from: http://iac.dtic.mil/hsiac/Health_Hazards.htm. In recent directives, the U.S. military refers to ISO and U.S. standards, which are given priority over the MIL standards. Many documents have been published by other standards-developing agencies in the U.S. These include: • The Occupational Safety and Health Act of 1970 (http://www.osha.gov/) • The Americans with Disabilities Act (ADA) of 1990 (http://www.access.gpo.gov/) • The California Ergonomics Standard of 1997, which addresses primarily the prevention of repetitive motion injuries (RMI) (http://www.dir.ca.gov/) The development of standards in the U.S. has had its ups and downs, and often seems to be driven by political arguments rather then measures of benefit-cost. The U.S. Ergonomics Program Standard developed under the Clinton administration was repealed by the Bush administration. Similarly, the Ergonomics Rule implemented in Washington State in 2000 was removed in 2003. The American National Standards Institute ANSI Z365 committee was formed in 1991 with the objective of developing a standard related to musculoskeletal injuries. The National Safety Council withdrew its role as chair for this committee in 2003, and the committee’s work seems to be in limbo. Industry appears to have realized that repetitive motion injuries are not important to standardize—they seem to be less important than ergonomic injuries caused by manual materials handling. Under the umbrella of the American National Standards Institute, several successful developments have occurred. The first was ANSI/HFS 100–1988, American National Standard for Human Factors Design of Visual Display Terminals. This document inspired several ISO standards, in particular a series of documents: ISO 9241–1 to ISO 9241–8. The Z10 Committee, Occupational Safety and Health, was formed in 2001. It has the objective of developing a standard for management of occupational safety and health. This committee recently delivered a draft standard (American Industrial Hygiene Association, 2004). European Standards European standards are developed by the European Union. In selecting areas for standardization, the E.U. collaborates with ISO. Members of the E.U. will support important ISO efforts but take their own initiatives in areas in which ISO has not yet proposed a standard. Directive 89/39/EEC, given in 1989, regulates workplaces, use of work equipment, use of personal protective equipment, work with display screen equipment, safety signs, plus a few other activities (europe.osha.eu.int/legislation/directives/al.php3). Three particularly important safety standards have been developed: • EN 1005–1:2001 Safety of machinery. Human physical performance. Terms and definitions • EN 1005–2:2003 Safety of machinery. Human physical performance. Manual handling of machinery and component parts of machinery


22

• EN 1005–3:2002 Safety of machinery. Human physical performance. Recommended force limits for machinery operation Standards Development in Other Countries Many other countries have developed HFE standards, including most countries in Europe and several countries in Asia, plus Australia and New Zealand. Of particular interest are the “six pack” regulations developed by the Health and Safety Executive in the U.K. This set of standards gives sound and practical advice for design of working environments (http://www.hsebooks.co.uk/books/static/whatsnew.htm #SixPack). 1.10 Conclusion This chapter has given an overview of the historical and conceptual background of human factors and ergonomics. Systems design remains an essential methodology for the analyst/designer. This design serves the purpose of clarifying important dependent variables and interactions in design. If possible, it would be necessary to conceptualize and predict cause-effect relationships. A large battery of methods to use to analyze and/or design systems and artifacts exists. Design constraints will usually determine the method to choose. The ergonomics standards published by the International Standards Organization and other standardization organizations are based on thorough research and international consensus. These are very important in the forensic perspective. The HFE profession is driven by design requirements from users, markets, industries, organizations, and governments. HFE must be able to respond quickly to the changing needs of society. Training programs in HFE must be able to incorporate new areas of interest. Certification programs for HFE professionals must be flexible enough to reconsider changes in current needs, and teaching programs must incorporate new knowledge. Formal education in HFE is required in order to understand how methods for analysis and design of HFE systems can be used. For forensic purposes it is most appropriate to refer to HFE experts who have been certified. In the U.S., certification is handled by the Board of Certification in Professional Ergonomics (BCPE; see http://www.bcpe.org/) under the auspices of IEA. In European countries, many of the ergonomics. societies can certify members under the auspices of CREE (www.eurerg.org/aboutCREE.htm) and IEA. Acknowledgment I acknowledge the help of Halimahtun M.Khalid for her comments on and editorial input into this chapter.


23

References American Industrial Hygiene Association. (2004). BSR AIHA Z10–200x, Occupational Health and Safety Management Systems. Fairfax, VA: American Industrial Hygiene Association. Braidwood, R. (1951). Prehistoric Men . Chicago, IL: Natural History Museum. Brown, O. Jr., Noy, I., and Robertson, M. (1995). Special survey of IEA Federated Societies. Santa Monica, CA: International Ergonomics Association, c/o Human Factors and Ergonomics Society. Chapanis, A. (1967). The relevance of laboratory studies to practical situations. Ergonomics , 10(5), 557–577. Chapanis, A. (1995). Ergonomics in product development: a personal view. Ergonomics , 38, 1625–1638. Chapanis, A. (1996). Human Factors in Systems Engineering . New York: John Wiley & Sons. Chengalur, N., Rodgers, S.H., and Bernard, T.E. (2004). Kodak’s Ergonomics Design for People at Work . New York: John Wiley & Sons. Childe, G. (1944). The Story of Tools . London: Cobbot. Christensen, J.M. (1962). The evaluation of the systems approach in human factors engineering. Hum. Factors , 4(1), 7–22. Czikszentmihalyi, M. (1990). Flow: the Psychology of Optimal Experience . New York: Harper Collins. DeGreene, K.B. (1973). Sociotechnical Systems . Englewood Cliffs, NJ: Prentice Hall. Eriksson, R. (1976). Personal communication. International Labor Organization, Geneva, Switzerland. Egan, D.E. (1988). Individual differences in human-computer interaction. In M.G.Helander (Ed.), Handbook of Human Computer Interaction . Amsterdam, The Netherlands: New Holland. Gero, J. (1990). University of Sydney. Personal communication. State University of New York at Buffalo, Buffalo, New York, April 1990. Goel, V. and Pirolli, P. (1992). The structure of design problem spaces. Cognitive Sci. , 16, 35–429. Guindon, R. (1990). Designing the design process: exploiting opportunistic thoughts. Hum. Computer Interaction , 5, 305–344. Helander, M.G. (1995). A Guide to the Ergonomics of Manufacturing . London: Taylor & Francis. Helander, M.G. (1997). Forty years of IEA: some reflections on the evolution of ergonomics. Ergonomics , 40, 952–961. Hendrick, H.W. and Kleiner, B.W. (2002). Macroergonomics. An Introduction to Work Systems Design . Santa Monica, CA: The Human Factors and Ergonomics Society. Hendrick, H.W. (1995). Future directions in macroergonornics. Ergonomics , 38, 1617–1624. International Ergonomics Association (2004). Downloaded from http://www.iea.cc/. Karwowski, W. (1991). Complexity, fuzziness, and ergonomic incompatibility issues in the control of dynamic work environments. Ergonomics , 34, 671–686. Khalid, H.M. (2003). ASEAN ergonomics and prioritizing research. Proc. 15th Triennial Congr. Int. Ergonomics Assoc. Seoul: The Ergonomics Society of Korea (CD-ROM). Klein, G. (1998). Sources of Power. How People Make Decisions . Cambridge, MA: MIT Press Konz, S. and Johnson, S. (1990). Work Design: Industrial Ergonomics . 5th ed. Scottsdale, AZ: Holcomb Hathaway Publishers, Inc. Kroemer, K., Kroemer, H., and Kroemer-Elbert, K. (1994). Ergonomics: How to Design for Ease and Efficiency . Englewood Cliffs, NJ: Prentice Hall. La Mettrie, J.O. (1748). L’homme Machine . Leyden, The Netherlands: Elie Luzac. Available in English at http://www.%20cscs.%20umich.%20edu/cr%20shalizi/LaMattrie%20ie/Machine/. Landauer, T. (1995). The Trouble with Computers . Cambridge, MA: MIT Press.


24

Licht, D.M., Polzella, D.J., and Boff, K.R. (1991). Human factors, ergonomics and human factors engineering. An analysis of definitions (Report No 89–01). Wright-Patterson AFB, OH: CSERIAC Program Office. Long, J. and Dowell, A. (1989). Conceptions for the discipline of HCI: craft, applied science and engineering. Proc. BCS HCI SIG Conf. , 9–32. UK: University of Nottingham. Luczak, H. (1997). Arbeitswissenschaft in internationalen Vergleich. In H.Luczak and E.Volpert, Eds., Handbuch der Arbeitwissenschaft . Stuttgart: Verlag Schaeffer-Poeschel. Neisser, U. (1976). Cognition and Reality . San Francisco, CA: Freeman. Meister, D. (1985). Behavioral Analysis and Measurement Methods . New York: John Wiley & Sons. Moray, M. (1994). The future of ergonomics—the need for interdisciplinary integration. Proc. IEA Congr. , 1791–1793. Santa Monica, CA: The Human Factors and Ergonomics Society. Ramazzini, B. 1940 (1713). In Wright, W. (translation). The Diseases of Workers . Chicago, IL: University of Chicago Press. Rosenbrock, H.H. (1983). Seeking an appropriate technology. Proc. IFAC Symp. Syst. Approach Appropriate Technol. Transfer , 127–134. Vienna Austria: IFAC. Salvendy, G. (1997). Handbook of Human Factors and Ergonomics . New York: WileyInterscience. Shaw, L. and Sichel, H. (1975). Accident Proneness . New York: Pergamon Press. Smith, K.U. (1988). Origins of ergonomics and the International Ergonomics Association. Human Factors Soc. Bull , 31(1). Santa Monica, CA: Human Factors and Ergonomics Society. Suh, N.P. (1990). The Principles of Design . New York: Oxford University Press. Vanwonterghem, K. (1994). Personal communication. Brussels, Belgium. Vincente, K. (1999). Cognitive Work Analysis . Mahwah, NJ: Lawrence Erlbaum. Wickens, C.D. and Hollands, J.G. (2000). Engineering Psychology and Human Performance , 3rd ed. New York: Prentice Hall. Wilson, J.R. and Corlett, E.N. (2002). Evaluation of Human Work. A Practical Ergonomics Methodology , 2nd ed. London: Taylor & Francis.

Appendix 1.1 HFE Standards Issued by International Standards Organization ISO Road vehicles 2575:2000 ISO 2631– Mechanical vibration and shock 1:1997 ISO Earth-moving machinery 3411:1995 ISO 3767– Tractors, machinery for agriculture and 1:1998 forestry, powered lawn and garden equipment ISO 3767– Tractors, machinery for agriculture and 2:1991 forestry, powered lawn and garden equipment ISO 3767– Tractors, machinery for agriculture and 3:1995 forestry, powered lawn and garden

Symbols for controls, indicators, and telltales. Cor 1:2001; Amd 1:2001; Amd 4:2001 Evaluation of human exposure to wholebody vibration—part 1: general requirements Human physical dimensions of operators and minimum operator space envelope Symbols for operator controls and other displays—part 1: common symbols Symbols for operator controls and other displays—part 2: symbols for agricultural tractors and machinery Amd 2:1998 additional symbols Symbols for operator controls and other displays—part 3: symbols for powered


equipment ISO 3767– Tractors, machinery for agriculture and 5:1992 forestry, powered lawn and garden equipment ISO Passenger cars 3958:1996 ISO Road vehicles 4040:2001 ISO Agricultural tractors 4253:1993 ISO 4254– Tractors and machinery for agriculture and 1:1989 forestry ISO Hydraulic fluid power 4413:1998 ISO Pneumatic fluid power 4414:1998 ISO 5349– Mechanical vibration 1:2001 ISO Tractors for agriculture 5721:1989 ISO Mechanical vibration and shock 5982:2001 ISO 6242– 1:1992 ISO 6385:2004 ISO 6549:1999 ISO 6682:1986 ISO 6858:1982 ISO 6897:1984

Building construction

lawn and garden equipment Symbols for operator controls and other displays—part 5: symbols for manual portable forestry machinery Driver hand-control reach Location of hand controls, indicators, and tell-tales in motor vehicles Operator’s seating accommodation— dimensions Technical means for ensuring safety— part 1: general General rules relating to systems General rules relating to systems Measurement and evaluation of human exposure to hand-transmitted vibration— part 1: general requirements Operator’s field of vision Range of idealized values to characterize seated-body biodynamic response under vertical vibration Expression of users’ requirements—part 1: thermal requirements

Ergonomic principles in the design of work systems Road vehicles Procedure for H- and R-point determination Earth-moving machinery Zones of comfort and reach for controls. Amd 1:1989 Aircraft Ground support electrical supplies— general requirements Guidelines for the evaluation of the response of occupants of fixed structures, especially buildings and off-shore structures, to lowfrequency horizontal motion (0.063 to 1 Hz) Information and documentation Presentation of catalogues of standards

ISO 7220:1996 ISO Hot environments 7243:1989 ISO 7250:1996 ISO 7397– 1:1993

25

Estimation of the heat stress on working man, based on the WBGT-index (wet bulb globe temperature)

Basic human body measurements for technological design Passenger cars Verification of driver’s direct field of view—part 1: vehicle positioning for static measurement


ISO 7397– 2:1993 ISO 7726:1998

Passenger cars

26

Verification of driver’s direct field of view—part 2: test method Instruments for measuring physical quantities Determination of the PMV and PPD indices and specification of the conditions for thermal comfort Danger signals for public and work areas—auditory danger signals Analytical determination and interpretation of thermal stress using calculation of required sweat rate Combine harvesters—test procedure

ISO 7730:1994

Ergonomics of the thermal environment Moderate thermal environments

ISO 7731:2003

Ergonomics

ISO 7933:1989

Hot environments

ISO 8210:1989 ISO 8230:1997

Equipment for harvesting Safety requirements for dry-cleaning machines using perchloroethylene Forestry machinery Portable chain saws—determination of balance Ship’s bridge layout and associated Requirements and guidelines equipment Mechanical vibration and shock Human exposure—biodynamic coordinate systems Lighting of indoor work places

ISO 8334:1985 ISO 8468:1990 ISO 8727:1997 ISO/CIE 8995:2002 ISO 8996:1990 ISO 9000:2000 ISO 9004:2000 ISO/IEC TR 9126–2:2003 ISO/IEC TR 9126–3:2003 ISO 9241– 1:1997 ISO 9241– 2:1992 ISO 9241– 3:1992 ISO 9241– 4:1998 ISO 9241– 4:1998/Cor 1:2000

Ergonomics Quality management systems Quality management systems Software engineering

Determination of metabolic heat production Fundamentals and vocabulary Guidelines for performance improvements Product quality—part 2: external metrics

Software engineering

Product quality—part 3: internal metrics

Ergonomic requirements for office work with visual display terminals (VDTs) Ergonomic requirements for office work with visual display terminals (VDTs) Ergonomic requirements for office work with visual display terminals (VDTs) Ergonomic requirements for office work with visual display terminals (VDTs)

Part 1: general introduction. Amd 1:2001

Part 2: guidance on task requirements

Part 3: visual display requirements. Amd 1:2000 Part 4: keyboard requirements


ISO 9241– 5:1998 ISO 9241– 6:1999 ISO 9241– 7:1998 ISO 9241– 8:1997 ISO 9241– 9:2000 ISO 9241– 10:1996 ISO 9241– 11:1998 ISO 9241– 12:1998 ISO 9241– 13:1998 ISO 9241– 14:1997 ISO 9241– 15:1997 ISO 9241– 16:1999 ISO 9241– 17:1998 ISO 9355– 1:1999 ISO 9355– 2:1999 ISO 9886:1992 ISO 9920:1995 ISO 9921:2003 ISO 10068:1998 ISO 10075:1991 ISO 10075– 2:1996

27

Ergonomic requirements for office work with Part 5: workstation layout and visual display terminals (VDTs) postural requirements Ergonomic requirements for office work with Part 6: guidance on the work visual display terminals (VDTs) environment Ergonomic requirements for office work with Part 7: requirements for display visual display terminals (VDTs) with reflections Ergonomic requirements for office work with Part 8: requirements for displayed visual display terminals (VDTs) colors Ergonomic requirements for office work with Part 9: requirements for visual display terminals (VDTs) nonkeyboard input devices Ergonomic requirements for office work with Part 10: dialogue principles visual display terminals (VDTs) Ergonomic requirements for office work with Part 11: guidance on usability visual display terminals (VDTs) Ergonomic requirements for office work with Part 12: presentation of visual display terminals (VDTs) information Ergonomic requirements for office work with Part 13: user guidance visual display terminals (VDTs) Ergonomic requirements for Part 14: menu dialogues office work with visual display terminals (VDTs) Ergonomic requirements for Part 15: command dialogues office work with visual display terminals (VDTs) Ergonomic requirements for Part 16: direct manipulation dialogues office work with visual display terminals (VDTs) Ergonomic requirements for Part 17: form filling dialogues office work with visual display terminals (VDTs) Ergonomic requirements for the Part 1: human interactions with displays and design of displays and control control actuators actuators Ergonomic requirements for the Part 2: displays design of displays and control actuators Evaluation of thermal strain by physiological measurements Ergonomics of the thermal Estimation of the thermal insulation and environment evaporative resistance of a clothing ensemble Ergonomics Assessment of speech communication Mechanical vibration and shock Free, mechanical impedance of the human handarm system at the driving point Ergonomic principles related to General terms and definitions mental workload Ergonomic principles related to Part 2: design principles mental workload


ISO 10303– 214:2003

Industrial automation systems and integration

ISO 10333– 6:2004 ISO 10472– 1:1997 ISO 10535:1998 ISO 10551:1995 ISO 10968:1995 ISO 11064– 1:2000 ISO 11064– 2:2000 ISO 11064– 3:1999 ISO 11111:1995 ISO 11112:1995 ISO 11199– 1:1999 ISO 11199– 2:1999 ISO 11226:2000 ISO 11228– 1:2003 ISO 11334– 1:1994 ISO 11334– 4:1999 ISO 11393– 4:2003 ISO 11399:1995 ISO 11428:1996 ISO 11429:1996 ISO 11553:1996 ISO 11680– 1:2000

Personal fall-arrest systems

28

Product data representation and exchange—part 214: application protocol: core data for automotive mechanical design processes Part 6: system performance tests

Safety requirements for industrial Part 1: common requirements laundry machinery Hoists for the transfer of disabled Requirements and test methods persons Ergonomics of the thermal Assessment of the influence of the thermal environment environment using subjective judgment scales Earth-moving machinery Operator’s controls Ergonomic design of control centers Ergonomic design of control centers Ergonomic design of control centers Safety requirements for textile machinery Earth-moving machinery

Part 1: principles for the design of control centers

Walking aids manipulated by both arms Walking aids manipulated by both arms Ergonomics

Requirements and test methods—part 1: walking frames Requirements and test methods—part 2: rollators

Ergonomics

Manual handling—part 1: lifting and carrying

Part 2: principles for the arrangement of control suites Part 3: control room layout. Cor 1:2002

Operator’s seat—dimensions and requirements

Evaluation of static working postures

Walking aids manipulated by one Requirements and test methods—part 1: elbow arm crutches Walking aids manipulated by one Requirements and test methods—part 4: walking arm sticks with three or more legs Protective clothing for users of Part 4: test methods and performance handheld chain saws requirements for protective gloves Ergonomics of the thermal Principles and application of relevant environment international standards Ergonomics Visual danger signals—general requirements, design and testing Ergonomics System of auditory and visual danger and information signals Safety of machinery Laser processing machines—safety requirements Machinery for forestry

Safety requirements and testing for pole-mounted powered pruners—part 1: units fitted with an integral combustion engine


ISO 11680– 2:2000

Machinery for forestry

ISO 11681– 2:1998 ISO 11690– 1:1996

Machinery for forestry

ISO 11748– 2:2001 ISO 11806:1997 ISO 11850:2003 ISO 12100– 2:2003 ISO/IEC 12119:1994 ISO/IEC 12207:1995 ISO 12214:2002 ISO 12239:2003 ISO 12648:2003 ISO 12894:2001 ISO 13090– 1:1998

Acoustics

Road vehicles Agricultural and forestry machinery Machinery for forestry Safety of machinery Information technology Information technology Road vehicles Fire detection and fire alarm systems Graphic technology Ergonomics of the thermal environment Mechanical vibration and shock

ISO 13091– 1:2001

Mechanical vibration

ISO 13091– 2:2003

Mechanical vibration

ISO 13406– 1:1999

Ergonomic requirements for work with visual displays based on flat panels Ergonomic requirements for work with visual displays based on flat panels Human-centered design processes for interactive systems Protective clothing

ISO 13406– 2:2001 ISO 13407:1999 ISO 13688:1998 ISO

Ergonomics of the thermal

29

Safety requirements and testing for pole-mounted powered pruners—part 2: units for use with a backpack power source Portable chain saws—safety requirements and testing—part 2: chain saws for tree service Recommended practice for the design of low-noise workplaces containing machinery—part 1: noisecontrol strategies Technical documentation of electrical and electronic systems—part 2: documentation agreement Portable handheld combustion engine-driven brush cutters and grass trimmers—safety Self-propelled machinery—safety requirements Basic concepts, general principles for design—part 2: technical principles Software packages—quality requirements and testing Software life cycle processes. Amd 1:2002 Direction-of-motion stereotypes for automotive hand controls Smoke alarms Safety requirements for printing press systems Medical supervision of individuals exposed to extreme hot or cold environments Guidance on safety aspects of tests and experiments with people—part 1: exposure to whole-body mechanical vibration and repeated shock Vibrotactile perception thresholds for the assessment of nerve dysfunction—part 1: methods of measurement at the fingertips Vibrotactile perception thresholds for the assessment of nerve dysfunction—part 2: analysis and interpretation of measurements at the fingertips Part 1: introduction

Part 2: ergonomic requirements for flat-panel displays

General requirements Vocabulary and symbols


13731:2001 environment ISO/TS Ergonomics of the thermal 13732–2:2001 environment

30

Methods for the assessment of human responses to contact with surfaces—part 2: human contact with surfaces at moderate temperature Safety-related parts of control systems—part 1: general principles for design Two-hand control devices—functional aspects and design principles Pressure-sensitive protective devices—part 1: general principles for design and testing of pressure-sensitive mats and pressure-sensitive floors Content and drafting of a functional specification

ISO 13849– 1:1999 ISO 13851:2002 ISO 13856– 1:2001

Safety of machinery

ISO 13879:1999 ISO 13880:1999 ISO 14123– 2:1998

Petroleum and natural gas industries Petroleum and natural gas industries Safety of machinery

ISO/IEC 14598–4:1999 ISO 14644– 4:2001 ISO 14738:2002 ISO 14740:1998

Software engineering

ISO 14915– 1:2002 ISO 14915– 2:2003 ISO 14915– 3:2002 ISO 14969:1999 ISO 15005:2002

Software ergonomics for multimedia user interfaces Software ergonomics for multimedia user interfaces Software ergonomics for multimedia user interfaces Quality systems

ISO 15007– 1:2002

Road vehicles

ISO 15008:2003

Road vehicles

ISO 15027– 3:2002 ISO 15190:2003

Immersion suits

Medical devices—guidance on the application of ISO 13485 and ISO 13488 Ergonomic aspects of transport information and control systems—dialogue management principles and compliance procedures Measurement of driver visual behavior with respect to transport information and control systems—part 1: definitions and parameters Ergonomic aspects of transport information and control systems—specifications and compliance procedures for in-vehicle visual presentation Part 3: test methods

Medical laboratories

Requirements for safety

Safety of machinery Safety of machinery

Cleanrooms and associated controlled environments Safety of machinery Forest machinery

Road vehicles

Content and drafting of a technical specification Reduction of risks to health from hazardous substances emitted by machinery—part 2: methodology leading to verification procedures Product evaluation—part 4: process for acquirers Part 4: design, construction and start-up Anthropometric requirements for the design of workstations at machinery Backpack power units for brush cutters, grass trimmers, pole cutters and similar appliances—safety requirements and testing Part 1: design principles and framework Part 2: multimedia navigation and control Part 3: media selection and combination


31

ISO/IEC 15288:2002 ISO/IEC 15411:1999 ISO 15534– 1:2000

Systems engineering

System life cycle processes

Information technology

Segmented keyboard layouts

Ergonomic design for the safety of machinery

ISO 15534– 2:2000 ISO 15534– 3:2000 ISO 15535:2003

Part 1: principles for determining the dimensions required for openings for whole-body access into machinery Part 2: principles for determining the dimensions required for access openings Part 3: anthropometric data

Ergonomic design for the safety of machinery Ergonomic design for the safety of machinery General requirements for establishing anthropometric databases Urine-absorbing aids General guidance on evaluation

ISO 15621:1999 ISO 15667:2000 ISO/TS 16071:2003 ISO 16091:2002 ISO 16100– 1:2002 ISO 16273:2003 ISO/TR 16982:2002 ISO 17287:2003 ISO 17776:2000 ISO/IEC 18019:2004 ISO/IEC 18035:2003 ISO/PAS 18152:2003 ISO/TR 19358:2002

Acoustics

Guidelines for noise control by enclosures and cabins

Ergonomics of human- Guidance on accessibility for human-computer interfaces system interaction Space systems Integrated logistic support Industrial automation systems and integration Ships and marine technology

Manufacturing software capability profiling for interoperability—part 1: framework Night vision equipment for high-speed craft—operational and performance requirements, methods of testing, and required test results Ergonomics of human- Usability methods supporting human-centered design system interaction Road vehicles Ergonomic aspects of transport information and control systems—procedure for assessing suitability for use while driving Petroleum and natural Offshore production installations—guidelines on tools and gas industries techniques for hazard identification and risk assessment Software and system Guidelines for the design and preparation of user engineering documentation for application software Information technology Icon symbols and functions for controlling multimedia software applications Ergonomics of human- Specification for the process assessment of human-system system interaction issues Ergonomics Construction and application of tests for speech technology


32

IEC 60601– Medical Part 1–8: General requirements for safety—collateral standard: 1–8:2003 electrical general requirements, tests and guidance for alarm systems in equipment medical electrical equipment and medical electrical systems Notes: The words “ergonomics” or “human factors” appear in titles or texts of these standards. The numbers indicate the number of the standard and the latest publication date. Cor=correction; Amd=amendment; TR=technical report; TS = technical specification; CIE=International Commission on Illumination; IEC=International Electrotechnical Commission; PAS=publicly available specification.

2 Preparing and Presenting Evidence in Court Daniel A.Bronstein Michigan State University 0-415-28870-3/05/$0.00+$1.50 © 2005 by CRC Press

2.1 Objectives and Scope of This Chapter This chapter will operate under the assumption that it has already been decided that you can qualify to testify under the standards used in the jurisdiction, for example, under the test of Daubert v.Merrill Dow Pharmaceuticals, 1993, in the United States federal courts. What we will discuss in this chapter is how to organize your testimony, prepare exhibits and use illustrative materials, and make a good impression on the fact finder. We will also discuss some things to watch out for on cross-examination. 2.2 Using Exhibits and Demonstrations One should always keep in mind the difference between things introduced into evidence at a trial and things used to illustrate points to be made. The two formal rules regarding things that are actually introduced into evidence and marked as exhibits are: the exhibits must be authenticated and they must be verified. Authentication is proving that the exhibit actually is what it claims to be. This means showing that the six-page report from the laboratory, for example, really is the report from the laboratory. This is a legal type of issue and is for the attorney to handle, not you as a witness. Verification, on the other hand, deals with the accuracy of the content of the exhibit and is a factual issue. As such, it is one you may have to handle. The questioning would go somewhat along the following lines: Q: Is this the report regarding the tests you ran on the control console? A: Yes. Q: And these tests were performed by you or under your supervision? A: Yes. Q: And the report is accurate regarding your findings? A: Yes.


34

The reason for these formal requirements for things introduced into evidence is that, with the approval of the judge, such exhibits can be taken into the jury room for examination during deliberations. Thus, they can be particularly subject to abuse because the jury can examine them at its leisure. Things you do to help the fact finder understand your testimony, on the other hand, are not allowed into the jury room. All the jury has is its collective memory of what you did or showed while testifying. Thus, there are few restrictions on demonstrations or things used for illustrative purposes. Legally, all that is necessary is the following exchange: Q: Do you have something that would help explain your testimony? A: Yes, I have two signs, one with white letters on a green background and one with black letters on a white background so that you can see how much easier it is to read the first. Psychologists tell us that things that people see and hear make more of an impact than things that are only heard. For this reason, it is very useful to have things to show during the course of your testimony. The limits are those imposed by the physical layout of the courtroom and by the judge’s patience. Of the two, the physical limitations of the courtroom are frequently the greater problem. In an “old style” courtroom, it is very difficult to use such things as slides, movies, videos, or computer animations. The room must be darkened, equipment set up, power lines run, testing performed, etc. If you and the attorney agree that the things of this type that you wish to use are really valuable to your testimony, the attorney will probably try to get you scheduled at the beginning of the day or after the lunch break so that the equipment can be set up and tested without disrupting the courtroom. It is very distracting to everyone if you must interrupt the middle of your testimony to set this equipment up. In a “modern” courtroom, on the other hand, most of the equipment will already be in place and connected. All you will need to do is provide the material. This is an issue you must discuss with the attorney in the case long before trial (and he may discuss with the opponent and the judge), because it would be a waste of time and money to prepare materials that are never used due to the physical limitations of the court facilities. This use of illustrative materials can be most helpful for a clear presentation of your points. In a case to which I frequently refer, Rogers v.Raymark Industries (1991), a doctor was arguing that the plaintiff did not have pleural plaques in his lungs because pleural plaques are three-dimensional and the spots on the patient’s x-rays only showed in the front-to-back pictures, not in the side-to-side pictures. To show what he meant, he brought into court and showed x-rays of some other, unidentified patient in which the spots were visible in both views. I am sure this really helped explain his point. If the equipment is simple and portable enough, you can bring it into the courtroom and perform a demonstration right there. In any case, you should probably show some pictures of the equipment and explain how it is operated. A forensic chemist, for example, can bring in a picture of the GC/MS and explain where the sample is inserted and where the results appear, and then say, “When I inserted the sample and the standard, I got the following printouts.” Then he can show the printouts and explain how the peaks correspond, thus making visual his conclusion that the control substance is present in the tested sample.

Preparing and presenting evidence in court

35

The limitations on visual presentations, then, are not really difficult. As long as you and the attorney discuss things in advance, the only limitations are those that the judge might impose due to problems with the courtroom. 2.3 Using “Learned Treatises” This is a problem in the law of hearsay. After all, the author of the book or article is not present to be cross-examined. The rules of evidence on this topic vary by jurisdiction. In many U.S. states and the federal courts, you cannot quote from treatises and journal articles in your direct testimony; in others you can. In either case, however, you can always testify that you relied on a particular item in the professional literature in performing your analysis; the limitation, if it exists in your jurisdiction, is simply whether you can actually quote from it. We will discuss the use of such professional literature in cross-examination later in this chapter. 2.4 Organizing Your Testimony The first thing you need to do when organizing your testimony is to decide which are the strongest and weakest points in your presentation. You should be aware that each point can be strong in two dimensions. It can be strong in that it is a major support to your ultimate conclusion or weak in that it does not greatly contribute to your conclusion. It can also be strong in that there is a great deal of laboratory or other evidentiary support for it or weak in that there is not a great deal of support for it in the evidence. It is necessary for you to decide which are your strong points, which are your weak points, and which points are between them. Let us assume that seven subsidiary points lend support to your conclusion. For example, look at Figure 2.1. We would all agree that item number one is the strongest point; it lies on the 45° line, indicating good support in the evidence and that it is a strong aid to the ultimate conclusion. We would probably also agree that number two is the second strongest point, although it is perfectly possible to argue that numbers three and four should be switched. Items five, six, and seven are clearly significantly weaker than the others. The idea that you should use for organizing your testimony is to try to end with the strongest point you can make. This does not mean that you should simply take the list and invert it because that would not get people’s interest when you first start your testimony. We would like to get them interested by starting with a reasonably strong point, perhaps point three. You then go quickly down to your weakest point and build up to end with your strongest point. For example, you might use the sequence: three, six, seven, five, four, two, one.


36

FIGURE 2.1 The dimensions of strength of points. This, of course, is rather idealized. In the real situation, you may need to cover point four before you can cover point seven because point seven relies upon point four. If this is so, then, in our example, you would invert points six and four. In other words, do not let your striving for effect interfere with the technical accuracy of the testimony you give. Next, you sit down and write the question for the lawyer to ask you that will raise each of the points. After the question, write some short notes for yourself regarding what you intend to offer in your answer to the question. Included in the notes should be small reminders of the documents to which you will refer, the exhibits that you will introduce into evidence, and the things that you will use to illustrate your points. I frequently suggest that these be included in “curly brackets” like this: {letter from OSHA to defendant} or {photo of scene} or {diagram of work station}. After you have done this, it is time to consult with the attorney and get his input. After all, it is his case that you are attempting to support, so he certainly has a say in how your testimony is presented. My preferred method would be face to face, but e-mailing the outline as attachments can also work. Be sure to give careful consideration to what the attorney says, but do not necessarily agree to all of the suggestions. After all, you are the expert and, if you disagree with some of the suggestions, do not hesitate to say so. However, in the end, he will have the final say. After you and the attorney have agreed on your testimony, redo the list of questions and the notes. You will take this on the witness stand with you and the attorney will have a copy. That way both of you will know what you intended to say and, if one of you misses an item, the other can bring it up. Do not worry about taking notes to the witness stand; it does not make you look weak or uncertain. Everybody knows that the President


37

speaks from a teleprompter; the fact that you need notes does not detract from your expert standing. The rules in the U.S. say that the other side is entitled to see anything you take on the stand with you, but this is not important. They have already read your reports and, probably, taken your deposition, so they already know what you are going to say. The last point is more a personal preference of mine than a suggestion, but I do believe it makes a difference. Start out with questions asking for your ultimate conclusion: Q: Do you have an opinion regarding the safety of the design of this control panel? A: Yes. Q: What is your opinion? A: I believe that it was properly designed. This provides the listeners with a “road map”; they know where you are going, even if they do not know the exact route you will take. At the end of your testimony, briefly recap the major points you have made: Q: Now, because you have been testifying for some time today, could you please briefly restate the reasons for your conclusion? A: Yes. This device was designed to enable the operator to switch easily among operations A, B, and C. It was never intended that the operator be able to access functions D or E. The design is better than that of most competitive consoles. If it had been used as designed and if it had not been modified, it would never have allowed the problems encountered in this case to occur. Thus, I believe it was properly designed and manufactured. Again, consult with the attorney, but I believe he will greatly approve of such introductions and summaries. 2.5 The Role of the Expert Witness The most important thing to remember when testifying is that your role is that of an educator. You are there to teach the judge and/or jury enough about your specialty so that they can see that your conclusions are obviously correct and should be believed. This brings up the question “What level class am I teaching?” If you have a bench (judge without jury) trial, then it is safe to assume that the judge has a college education, although you cannot assume anything about his scientific or technical knowledge. That information is available to the attorney, however, who can check the judge’s biography and learn if it contains some scientific content. If it does not, then, obviously, you are teaching to the “educated layman.” If you have a jury trial, then the normal assumption most attorneys make is that you should teach to the secondary school graduate. The basis for this is the belief that any college graduates on the jury will understand why you are teaching at that level, but people who stopped their education at that point might be insulted if you were to teach to


38

a lower level. Again, information about the jury is available to the attorney and you should discuss the issue with him or her before you take the stand. 2.6 Use of Language It is not often stated, even by attorneys who are preparing expert witnesses, but I and many other trial lawyers firmly believe that, when you are on the witness stand testifying, you should use completely different language from that which you use when writing reports. Following are a few examples: • Use the active voice, not the passive voice. The idea is to present yourself to the fact finder as a likable, believable person who also has some specialized knowledge and not to appear as a “walking, talking textbook.” In your written report, you probably said something like “the examination of the control panel revealed that it had been modified to provide extra functions”; however, when you are on the witness stand, you should say, “I looked at the panel and saw that it had been changed.” Notice, in this example, the use of the active voice, the first person, and simpler language. Again, the idea is to appear human. You are an actor in the events you are narrating and should speak that way. • Avoid jargon. Technical terms have little content to persons not intimately familiar with your specialty. Avoid them whenever you can. Hopefully, your attorney will catch it when you slip back into “tech speak” and interrupt to ask you to clarify what you mean. • Use analogies. This is the “painting of verbal pictures” and can get as much attention as the use of demonstrations and illustrations. I still remember one I encountered back in 1968 when I was in practice. An expert defined “one part per million” as “the same as putting one teaspoon of dye into an Olympic swimming pool.” I have never done the calculations to see if that is really correct, but it certainly draws a mental and verbal picture for everyone of how small one part per million is. • Tell stories. By this I mean that you should be writing a detective story. The mystery is “what happened and why?” We know what your conclusion is, but you are giving us the clues and your interpretation of them as you go through your testimony. You can also tell stories to illustrate your points as you go. I did just that in the previous paragraph with the “teaspoon in a pool” example. This helps understanding and also provides a bit of a break from straight technical testimony, which all in the courtroom will appreciate. • Be likeable, honest, and human. Even if the jury cannot understand all of your testimony, if you impress them as a person they would “buy a used car from,” they will give your conclusions greater weight than they would to a “talking computer.”

2.7 Courtroom Behavior The most important thing to do when testifying is to make eye contact with the jury. The same holds true at a bench trial if you are in a “modern” courtroom and can see the judge. Do not get into a conversation with the attorney; you are not attempting to convince him


39

or her. Because none of us really trusts a person who does not make eye contact, you should look at the attorney when he or she asks you a question and then turn your whole body in the witness chair and look at the jury (or judge in a bench trial). When you have finished your answer, turn your whole body again to look at the attorney for the next question. When using exhibits or conducting a demonstration, make certain that you do not block the view of the jury or judge. You should make certain that you stand to the side so that the beautiful thing you prepared can be appreciated by its intended audience. The attorney is not the intended audience, and you can block his or her view without worrying about it. Whenever the opportunity presents itself during your testimony, get out of the witness chair and move around the room a little. You know what advertisers think of “talking heads” on television; they hate them. Even the network TV news anchors get up out of their chairs at some point during their shows. If you are stuck in the same place the entire time, it gets boring for the jury. Talk with the attorney in advance to find places in your testimony during which you can get out of the chair, and then do it. It sounds silly, but clothing can be important. Logically, the validity of the argument is not dependent on the appearance of the person making it. Many judges, however, appear to believe that people who are badly dressed are showing disrespect for the court. Therefore, if you appear in what the judge believes is improper clothing, the judge’s nonverbal clues to the jury will say “ignore this lout.” Proper dress is a solid-color, conservative suit with quiet striped or foulard tie (for men) or blouse (for women). Pantsuits are perfectly acceptable for women. 2.8 Cross-Examination The fundamental rule for you to remember when being cross-examined is that you are the expert; the lawyer asking the questions is not. This has several important implications. First, if the cross-examiner manages to “back you up against a wall,” you always have the safe exit of “that is my expert opinion.” Second, only you know when you have finished answering the question. On television we see lawyers saying, “Thank you witness, but that is enough.” You, as an expert, have the right to say, “I am sorry, but I do not believe I have finished answering,” and then go on to finish what you wanted to say. The judge, of course, has the right to cut in and say, “Well, I think you are done,” but I have never heard of this happening. If you have more you want to say, do not let the lawyer cut you off. It is very important that you never agree in advance to respond to questions with only a “yes” or a “no.” This is an old trick of lawyers. If asked to do this, you can reasonably say, “Well, until I hear the questions, I do not believe I can agree to that; I do not know what you are going to ask me.” To return to “learned treatises,” beware of the lawyer quoting out of context. The article may really say, “One should always be aware of the small but nevertheless real possibility of A,” but all the lawyer reads to you is “one should always be aware of A.” If what is read to you seems very unlikely or unreasonable, you have the right to ask to see the quoted publication. Then you can point out that the quote was taken out of context.


40

Similarly, you can refute the quoted material by saying, “Well, we used to think that, but more recent research has shown…” or, you might say, “Well, Dr. Ivory is sort of out there alone in that opinion; the majority view, as most recently summarized in the review article by Professor Tower, is that…” or, again, “Well, the journal you are quoting from is not regarded particularly well in our field; it would be classified as a fourth-tier publication and very few people pay attention to what appears there.” However, this does mean that you need to keep up with the top journals in your field and be aware of the most recent developments. In all cases, remember that you are the “living, breathing” person the jury sees. We, as professionals, have much more respect for the printed word than do most people. To most people, it is more important that you, a knowledgeable, personable expert they can see, says “A is true” than that some unknown other person says “B is true.” On the topic of out-of-context quotes, if the lawyer asks you, “Now, did you not say X in your direct testimony?” and you are reasonably sure you did not, you can say, “I do not believe that I said that. Could we have the reporter read it back?” That is the reason, after all, that we use court reporters instead of depending solely on tape recorders. Do not answer a question unless you are sure you understand it. This means that if the lawyer misuses a technical term, ask him or her to rephrase the question; do not simply assume that you know what he or she is driving at. Also, be sure to confine yourself to your specialty. I am sure you know a great deal about adjacent subject areas, but you are not an expert in them. If you get led into them, you might make a mistake, so stop the process before it gets started. For example, if an environmental chemist testifies that substance X was present in the groundwater at 0.1 ppm, upon being asked, “Is 0.1 ppm dangerous?” he or she should reply, “Well, I think you need to ask a toxicologist that question.” “I don’t know” is an acceptable answer, but be sure to explain why, for example, “I did not think that important to my opinion in this case.” Be certain to add, “But I can find out if you wish.” It is also acceptable to say “I know I know that; it is on the tip of my tongue. Could we come back to it?” All of us have that sort of experience occasionally, and it does not hurt your professional expertise to admit it when it happens. If asked where a number came from, if it came from a recognized source, e.g., the Handbook of Physics and Chemistry or the Merck Manual you are entitled to say so. Furthermore, your attorney can, if he or she wishes, move to have the source admitted into evidence under either of two evidentiary rules, one related to “published compilations” (Federal Rules of Evidence 803(17)) and the other to “judicial notice” (Federal Rules of Evidence 201). Always watch for what are called “argumentative questions.” These are exemplified by the classic “Have you stopped beating your spouse?” Point out that the question cannot be answered in the form presented, and explain why. You should regard cross-examination as an opportunity to repeat as much of your direct testimony as you can. When answering, you can say, “Well, as I said earlier,…” A good lawyer will not give you the chance to do this by not asking such questions, but if presented with the chance, grab it and run with it. Again, remember that it is the jury or the judge you are trying to convince. Make eye contact with them when answering, not with the lawyer asking the questions.


41

Perhaps the two most important things to remember when being cross-examined are “do not get mad” and “be sure brain is engaged before putting mouth into gear.” The first means you should not get into a shouting match with the lawyer, even if he or she resorts to personal attacks. The second means you should always stop and think for a second before starting to answer the question. 2.9 Summary The techniques and tips presented in this chapter are not exclusive. They are based on my years of trial practice and as a consultant helping to prepare experts to testify. I am sure that the attorney with whom you are working will have other, and possibly contradictory, advice for you. Pay attention to what the attorney says, but always remember that you are the expert and must present the material on the witness stand, not the attorney. In any case, what was presented here should serve as a helpful starting point for you. References Daubert v.Merrill Dow Pharmaceuticals , 509 U.S. 579 (1993). Rogers v.Raymark Industries , 922 F. 2d 1426 (9th Cir. 1991).

Further Information For readers in the U.S., two items can be recommended for further reading: Bronstein, D.A., 1999, Law for the Expert Witness , 2nd ed., Boca Raton: Lewis Publishers; discusses the legal issues regarding being an expert witness. Kumho Tire Co. Ltd. v.Carmichael , 526 U.S. 137 (1999) is a recent Supreme Court opinion regarding the admissibility of expert testimony.

3 Presenting Behavioral Science Data as Legal Evidence: Legal Standards that the Ergonomic and Human Factors Expert Needs to Know Allen K.Hess 0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

Auburn University at Montgomery For most of the last millennium, educated people entered four professions: the priesthood, engineering, medicine, and the law. During the past century, a profusion of professions abounded. The mission of universities shifted from educating a well-rounded gentleman capped by awarding the baccalaureate degree to offering a staggering menu of advanced (graduate and professional) degrees. The explosion of knowledge, the fruit of the Enlightenment, led to differentiation and specialization of knowledge. Consider how few physicians are general practitioners; even the “GP” is now a specialist in “family practice.” This proliferation of knowledge and technical specialties has implications for the justice system. A jury of one’s peers was once composed of educated gentlemen who could understand evidence presented at trials. In fact, most cases were (and still are) heard by judges. As knowledge and technical developments blossomed, the evidence presented in courts grew beyond the grasp of most fact finders (judges and jurors). The duty of a lawyer is to provide a complete and spirited advocacy of his or her client. If technical information may be probative or informing and helpful to a case, an ethical lawyer is bound to seek the best and most favorable testimony for his or her client. At the beginning of the 20th century, Terman introduced an early version of the test he was to adapt from France, which would be known as the Stanford-Binet, into court in defense of a murder suspect. In Europe, the Sterns (current specialization would classify them as developmental psychologists) experimented with the ability of children to provide veridical testimony. This introduction of the “softer” or social and behavioral sciences to the Court reached its apotheosis with two noteworthy figures. Munsterberg issued his clarion call for psychology to enter court with his book On the Witness Stand (1908) and his demonstration of the inaccuracy of eyewitness testimony. He conducted the iconic experiment of having a mock shooting in class and then asking students for their account of what happened. His call was answered by Wigmore’s (1909) satire, which pilloried psychology and Munsterberg as not ready for the Court by a long


44

shot. Munsterberg’s student, Marston, received his Ph.D. and law degree and applied his interests to a number of forensic psychology concerns. A frequent consultant to the police, he determined …that written evidence was superior to oral evidence; that free narration, while less complete, was more accurate than cross examination or direct questioning (see recent cognitive psychology literature concerning recall vs. recognition—how often do we simply put old wine in new bottles!) 1 ; that a witness’s caution in answering was a good indicator of accuracy; and that female jurors considered evidence more carefully than male jurors (Bartol, 1999, p. 8). Marston created the successful comic strip “Wonder Woman,” under the pen name of Charles Moulton. Foreshadowing his scientific and forensic thinking, he had Wonder Woman wear a bracelet that could detect falsehoods in people with whom she interacted. Marston conducted research that found significant elevation of systolic blood pressure when someone lied, thus creating the modern polygraph. The polygraph plays a key role in the development of the concept of the expert witness because of its role in the Frye case. 3.1 Frye Mustering all the evidence possible in the client’s defense, a defendant’s attorney presented evidence through an expert witness that the client was not deceptive. The courts held that …just when a scientific principle of discovery crosses the line between the experimental and demonstrable stages is difficult to define. Somewhere in this twilight zone the evidential force of the principle must be recognized, and while the courts will go long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs (Blau, 1998, p. 422). Frye shows how evidence, in this case the relationship between biophysical measures and emotions, became more technical and required experts to present and explain its features and significance. More specifically, the Frye test or standard of “general acceptance” became the criterion by which scientific evidence became admissible or inadmissible for most of the rest of the century. Many jurisdictions still use the Frye test to determine whether expert testimony is admissible as legal evidence. 1

Author’s insert in parentheses.

Presenting behavioral science data as legal evidence

45

3.2 Evidence Perhaps it is a fortuitous time to introduce a more formal definition of the term we have used in a general sense. “Evidence” is used in several senses. It means “to prove an unknown or disputed fact”; it is “oral and documentary evidence of reasonable and substantial character to prove a point, not to raise a mere suspicion”; and it is “that which demonstrates or makes clear or ascertains the truth of the fact or point in issue, on either one side or the other” (Klotter, 2000). As specialization of knowledge occurred, what constituted a demonstrable truth, fact, or point in issue grew beyond the ken of the typical juror and jurist. The latter is a point that we will revisit when we consider Joiner (General Electric v.Joiner, 1997). 3.3 Jenkins For many years, psychologists were considered somewhat less than their medical brethren because they were not “real doctors.” Lawyers claimed psychologists did not cut patients and prescribe medications and thus were less than physicians and not really entitled to the appellation “doctor.” This challenge to psychologists is still heard in courts. Jenkins v. United States (1962) held that The determination of a psychologist’s competence to render an expert opinion based on his findings as to presence or absence of mental disease or defect must depend upon the nature and extent of his knowledge. It does not depend upon his claim to the title “psychologist.” And that determination, after hearings, must be left in each case to traditional discretion of trial court subject to appellate review (Blau, 1998, p. 346). The Jenkins court further defined the expert witness as qualified to testify because he has firsthand knowledge of the situation or transaction at issue that the jury does not have. The expert has something different to contribute: the power to draw inferences from the facts that the jury would not be competent to draw. To warrant the use of expert testimony, then, two elements are required. First, the subject of the inference must be so distinctively beyond the ken of the average layman. Second, the witness must have such skill, knowledge, or experience in that field or calling as to make it appear that his opinion of [sic] inference will probably aid the trier of fact in his search for truth. The knowledge may in some fields be derived from reading alone and in some from practice alone, or as is more commonly the case, from both (Beis, 1984, p. 234). The key point in Jenkins for all psychologists, not just clinical psychologists, is that challenges for ergonomics and human factors experts as not being engineers or some other “hard” science experts is hardly a disqualifying objection. Although experts concerning pain, fatigue, disability, and other subjectively experienced phenomena will be subject to just as scathing an attack on their presence in court as clinical psychologists


46

have been, their presence in court is just and justifiable. Jenkins, then, becomes a compelling line of reasoning for the expert and the retaining attorney. However, time does not stand still and, following Jenkins, the Supreme Court had a commission study the rules of evidence. In 1975, Congress approved the Federal Rules of Evidence (Green and Nesson, 1984) under which federal jurisdictions now function. 3.4 Federal Rules of Evidence (FRE) Although the whole of the rules is fascinating, the pertinent part concerns Article VII. These rules are promulgated for the federal bench; however, such rules form the template for state courts. Many state courts have adapted them or a variation of the rules. Rule 701 If the witness is not testifying as an expert, his testimony in [the] form of opinions or inferences is limited to those opinions or inferences which are (a) rationally based on the perception of the witness and (b) helpful to a clear understanding of his testimony or the determination of a fact in issue. The sum and substance of this rule introduces several predicates. First, there will be a distinction between the lay and expert witness. Then, any lay opinions must be tied to direct experience and helpful to the trier of fact (judge or jury). Rule 702 If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or determination of a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education may testify thereto in the form of an opinion or otherwise.

The upshot of this key rule is (1) to admit specialized knowledge or expertise, (2) to establish how an expert attains that status (knowledge, skill, education, experience, or training), and (3) to allow the expert to proffer an opinion, as opposed to other experts who are limited in testifying to evidence based on their sense experience.


47

Rule 703 The facts or dat[a] in the particular case upon which an expert bases an opinion or inference may be those perceived by or made know[n] to him at or before the hearing. If of a type reasonably relied upon by experts in the particular field in forming opinions or inference upon the subject, the facts or data need not be admissible in evidence.

The import of Rule 703 is to allow three sources of information from the expert: (1) that which is traditionally based on sense experience, (2) that of allowing the expert to answer hypothetical questions by way of illuminating the current case, and (3) that having thirdparty evidence admitted through the expert. The latter might include scientific findings by others, evidence from research conducted by members of the expert’s research team, and such third-party data as public opinion polls. For example, the infamous case of preschoolers who recognized Joe Camel more than Mickey Mouse relied on a pediatrician’s questioning of several hundred 3- to 6-year-old children. This rule allows such data entry to the Court without its being barred per se by hearsay rulings. Hearsay rulings generally do not admit testimony from a third party that is not subject to crossexamination. Rule 704 Testimony in the form of an opinion or inference otherwise admissible is not objectionable because it embraces an ultimate issue to be decided by the tried fact.

The import of the deceptively simply stated Rule 704 is to allow the expert witness to use the word “did” rather than the words “might” or “could.” For example, testifying as to the ultimate issue or whether the individual did or did not have criminal responsibility or intent was no longer “usurping the province of the jury.” The expert could now make stronger assertions in such matters as intoxication, speed of vehicles, handwriting similarity, and other provinces of the expert. Rule 705 The expert may testify in terms of opinion or inference and give his reasons therefore without prior disclosure of the underlying facts or data, unless the court requires


48

otherwise. The expert may in any event be required to disclose the underlying facts or data on cross-examination. This rule lifts the requirement leading to the “song-and-dance” routine of introducing predicates of a hypothetical question. However, the burden now shifts to the opposing counsel as to questioning the basis for the expert’s opinions. The judge, too, can ask for preliminary disclosure for the opinion’s basis. Rule 706 (a) Appointment. The court may on its own motion or on the motion of any party enter an order to show cause why expert witnesses should not be appointed, and may request the parties to submit nominations. The court may appoint any expert witnesses agreed upon by the parties, and may appoint expert witnesses of its own selection. An expert witness shall not be appointed by the court unless he consents to act. A witness so appointed shall be informed of his duties by the court in writing, a copy of which shall be filed with the clerk, or at a conference in which the parties shall have opportunity to participate. A witness so appointed shall advise the parties of his findings, if any; his deposition may be taken by any party; and he may be called to testify by the court or any party. He shall be subject to cross-examination by each party, including a party calling him as a witness. (b) Compensation. Expert witnesses so appointed are entitled to reasonable compensation in whatever sum the court may allow. The compensation thus fixed is payable from funds which may be provided by law in criminal cases and civil actions and proceeding involving just compensation under the Fifth Amendment. In other civil actions and proceedings the compensation shall be paid by the parties in such proportion and at such time as the court directs. And thereafter charged in like manner as other costs. (c) Disclosure of Appointment. In the exercise of its discretion, the court may authorize disclosure to the jury of the fact that the court appointed the expert witness. (d) Parties’ Experts of Own Selection. Nothing in this rule limits the parties in calling expert witnesses of their own selection.


49

This lengthy rule tempers the shopping for expert or “hired gun” practice by specifying an already existing option for the Court—that of appointing its own expert. This is designed to make the counsels aware that their experts might be impeached by the Court’s expert. Although federal courts since 1946 (Rule 46 of the Federal Rules of Criminal Procedure) allowed for such experts, Rule 706 broadens this access to civil cases. These rules have percolated through the judicial system for the last three decades. However, events in the 1980s and 1990s led to several important findings that have changed the face of articulating scientific, technical, and expert evidence to courtroom testimony. 3.5 The Daubert-Joiner-Kumho (DJK) Trilogy In the two decades following the advent of the Federal Rules of Evidence, there arose a flood of experts into the courts and defenses that stretched credulity (e.g., Mike White’s “Twinkie” defense, the Zamora “TV made me do it” defense, and the “lawn-fertilizermade-me-crazy” defense). This flood swept the term “junk science” into the public lexicon as a term of scorn. Books such as Galileo’s Revenge (Huber, 1993) and Whores of the Court (Hagan, 1997) dissected the horrors visited by some “experts” upon the courts. This public tumult led to three cases decided by the Supreme Court that now define the expert and expert testimony. Daubert Benedictin prescribed for mothers suffering morning sickness resulted in children born with limb reduction birth defects. In Daubert v.Merrell (1993), the U.S. Supreme Court established a four-pronged test (or set of definitions) to guide judges in screening experts and their testimony. Daubert replaced Frye with the FRE; it emphasized FRE Rule 702 by having the judge ensure that the testimony has a reliable foundation and is relevant to the task at hand and offered the four-pronged test. The first prong is based on the philosophy of the Vienna Circle, which holds that (1) a theory must be falsifiable (or is the theory so amorphous as to be untestable, incapable of being disproved, and philosophically and scientifically useless?). Second, Daubert holds a theory or technique in better regard if it has been (2) subjected to peer review and publication. Although not a sine qua non, publication in peer-reviewed journals increases the likelihood that substantive flaws in methodology will be detected. To be sure, we seem to have a scientific fraud eruption at least twice a decade in medicine and physics. Highly regarded breakthrough papers that top journals compete to publish first (establishing priority) are sometimes retracted in small print. Although published peerreviewed papers are an excellent filter, neither we nor the courts should regard them as fool- or tamper-proof. No measure exists without variation or “error” This error is often translated in psychological, physical, and pharmaceutical measurement as one of the family of standard error terms (e.g., standard deviation, standard error of measurement, standard error of the means) or in “hits and misses” in diagnostics. Thus, the third prong calls for


50

consideration of (3) the known or potential rate of error. Finally, (4) the Frye test of general acceptance has a place. It defines the relevant scientific community and allows inference based on whether widespread or only minimal support within the scientific community lends to acceptance by the Court or leads the Court to view the evidence with skepticism. More detailed examination of Daubert, subject to many articles in the legal literature, is beyond our scope. Suffice it to say that Daubert defined the contours of experts and their evidence but did leave a few questions unanswered. To be sure, Daubert is not a one-sided affair. It does not exclude wholesale and, in an uncompromising fashion, any evidence that is not generally accepted (Frye). Because the pace of scientific progress seems to have accelerated, some valid and probative findings may take time to become generally accepted, but might have served justice if admitted before being wholly or generally accepted canon. Daubert says, then, that expert evidence is subject to cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof. However, who says what can enter the Court and how is that to be decided? Joiner Did Daubert liberalize the rules rather than restrict junk science entering the Court? Are we in for a “free-for-all?”’ General Electric v. Joiner concerned whether Robert Joiner, working as an electrician in and around Thomasville, Georgia’s electric transformers, was exposed to polychlorinated biphenyls (PCBs), which led to his contracting small cell lung cancer. The testimony—about quite technical matters as to whether PCBs by themselves or with the aid of furans and dioxins caused the cancer—was central to deciding the case. In order to grasp the essential issues in a case brought before it, a jury of one’s peers needs experts. Who decides what is admitted in court? Joiner holds the judge, as the gatekeeper, responsible for deciding what is admissible to the Court’s consideration, subject, as usual, to the “abuse of discretion” standard. That is, the judge in Joiner decided animal research should be excluded because the evidence “did not rise above ‘subjective belief or unsupported speculation’” (Hess, 1999, p. 525). The Court held that the judge must ignore logic, settled law, relevance, or reliability issues to have abused his or her discretion. Going from Frye to Daubert did nothing to diminish the judge’s gatekeeper role. In fact, one real problem remaining is that judges, schooled in the law, may not have a clue about force vectors in an automobile seat belt case, about psychosis in an insanity defense, or about potentiation of PCBs by furons in the Joiner case. Yet, the status is that the judge determines whether the witness and the testimony proffered are more probative than prejudicial and that is subject to cross-examination, contradictory evidence (and witnesses), and instructions as to the burden of proof in the case at hand. The judge really must “lose it” in terms of logic and the settled law and basic rules of reliability of evidence to abuse his or her discretion. However, one more question confronts us. Kumho What happens when a fire breaks out in a restaurant and the owners claim the hundreds of thousands of dollars of wine stored in the wine cellar are now tainted and have lost all


51

their value? How does one value a one-of-a-kind gem? When science meets application, the relevant expert might be more of an artist than a scientist. Do Daubert and Joiner speak to the nonscience expert? Patrick Carmichael’s tire blew out and his vehicle rolled over, killing one person and injuring others. Tire failure expert Dennis Carlson, Jr., intended to testify that a defect in the tire’s manufacture or design caused the blowout. However, although accepting Carlson as an expert, the Court found so many problems with the consistency of his testimony and the bases or methodologies he employed that it dismissed his testimony as unreliable. The essence of Kumho for us is that a lower court held Daubert’s factors did not apply to Carlson because his evidence was based on skill or experience. However, the Court held that technical and expert testimony was subject to Daubert, which was not limited to scientists but applicable to all experts ordered by the courts or offered to the courts. Thus, although perfume experts who rely on physical measures would be welcome to the Court, so would experts who could testify as to the noxiousness of certain aromas, subject to the DJK trilogy. 3.6 Presenting Behavioral Science Evidence It took time to lay the groundwork in reviewing the legal basis for the expert witness and expert testimony. Now how do we articulate and judge psychological testimony relative to the DJK (and Frye, which is still used in many states)? Actually, psychological data stand in good stead. Qualifying the Expert Witness First, a potential witness must qualify through a process called voir dire. The potential expert should bring a complete curriculum vitae (I always bring copies for the judge, court clerk, and opposing counsel as well as one for the attorney retaining me and for myself), which he or she has reviewed and updated recently. The attorneys will then have the expert recite qualifications based on knowledge, skills, education, experience, and training. If the qualifications seem to be probative or will lend to some factual understanding of the case, the judge qualifies the expert or accepts him or her. The judge can qualify the expert for some areas of inquiry but not others. For example, I was asked to testify as to whether a defendant could understand his Miranda warning against selfincrimination despite his consuming between a pint to a quart of bourbon while fishing. I was disqualified as an expert regarding the effects of alcohol on his cognitive capacities but qualified regarding my estimates of his intelligence vis-à-vis understanding a text such as Miranda, irrespective of alcohol intake. The Nature of Psychological Theory and Data Behavioral science has probably the most voluminous literature as well as journals with high rejection rates. Thus, a tremendous font of knowledge is available and the standards for its peer-reviewed published literature are high. Before presentation to the Court, in order to be admissible, the data upon which our opinions rest should pass the Daubert


52

screens: generally accepted, peer-reviewed, testable, and having a known error rate. The adversarial attorney should insist on no less; the conscientious and ethical witness, too, should offer to his or her retaining attorney no less than “Daubertized” data or the ugly spector of malpractice will arise. Do one’s data and opinions need to pass each criterion? No, the courts use a criterion sometimes termed the “totality” of the evidence in reaching a conclusion, in this case as to admissibility of evidence. That is, unpublished data might have a well-established error rate and be so compelling and probative that the Court will look at the totality of the evidence and admit them. Conversely, well-published evidence, meeting all the other three Daubert criteria as well, simply may not be relevant and might be seen by the judge as more prejudicial than probative and thus not be admitted. Is the key issue the data or the witness? Validity What the courts know as “reliability,” social and behavioral scientists term “validity.” The Standards for Education and Psychological Testing (American Educational Research Association, American Psychological Association, and National Council on Measurement in Education, 1999) defines validity as: …the degree to which evidence and theory support the interpretation of test scores entailing by proposed uses of the test. Validity is, therefore, the most fundamental consideration in developing and evaluating tests. The process of validation involves accumulating evidence to provide a sound scientific basis for the proposed score interpretations. It is the interpretations of test scores required by proposed uses that are evaluated, not the test itself. When test scores are used or interpreted in more than one way, each intended interpretation must be validated (Standards for Educational and Psychological Testing, 2000, p. 9). Thus, the crux of expert testimony is in the soundness of the interpretations rendered by the expert. This is necessarily so. Consider the Rorschach test; more heat than light seems to have been showered upon this controversial assessment method. Although some have condemned the Rorschach as the current incarnation of phrenology or Tarot cards, others stand by the technique with equal resoluteness. In fact, I have seen some licensed psychologists who should never be allowed near a set of inkblots. However, I have also seen a few masters of the technique in whose hands this instrument is of superb assistance in clinical and forensic arenas. How does one judge the admissibility of such a method? Simply put, has the expert mastered the literature; had experience, training, and education in its application; and developed skills that would endow the proffered testimony with validity that helps in resolving the question at hand? In the case of the Rorschach, Hess and colleagues (2001) have found some indices on the Exner system of the Rorschach to be as refined and confirmed as any of psychology’s objective inventories tests, while other indices are unproven as yet or have no claims to validity. The Rorschach, as with most measures, has a more nuanced literature than partisans on


53

either side of the debate care to acknowledge. However, the critical question for us rests not solely or even mostly with the measure, but, as in the Joiner case, with the way in which the expert has mastered the subject matter and applies it to the question at hand. 3.7 Conclusion Unlike almost a century ago when psychology was still young and the concept of the expert witness was yet to be distilled by the courts, the social and behavioral sciences are not merely ready for the courts but also indispensable in helping the courts decide questions. The remaining questions for the potential expert are whether he or she is qualified in an area and whether the expert and attorney are able to articulate the legal questions with the psychological answers. We should be careful in answering these questions so that we are not purveyors of “junk science,” but rather can bring data and evidence to court that serve justice and better society. Acknowledgment I appreciate the contributions of Kathryn D.Hess in the development of this chapter. References American Educational Research Association, American Psychological Association, and National Council on Measurement in Education. (1999). Standards for Education and Psychological Testing . Washington, D.C.: Author (APA). Bartol, C. (1999). History of forensic psychology. In A.K.Hess and T.E.Weiner, Eds., The Handbook of Forensic Psychology , 2nd ed. New York: John Wiley & Sons, 3–22. Beis, E.B. (1984). Mental Health and the Law . Rockville, MD: Aspen. Blau, T.H. (1998). The Psychologist as Expert Witness , 2nd ed. New York: John Wiley & Sons. Daubert v. Merrell Dow Pharmaceuticals , 509 U.S. 579 (1993). Frye v. United States, 293 E 1013 (D. C. Cir. 1923). General Electric Co. v. Joiner , 522 U.S. 136 (1997). Green, E.D. and Nesson, C.R. (1984). Federal Rules of Evidence: with Selected Legislative History and New Cases and Problems . Boston: Little, Brown. Hagan, M.A. (1997). Whores of the Court: the Fraud of Psychiatric Testimony and the Rape of American Justice . New York: HarperCollins. Hess, A.K. (1999). Serving as an expert witness. In A.K.Hess and T.E.Weiner, Eds., The Handbook of Forensic Psychology , 3rd ed. New York: John Wiley & Sons, 521–555. Hess, A.K., Zachar, P., and Cramer, J. (2001). Rorschach [review of the Rorschach inkblot method]. In B.S.Plake and J.C.Impara, Eds., The Fourteenth Mental Measurements Yearbook . Lincoln, NE: Buros Institute, 1033–1038. Huber, P.W. (1993). Galileo’s Revenge: Junk Science in the Courtroom . New York: Basic Books. Jenkins v. United States , 307 F. 2d 637 (D. C. App. 1962). Klotter, J.C. (2000). Criminal Evidence , 7th ed. Cincinnati, OH: Anderson. Kumho Tire Co. v. Carmichael , 526 U.S. 137 (1999).


54

Munsterberg, H. (1908). On the Witness Stand: Essays on Psychology and Crime . New York: McClure. Wigmore, J.H. (1909). Professor Muensterberg and the psychology of testimony: being a report of the case of Cokeston v. Munsterberg . Illinois Law Rev. , 3, 399–445.

4 Practical Ethics for the Expert Witness in Ergonomics and Human Factors Forensic Cases Allen K.Hess 0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

Auburn University at Montgomery Many texts concerning the intersection of law and applied disciplines address ethical concerns in passing or do not address them at all. Ethical concerns are like our skin. Skin is our largest organ and the one that interacts with the world around us. However, it is simply overlooked until something goes wrong—and so it is with ethics. If ethical concerns are unaddressed, they will surely emerge during forensic consultation, with potentially disastrous results. Thus, I am happy the editors of this volume see fit to feature a chapter on ethics and the expert witness in forensic ergonomics and human factors. The goal of this chapter is to alert the reader to some concerns which, if addressed appropriately, lead to the immense satisfaction of doing a job professionally and serving society. We will address the ethical questions that face the expert at the beginning of a case, followed by some questions that might arise during the datagathering and consultation stage, and conclude with questions that arise during the concluding stage. 4.1 Accepting the Forensic Referral There are more pitfalls at this stage than at any other. The expert needs to consider the person (usually an attorney) who asks for his or her help. The expert needs to determine whether (1) any parties in the case might contribute to a conflict of interest or the appearance of such a conflict and (2) this attorney understands the articulation of the legal and psychological issues. Conflict of Interest A conflict of interest is when a public duty is compromised or appears to be compromised by a private or a pecuniary benefit. Thus, ethical practice dictates that the

Practical ethics for the expert witness in ergonomics and human factors forensic cases

57

practitioner should ask the attorney to identify the parties in the case: individuals such as attorneys on both sides of the case, the plaintiff and defendant, and the judge hearing the case. If the professional knows any of the parties or has a financial interest in a plaintiff or defendant company or agency, he or she should disclose this to the attorney. In practical terms, within many communities, roles such as psychologist, coach, adjunct instructor at a local college, and scout leader or parent of a scout can intersect. Suppose an opposing attorney is the coach of a gymnastics team on which the psychologist’s child participates. Will this possible role conflict jeopardize the child’s chances of competing in tournaments as the first uneven parallel bar contestant? If this is a concern to the potential expert witness so that he might modify forthcoming testimony to the detriment of the client attorney, a conflict of interest exists and the professional might be advised to recuse himself. With other people this might not be a factor at all and no conflict of interest would be present. The key is to be aware of the possible conflicts of interest that could arise, to bring this issue up with the attorney during the first phone call, and to be clear about one’s own vulnerabilities to influence. Interestingly, these factors are not engraved in stone but involve judgments about a shifting set of variables. For example, with a large litigation such as a class-action suit, the boundaries might be different than with a more local case. With the larger case, one might be practicing in another community and not know any of the parties. On the other hand, with a class-action suit, one might be personally affected or be more likely by chance to know someone in the suit. Another variable is the size of the community in which the suit and the expert are situated. In a smaller community, one is more connected with other parties; if the judge or attorneys were to start excluding anyone who might know anyone else, there is little chance the case would be heard. In such a case, the expert might be asked whether he can work with disregard to the interests of the people in the case and, instead, provide unbiased and professional testimony. In smaller and more isolated communities, a professional might be asked to practice on the fringes of his competence because other, more knowledgeable experts might not be available. Then the question becomes one of whether, by reading and consultation, the professional can become expert enough to serve the case competently. We will return to considerations of competence shortly, but now consider personal and political philosophies. These examinations of interests and competencies extend to personal and political philosophies as well. Consider a person who is a champion of the “little people” and has written and spoken out on issues regarding automobile safety. Could this person serve as an expert witness in a personal injury case in which the issue is whether the large automobile companies provided enough safety belt protection or cut costs to use a cheaper and less safe material in the belts? Will the astute opposing attorney bring up past statements to impeach the integrity of the expert’s testimony? A person can hardly live a textured life without having interests, so a frank discussion with the attorney can help anticipate a possible ethical problem or an attack by opposing counsel as to the motives of the expert witness in an attempt to impeach the testimony as serving a personal, financial, or political interest. On a tactical level, the hiring attorney should be impressed by the professional’s sensitivity to these issues that could hurt the case.


58

Attorney’s Qualities If the attorney does not understand the psychological issues or has no interest in being educated about them, several unfortunate consequences will result. During the case, the expert will not be regarded well, so access to people and records will have a lower priority for the lawyer. Then the expert’s work product will suffer because of gaps in the information that he has gathered; the expert can be vulnerable to cross-examination that he has not prepared his testimony in a professional manner. Attorneys who do not value the expert will be resistant to paying the witness. Also, the attorney will not be able or willing to rehearse the types of questions that will elicit the best testimony from the expert during the trial. Naturally, attorneys are interested in winning their cases—their professional survival depends on winning. It would be inconsistent with their interests to secure a witness who will not help the case or even proffer testimony that hurts the case. (Note that attorneys might retain an expert who is a highly regarded proponent of the issues or theories that the other side of the case is pursuing. The retention of the expert ensures that the expert will not be testifying on behalf of the opposing side because a person hired by one side cannot be retained by the other.) However, the attorney’s zeal about winning can lead to the lawyer wanting an expert who is simply a “hired gun,” or one who will testify in whatever fashion the lawyer wants, irrespective of the facts of the case. For example, one psychologist I know, who has state contracts with the prison system and is on retainer with the district attorney’s office, has never found a defendant insane, no matter how much debilitation the person presents. How does the ethical expert find out about the lawyer’s orientation? During the initial telephone call, the expert needs to have the attorney articulate the theory that he or she is using in the case and how the psychologist might provide information that would support or at least test the theory. For example, the psychologist is asked whether he might be interested in a tort case concerning contracts that people have signed, not realizing the usurious interest rates being charged and that the consumers have been misled by advertisements concerning compressed speech on radio and television advertisements that virtually no one could have been reasonably expected to understand. At that point, the psychologist might mention one or two theories concerning the size of the sensory register in understanding information. If the attorney appears interested and can comprehend why the psychologist might be thinking about sensory register and shortterm information storage, then the lawyer is articulating the legal and psychological constructs. Also, this helps the attorney feel confidence in the psychologist. Because virtually all the activities of any profession are verbal, the expert witness and the lawyer have ample opportunities during the phone call to assess each other’s verbal facility. The initial call should not be jargon filled; an excellent rule of thumb is that the best professionals can explain $25 concepts in 254 terms. The initial call allows the attorney and the potential expert witness to see how each conceptualizes and whether they can work together. Most ethical problems arise when people are angry or disappointed with each other and communication and trust have not been established or have broken down.


59

The professional needs to be careful in judging the attorney’s skills and ethicality. The success of our practice depends largely on our talents and motivation. When practicing jointly, our efforts may be enhanced or diminished by others. However, in forensic work, a great deal of the effectiveness of the professional will depend on the retaining attorney’s efforts, talents, and ethics. How the forensic psychologist’s evidence and case are presented is almost wholly a function of the attorney. Consequently, the lawyerexpert interaction is vital in determining whether forensic practice is satisfying and successful, or not. We will decline to work with some attorneys. The crucial reason is that they have demonstrated an inability to use the psychologist’s efforts in a way that benefits the client. In one case, the attorney seemed unaware that the expert must be disclosed to the opposing counsel a minimum of 10 days before trial. This allows the opposing counsel the opportunity to depose the witness and discover the nature of his testimony, a firstyear law school topic. In this eye-witness accuracy case, the judge accepted the psychologist as an expert but did not let him testify because the opposing counsel had no opportunity to impeach the expert’s testimony. The defendant was eventually sentenced to 20 years for a simple assault when the eye-witness accounts were clearly in error (in the psychologist’s opinion); the witnesses described a 5-ft 11 in. narrow-faced, righthanded assailant, but the accused was 5 ft 4 in., round faced, and left handed (facts that the police and lawyers did not present during the case but were to be presented by the expert witness). This was the critical evidence in the case; no other evidence— fingerprints, DNA, other witnesses, stolen property, guns, wigs used by the perpetrator, or any other evidence—was presented. Although the psychologist was paid and stayed on to help the attorney structure the closing argument, the case left a bitter taste. Imagine how unsatisfying it is to be a party to sending a most likely innocent 24-year-old person to a career as a felon because the attorney did not do what a first-year law student learns to do—present a witness list to the opposing attorney. Who, besides the attorney, might try to retain the expert? From time to time, people other than attorneys might call the potential expert witness. Too often, the call might lead to undue complications. For example, a patient with temporal mandibular joint disorder asked her psychotherapist (treating her for suppressed anger, which led to jaw clenching) to testify as an expert witness. He asked her to consult with her attorney about contacting him. The lawyer specializing in TMJ disorder litigation might have a stable of experts he or she uses regularly and might not be able to use the therapist within the regular preparation of the case, might prefer not to use the therapist as a potentially biased professional, or might prefer to use the expert to establish damages during that phase of the trial. Thus, the expert is almost always best served by having the attorney as the client. Also, this makes billing and privilege more easily handled. With the attorney as client, the expert and attorney can fully disclose to each other. With a third person retaining the expert, the expert’s first allegiance would be owed to the third party and that person might reveal information to the expert, not wanting it to go to the attorney, which then places the expert in a dilemma about withholding information that the attorney should have in the best interests of the case. Billing is generally easier if the attorney has signed some form of contract with the expert that the attorney is to pay the expert. That way, the expert’s time later in the case


60

and even after the case is not consumed by chasing the third party. Attorneys are more expert at collecting fees than are most experts. Thus, expert time is saved by billing and receiving payments from the attorney. Again, this is best clarified during the first phone call. More money matters will be discussed later. Expertise or Competence Personal Skills Perhaps before answering or returning the attorney’s initial phone call, the potential expert should have assessed himself to see whether he has the requisite skills for courtroom combat. Some people are naturally averse to conflict. The American legal system is adversarial. If the expert is uncomfortable with the professional, and sometimes personal, attacks that the opposing attorneys will mount (they owe their clients a spirited representation, which might include a virtual assault on the expert’s professional life and personal integrity), forensic work may not be his métier. Is the expert able to cope with this kind of attack? Professional Competence Does the professional have the expertise needed for the question at hand? While the psychologist whom the attorney called might be an “expert forensic psychologist,” is he versed in psycholinguistic processing? Does the psychologist have the skill, knowledge, education, experience, and training in how people process compressed speech to the extent that he should be offering professional services? In the American Psychological Association’s Code of Professional Conduct (1992) and in every other profession’s code of ethical conduct, the issue is framed as “competence”: is the expert professionally competent in the area of service offered? This question is seemingly simple; however, the most skilled psycholinguist might not be able to articulate the research on compressed speech with the legal concept of the reasonable person being duly informed by the advertisement. Even more subtly, can the way the advertisement was posed actually lead to the inference that the company offering the loan misled the consumers? This implies the legal construct of “intent.” The psychologist who was contacted might not have mastered the subtleties of the psycholinguistics research base, but that might not be the level of discourse needed in the case. The case might simply need someone who can read and metabolize the scientific literature and then articulate the psychological knowledge about sensory register capacity and short-term memory development to testify that no reasonable or even highly educated person could have registered or stored the disclaimers in the advertisement on loans. If the information needed is more complex than the expert’s skill level, might the expert then be able to retain the services of other experts—in effect, forming a team? Perhaps some research directly related to this case—taping the putatively offending advertisements and showing them to a sample of college students to show that they were unable to process the disclaimers—might be subcontracted to a psycholinguist. To the extent to which they can be anticipated, these issues are best discussed at the inception of the case. Even more so, the attorney may feel comforted if the expert can draw on the


61

skills of a team, to the extent to which the case financially warrants this degree of expertise. Attorneys are used to working in teams in which a partner will assist on certain technical features of a case and are aware of billing issues in which extra expenses need to be justified. Standards of Practice In areas such as child custody, in which psychologists have been routinely employed for decades, standards have been developed concerning how psychologists should function in child custody evaluations. The expert should know the guidelines and standards of performance available before proceeding in a case. Sometimes the standards might be available in a cognate or related discipline. Although they may not be binding for people who are not members of that profession, they are informing. For example, if certain tests are conducted on products in an industry, the expert should be knowledgeable about these tests, even if they are not ones that he is qualified to run. Being informed about them, as well as the results of such testing in the particular case, might inform the expert’s testimony; at the least, he would not seem adrift if queried about such information by opposing counsel. Even more telling, knowledge of these standards will show the expert to be ethical and careful, with an awareness of professional issues and standards. For example, in a case in which the levels of noise in a factory are at issue because of damaged hearing ability in workers, the acoustical expert witness psychologist would need to know how acoustical engineers conduct their tests. In a real sense, such an informed expert is at a great advantage. The expert psychologist, conversant with the engineer’s tests and able to articulate them with how a worker’s hearing is damaged, is much more valuable than the engineer who cannot address the subsequent functional and emotional damages resulting from the physical conditions. The engineer’s tests and standards are not binding on the psychologist, but do provide interesting approaches to the case. Such knowledge helps the psychologist understand how other professionals might approach the case and allows the psychologist to subsume the other side’s theory and evidence. This kind of awareness and professionalism is what is called practicing with aspirational ethics, as opposed to practicing to minimal standards. Standards of practice are held to be across the profession as opposed to practicing to the standards of local practitioners. There are emerging standards for expert witnesses (Committee on Ethical Guidelines for Forensic Psychologists, 1991; Perrin and Sales, 1994) that the expert-witness psychologist should know. Similarly, when the expert witness is testifying about another professional’s conduct or in a product liability case, the material standards are national, not local. In one case, an anesthesiologist’s defense about leaving a sedated patient who subsequently suffered irreversible brain damage was based on the fact that, in that rural area, no one else was available to assist in another surgery at another hospital. In a more urban area, he argued, other anesthesiologists would have been available to cover at the other surgery. Thus, the anesthesiologist left the patient, assuming he would come out of the anesthesia in good shape, but he did not. The verdict held that standards of professional conduct in such cases are national. On the other hand, when there are state laws or more local practices, the contours of the case, of which the attorney should make the expert aware, might involve the “usual,


62

customary, and reasonable” customs of the local jurisdiction. The expert needs to be sensitive to these issues and inquire of the attorney; however, they are more the central concern of the attorney and another way of determining the professionalism of the attorney retaining the expert. Again, the forensic professional’s level of practice, to a greater extent than when practicing within his usual activities, is dependent upon the attorney retaining the practitioner. Legal Knowledge Although the expert is retained for his or her professional knowledge, how much law does the forensic expert need to know? The answer to this question focuses on the core question—one needs to know about the legal construct in order to articulate it with the professional’s knowledge base and to know enough about legal procedure so that the expert is aware, not ignorant, of court. On the other hand, the expert need not be a lawyer; to some extent, even if the expert has a law degree, he or she should not be practicing law, but proffering the professional service that the attorney is seeking. The attorney should be able to teach the professional the law concerning the case, and the expert would be well served to be able to understand how lawyers think about a case. The issue, for example, might not be whether people can understand compressed speech. After all, the law is clear: if the consumer has signed a written contract, that contract usually overrides the advertisements that brought the consumer to the loan agency. The attorney might now suggest that the question is a combination of whether the advertisement, the sales pitch, and the foreboding legal language of the contract reached such proportions that the cognitive psychologist, who is knowledgeable about information processing, can testify that the consumer never had a chance to understand the terms of the loan. The question now is whether the psychologist can understand enough of the theory of tort law to articulate cognitive science with the legal questions to be helpful. The psychologist can read learned law treatises or relevant legal codes in order to understand the legal constructs. The expert is not expected to know the nuances between thresholds in state and federal consumer law (most lawyers do not know such specialized areas). However, the expert should take the opportunity to be educated by the case, to the extent that he has the aptitude to master legal reasoning. Money Matters If the expert has contracted with the attorney, then the attorney is the paying client. Naturally, the attorney is retained by an injured party or by a corporation and is paid by the client. Thus, the expert is ultimately paid by the party or the company, although, in reality, the expert’s client is the attorney and, if the “ultimately paying party” does not pay, the attorney is still liable for the expert’s fees (and expenses), assuming that the wise expert structured the relationship correctly. As mentioned earlier, the expert should not be in the business of collecting fees from recalcitrant payers. The wise expert should ask for a retainer and be sure that billing is current so that the money flow is not backed up. It is good practice to receive payments so that one can devote full attention to the case and not be worried by the fees.


63

In no case should the expert work for contingency fees, as attorneys sometimes do. Contingency fees for an expert witness create an immediate conflict of interest in which the testimony must be skewed, or appear to be so, in order for the expert to be paid. In cases in which the attorney is working for a contingent fee, he or she must understand that the expert is working for a per-hour or per-case fee, not on a contingency basis. The expert should determine the suitable fee based on the type of work, the reputation of the expert, and the norm for the profession. The term of art is “usual, customary, and reasonable” fee. If the expert charges too much, the sensibilities of the lawyers, judge, and jurors might be offended. If the expert charges too little, the question might arise about whether he or she has a financial interest in the case. In a disability case regarding a relatively poor client, fees could be adjusted, although it would be better to charge one’s usual fees so that the court or juror does not see a question of bias due to pity for the client. In such a case, the expert might reasonably excuse part of the bill later, but even this should be done with legal guidance of the attorney so as not to prejudice any appeal or subsequent legal action. Of course, many professions have pro bono practice, whereby one donates services to indigent clients. This must be done with care so as not to have testimony appear biased. 4.2 Data-Gathering and Consultative Stage Once the attorney wishes to retain the expert and the expert agrees to consult with the attorney, a number of questions tinged with ethical implications arise. We will consider several of them. Who Else Has Access to the Expert? The expert and those retained by him or her should communicate only through the retaining attorney to any other parties. Thus, when the expert examines people for disability, for example, the contacts are made with the attorney’s knowledge of the type, time, and extent of the examination. When the expert examines people who helped design or manufacture a product, the examination is conducted through the attorney’s aegis. When the expert receives any inquiries from other parties, particularly the attorney or clients from the other side of the case, the attorney who hired the expert needs to know. Also, no information should flow from the expert or his office to the third parties. It is best to instruct one’s secretary to say that one is out of the office (rather than out of town or, heaven forbid, out of town working on the Smith v. Jones case). It is much better to instruct office personnel not to be afraid of seeming uninformed than to be flattered into revealing information that, once released, cannot be retracted. Legal cases make people desperate, and desperate people behave desperately. In one case, the father of a woman in a divorce case called the expert and offered him vast sums of money not to testify. This was immediately reported to the attorney retaining the expert. In another case, the law intern to the district attorney posed as a student interested in forensic psychology and went to the expert’s faculty office. In his guise as a student, he asked “innocent” questions of the professor, but he was soon discovered when he began to pose questions specific to the case in which the professor had been retained.


64

Again, the expert immediately terminated the interview and called the retaining attorney, who could then choose an appropriate course of action regarding the district attorney’s office. Defining Privacy, Confidentiality, and Privilege What obligations does the expert have in keeping information private? Who else does the expert need to consider in guarding information? Privacy In order for our day-to-day relationships to work, we observe courtesy. When a colleague tells us about his or her feelings about the boss, privacy would dictate that we hold these utterances to ourselves. Our colleague might be damaged by the revelation of his or her feelings. This is privacy. It is a duty we owe to others by virtue of our acquaintance. Violation of privacy is typically not actionable unless defamation with attendant damages results. However, it is a lamentable person who does not hold a friend’s or colleague’s comments, unless such a withholding can damage others. Confidentiality Confidentiality is typically owed to another in a fiduciary relationship. A supervisor owes a worker confidentiality regarding employee evaluations, salary considerations, medical leave information, and a host of other personal information that becomes available in order for the work relationship to function. (Others in the organization do have access to this information by virtue of their status in the organization, e.g., the supervisor’s supervisor, the human resource or personnel office, and the finance office.) One’s stockbroker holds a fiduciary role regarding revelation of one’s financial holdings. However, the expert witness, to the extent that the information is safeguarded by a practice law, might hold a privileged relationship. Privilege Privilege is a standard that calls for a higher level or threshold of urgency in order to breach it. That is, privilege is a protection against court-compelled disclosure of information deemed so important that the relationship cannot function without such protection. Nine relationships are guarded by privilege: • Husband-wife • Priest-penitent • Physician-patient • Psychotherapist-patient • Lawyer-client • Political voting • Trade secrets • State secrets • Informer identities


65

If the professional is covered by a licensing law in the state in which the information is gathered and the case is heard, then communication between the lawyer and expert is privileged. In some cases, the attorney may want to guard against a particularly wellknown professional being retained by the other side or may become aware of the expert’s planned testimony as not being helpful to the case. Then the lawyer, who has retained the expert, may never call the expert and his or her testimony might be regarded as privileged. If the attorney never reveals that the expert was retained, then the information is privileged. Privilege belongs to the client—in this case, the attorney—and cannot be breached by the professional. If the court or other side has learned of the expert’s involvement in the case, perhaps by virtue of the witness disclosure list furnished by the attorney to the court and other side, then the expert may be guided by the court in the event that he or she is ordered to testify. In some cases, the expert may need to retain an attorney for guidance through the thicket of what is privileged. The expert needs to err on the side of caution; therefore, when writing, lecturing, or simply seeking consultation with an attorney or trusted colleague, the expert may want to not disclose or disclose so that the revelation is so abstrusely related to the case as not to be identifiable. Of course, when the attorney offers the expert to the court, privilege has been waived. The expert cannot offer some testimony and withhold other evidence, claiming privilege. If the lawyer does not want potentially damaging evidence offered, then he or she has a choice of not questioning in that area and hoping the other side will not query there, or simply not calling the expert. The expert, when taking the oath, has pledged to testify to the truth, the whole truth, and nothing but the truth. To do otherwise, barring the judge ruling some areas inviolate, could lead to perjured testimony. All the people whom the expert retains in his or her employ are guarded under the umbrella of the expert’s license or confidentiality. The expert must guarantee that the employees understand this and do not reveal any information to third parties. The logic is clear: but for the employment relationship, the janitor, secretary, colleagues, or other people employed by the expert would not have known of the information. Under the “respondent superior” or “captain-of-the-ship” doctrine, the expert has ultimate authority for receiving, retaining, and releasing information. Hurry Up and Wait Attorneys seem to be in a chronic rush. Having retained the expert, the attorney may then seem to have forgotten about him. The expert needs to be sure that he understands the attorney’s calendar regarding the case. When the expert is retained, it is a good idea for him to discuss when the case calendar dates are upcoming. Perhaps the expert was retained so that the attorney could turn in a name on a witness list due the next Monday. Perhaps the case has a half-year postponement. After weeks or months of no contact, the expert may be contacted and asked to provide information within an impossible timetable. It is a good idea to discuss with the attorney any impending vacation plans; other time commitments the expert has made; and the amount of time any investigations, examinations, or studies will take. In that way, the attorney and expert can make a schedule that will not compromise the case or the professional functioning of either. It is


66

in their interests that they perform well. This may involve buttonholing the attorney so that a reasonable time frame can be sketched out. Of course, many variables, such as the judge’s calendar, other attorneys’ schedules, and supervention by other cases, can ruin a schedule. However, these events usually involve delay so that required assessment of a client or conducting of research may be carried out in a timely fashion. Science vs. Junk Science Two related forces led to the increasing use of expert witnesses. Because of the great strides in health and technology over the past several centuries, we live in an age that holds science in high regard. The prestige with which we regard science makes an expert witness particularly attractive for attorneys to use to present evidence to fact finders (jurors and judges). The prestige of science coupled with the increasingly complex technology that face fact finders led to the need for expert witnesses in court. With the advent of the expert witness, it was only a matter time before many charlatans presented themselves as experts. For scorching analyses of “junk science,” Huber’s Galileo’s Revenge: Junk Science in the Courtroom (1993) and Hagen’s Whores of the Court: the Fraud of Psychiatric Testimony and the Rape of American Justice Evidence (1997) are unforgettable primers for the ethical expert witness, matched only by Ziskin and Faust’s Coping with Psychiatric and Psychological Testimony (1994). The latter is a text on disassembling the expert witness and any testimony he or she offers. Any expert will be well served in reading one or all of these before testifying; failure to read these critiques may render the expert unarmed and unprepared for the combat awaiting him in court. Among the key lessons to keep in mind is to be sure that what one is presenting has a basis in fact and that any expression of opinion is accompanied by the appropriate statement of a certainty or probability of certainty or qualifying statements. A number of lessons are distilled in Chapter 3 in this volume, which articulates how the expert answers legal questions with psychological evidence. Recordkeeping Records need to be kept completely. An old correctional psychology dictum regarding correctional officers making their rounds says, “If it is not recorded, you did not make your rounds.” The ethical psychologist keeps complete records and is aware of the work product rule. This safeguards certain notes and records of attorney-expert interactions from discovery; without this, an attorney could not fully function in presenting his or her client’s case. Any questions about the inviolate nature of a particular record ought to be referred to the attorney. Records need to be kept securely. As mentioned earlier, privilege extends to one’s records and test data. If privilege is to be ensured, records must be kept secret and in a place available only to the expert and his direct adjuncts, who should be made aware of privilege. Records should be kept for an amount of time that exceeds any statutory limits that may be imposed on a case. In a case involving termination of a worker who was not fit to return to duty, the psychologist needs to learn from the attorney, and verify in the


67

statutes, the amount of time that the former employee has to file any claims and appeals regarding termination of employment. Also, when disposing of records, the psychologist must be sure that any information no longer needed to be stored is truly destroyed and will not appear on a discarded hard drive of a secretary’s computer. Recordkeeping guidelines have been broached by the American Psychological Association (1993). Getting All the Information or Making a Reasonable Attempt The ethical expert tries to get all needed information. If information from the opposing side is not made available, one’s retaining attorney needs to be notified. If he or she is still unable to obtain it, then records should be kept (e.g., unanswered phone calls and emails, certified mail receipts that show attempts to obtain the data) so that, when appearing in court, the expert can show that all possible attempts to present a complete record were attempted. Then any bad faith dealings by the opposing side may be made evident by the retaining attorney, as he or she sees fit. Conducting Research As part of the development of the case, the attorney and expert need to keep contact as to how much effort the expert needs to expend. In some cases, conducting examinations and research may be unnecessary. If the attorney knows that he or she may settle the case, the expenditures by the expert for research and examinations may be in vain. This calls for continual communication between the two. Pretrial Meetings The attorney and expert should be in communication about evidence that is favorable and unfavorable to the retaining side. The attorney does have the option not to present the expert or not to ask questions that would be adverse to the client. The expert should answer questions asked and is under no obligation to tell all that he ever learned. The expert should be honestly responsive to both attorneys’ queries in deposition and in court. Also, if unfavorable information may come to light in court, the attorney and expert certainly may discuss how to handle it, perhaps on redirect examination. Communication is key; the psychologist and attorney need to “be on the same page” during the case. 4.3 Presenting the Case Testifying All the expert’s efforts may be undone if testimony is presented poorly. Books have been written on presenting expert testimony (and how to tear experts apart), so our comments are simple and limited to the ethics of testifying. Make sure that you have copies of all your communications, in at least duplicate form. Any evidence might be seized by the court, so having a backup copy is indispensable. Be sure all records (including and especially your curriculum vitae) are accurate and that you are familiar with all the data.


68

When presenting evidence, the court-wise expert takes all needed time and teaches the court, without any arrogance. Errors can be made in any number of ways and the expert is sure to make one sooner or later. As soon as the expert realizes that he has made an error, at the next opportunity he should offer the correction in as low-key and sincere a manner as possible. The oath, as mentioned earlier, is to tell the truth, the whole truth, and nothing but the truth. This means that the court-wise expert does not simply volunteer information, but answers the question. If the evidence if unfavorable to the retaining side, it should not be a surprise to the retaining attorney. As mentioned earlier, evidence should be reviewed with the attorney beforehand. Hess (1999) describes more about preparing and presenting testimony, a topic deserving more attention than can be afforded in this chapter devoted to ethics. 4.4 After the Case There is little so final as ending a forensic case. The parties—plaintiff, victim, or defendant, or winner or loser—have been seared and scarred by the legal system. Few will have much time to delight or mourn with the expert. Attorneys who have been adversaries have a drink and perhaps dinner because they might be working together on the next case and both have another case the next morning to prepare and present. If the expert has performed ethically and professionally, he might be thanked, paid, and bid adieu. Records need to be retained; any appeals by either side need to be considered, although appeals are few. The best reward for the expert is the satisfaction in a job done well and perhaps to have served justice and presented psychology to the public as a wellregarded profession. Just perhaps, he will be asked to help again. Acknowledgments I appreciate the contributions of Kathryn D.Hess, Tanya H.Hess, and Steven Walfish to the development of this chapter. References American Psychological Association. (1992). Ethical principles of psychologists and code of conduct Am.Psychologist , 47, 2597–1628. American Psychological Association. (1993). Record keeping guidelines. Am. Psychologist , 48, 984–986. Committee on Ethical Guidelines for Forensic Psychologists. (1991). Speciality guidelines for forensic psychologists. Law Hum. Behav ., 15, 655–665. Hagen, M.A. (1997). Whores of the Court: The Fraud of Psychiatric Testimony and the Rape of American Justice Evidence . New York: HarperCollins. Hess, A.K. (1999). Serving as an expert witness. In A.K.Hess and I.B.Weiner, Eds., The Handbook of Forensic Psychology . 3rd ed. New York: John Wiley & Sons, 521–555.


69

Huber, P.W. (1993). Galileo’s Revenge: Junk Science in the Courtroom . New York: Basic Books. Perrin, G.L and Sales, B.D. (1994). Forensic standards in the American Psychological Association’s new ethics code. Prof. Psychol: Res. Pract. , 25, 376–381. Ziskin, J. and Faust, D. (1994). Coping with Psychiatric and Psychological Testimony . Beverly Hills, CA: Law and Psychology Press.

5 A Road Map for the Practice of Forensic Human Factors and Ergonomics William B.Askren Human Factors Services John M.Howard Crossroads Machine, Inc. 0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

5.1 Introduction This chapter is based on the authors’ many years of experience in the practice of forensic human factors and ergonomics for plaintiff and defense attorneys. The authors have observed that over the past 30 years human factors and ergonomics professionals have been used increasingly as expert witnesses in litigation. Peters (1971) described the early years of product liability and product safety. Initially, the use was by the plaintiff’s bar. Subsequently, the defense bar began employing these professionals to support their clients’ cases and to challenge the opinions of the plaintiffs’ human factors and ergonomics experts. With this increasing use of human factors and ergonomics professionals in litigation, it seems timely to provide a “road map” to guide professionals new to the field in the performance of their forensic practice. A useful background reference regarding product liability and forensic human factors and ergonomics is Weinstein et al. (1978). The types of litigation that serve as the authors’ experience base for this “road map” are personal injury and property damage accidents involving motor vehicles; consumer products; industrial machines and processes; and slips, trips, and falls for plaintiff and defense lawsuits. The human factors and ergonomics principles of the experience base cover a wide variety of topics, such as human use and misuse; human error; human senses, especially visual and auditory perception; biomechanics; anthropometries; decision making; reaction times; skill level; fatigue; training; and corporate safety and organizational policies. Also, physical technologies, such as electrical, mechanical, and chemical parameters that interact with the human factors and ergonomics issues, are part of the experience base.


72

5.2 Objective The objective of the chapter is to present a road map for the practice of forensic human factors and ergonomics from initial contact with the attorney through trial testimony. The intent is to list and discuss the general sequence of activities, grouped by phases, that should be addressed while the professional is serving as a forensic expert witness and consultant. The consultant role to the attorney often is as important as the expert witness role. As a consultant, the human factors and ergonomics specialist may help the attorney develop case strategy as it relates to human involvement in the incident, as well as provide specific services such as developing and reviewing interrogatory questions and developing questions for examining the opposing human factors and ergonomics expert during deposition or trial. 5.3 Principal Issues The principal issues are the five basic phases that make up the road map. The phases are case initiation; understanding the accident scenario; analysis, findings, and opinions; reports; and deposition and trial testimony. These are presented in this section in the order in which they usually are performed by the expert. Included in each phase are its general purpose and intent and a listing of topics and activities to be considered and/or performed during that phase of the road map. Phase I. Case Initiation This is generally the first contact with the attorney. The purpose is to understand the general nature of the case and the time schedule, to decide whether the expert’s credentials and interests are appropriate, to decide whether to work on the case, to establish a financial relationship with the attorney if the expert decides to get involved, and to obtain from the attorney the immediately available case information and data. The following topics and activities should be addressed in the first contact with the attorney: • How did the attorney learn of the specialist’s expertise and availability? This becomes important later, during deposition and/or trial, when he is examined by opposing counsel. Opposing counsel will want to reveal whether his work on the case is the result of aggressive marketing on his part, referral from another attorney, or another case with the same attorney. The opposing counsel will be attempting to set the stage to impugn the expert’s objectivity. • Is this a defense or plaintiff case? Some experts are more comfortable with defense cases and others with the plaintiff. In general, it is suggested that experts try to work on plaintiff and defense cases to indicate their professional objectivity.

A road map for the practice of forensic human factors and ergonomics

73

• Conduct an initial review of the incident with as much detail as possible. Included in this detail might be the accident scenario, the persons involved, description of the equipment involved, and the environment. • Determine whether any conflicts of interest could prevent the expert from being a witness in a particular accident litigation. For example, has he been approached to testify against a defendant whom he has defended before or with whom he has consulted previously? • What role does the attorney want the human factors and ergonomics specialist to serve? Does the attorney have other experts dealing with other areas of the litigation accident? Avoid being an expert on everything. Even within the field of human factors and ergonomics, the specialist will probably have specific areas of expertise. • What is the time schedule of the case? Critical dates are discovery cutoff, deposition, and trial. The expert should consider whether time is sufficient for investigation, analysis, and development of opinions. • Decide whether the expert should serve as the human factors and ergonomics specialist for the case, or possibly recommend another person. Are the human factors and ergonomics issues within his area of expertise? He should be selective about accepting a case. • Request available documentation on the case. Documentation might include accident reports; photographs, witness statements, depositions, interrogatory answers, motions to produce, manuals, etc. • Establish the financial arrangements. It is suggested that the expert have a printed schedule of charges and fees for services, which can be provided to the attorney. A clear understanding of payment of fees is necessary. Such a fee schedule should include: retainer amount (if required), hourly rates for services, what services are included, and fee for deposition and trial testimony if different from the hourly rate. • Provide a curriculum vitae (CV), or resume, to the attorney. Everything written in the CV probably will be examined during deposition and/or trial by opposing counsel. The CV will help establish expertise. • Set up a file with several sections, such as: (1) significant dates, e.g., first contact by counsel, date of accident, potential deposition and trial; (2) documents and material provided by the attorney; (3) documents and material compiled by the expert; (4) analyses and opinions; and (5) date log of the expert’s labor, trips, expenses, etc. During deposition the opposing counsel will thoroughly examine this file and probably will question the expert about every document in it. Phase II. Understand the Accident Scenario The purpose of this phase is to gain as complete an understanding of the nature of the accident as possible. This is essential for follow-on analysis and for development of findings and opinions of the human factors and ergonomics issues of the accident. This understanding will involve learning about any environmental, equipment, human, process, and organizational conditions relevant to the accident. This learning is a continuing, evolving activity over the early life of the case. In fact, the expert may be exposed to several different possible accident scenarios during his analysis. However, the final accident scenario selected must be supported by the analysis.


74

Generally, the more of these topics and activities that can be addressed, the better is the grasp of the accident scenario (see Section 5.6 for a more detailed description): • Read available case documentation, such as accident reports, police reports, insurance reports, witness statements, depositions, interrogatories, requests for admissions, medical reports, and product reports. • Review videotapes and photographs of the accident scene, equipment, etc. • Visit the scene of the accident. At a minimum, develop a plan view of the accident scene and place on the drawing the location of any witnesses. This assumes that the accident scene has not substantially changed since the time of the accident. The visit could possibly be coordinated to be close to the time of day of the accident and, in some cases, the time of year and similar weather conditions. • Inspect equipment or products involved. Develop a checklist for these inspection procedures. • Take photographs and measurements. If human visual perspective in an accident is to be evaluated, be sure to position the camera at the subject’s eye level. Also, to establish sizes and distances of objects, place a measuring tape next to the object being photographed. • Useful accident investigation tools include: still and video camera, audio recorder, 50-ft tape measure, roadway distance-measuring wheel, gauges for measuring push/pull/lift forces, inclinom-eter, level, marking implements such as chalk and adhesive tape, plumb bob, calculator, stop watch, and protractor. • Interview witnesses of the accident. Pay particular attention to the witness’s location or ability to evaluate the accident. Attention should be paid to such factors as the witness’s visual, auditory, timing, and accident-event-sequence perspectives of the accident. • Develop a timeline of events leading up to and during the accident. • Acquire weather reports for the date of outside accidents. Sometimes the sunrise/sunset times and moon phases are also important. Phase III. Analysis, Findings, and Opinions In general, the purpose of this phase is to compare the information and data collected in Phase II with human factors and ergonomics criteria that have been established to provide for adequate human performance and safety in the commercial, industrial, and technological world. This is followed by the development of opinions as to whether the human factors and ergonomics characteristics of the litigation case meet or fail these criteria and whether these characteristics are relevant to the cause of the accident. Generally, this is not a two-step process because opinions are continually formulated during analysis and then are accepted, refined, or refuted as the analysis and findings progress. Thus, analysis, findings, and opinion formulation is an iterative procedure. A suggested three-step technical approach to performing the analysis and developing opinions is given in Section 5.6. The three steps are: (1) a forensic accident analysis model, which guides the investigative procedure; (2) applying this model to a specific accident case; and (3) evaluating the analysis results and findings and developing opinions.


75

The suggested topics and activities to be addressed while performing the analysis, developing findings, and developing opinions are: • Compile and review the information and data collected in Phase II. • Adapt the forensic accident analysis model described in Section 5.6 to the specific type of accident and litigation. • Apply the adapted forensic accident analysis model to the accident scenario and determine potential human factors and ergonomics issues of the accident. • Evaluate the results from application of the forensic accident analysis model using the criteria described in Section 5.6. • Formulate opinions as to the adequacy or inadequacy of the preaccident considerations for the human factors and ergonomics parameters. Phase IV. Reports The purposes of the report are to provide a record of the expert’s credentials and qualifications, the analysis performed by the expert, the opinions developed by the expert, and the basis for the opinions regarding the accident that is the subject of the litigation. The report should be in a format that is acceptable to the expert, the attorney, and the court. In some litigation venues such as local, county, and state courts, a written report is not required. The written report may be in the form of a preliminary report, a final report, or an affidavit or declaration, both of which are notarized (witnessed) signed documents. When a preliminary report that probably will become a piece of evidence is provided, the title should reflect that these are the expert’s findings to date. The preliminary report should state that the analysis is not yet complete, state why the findings are tentative, and identify remaining analysis tasks. The final report should reflect the expert’s final opinions based upon the performed human factors and ergonomics analysis. However, it should state that the expert reserves the right to change or alter the opinions as the facts change or as the understanding of the facts changes. In general, the final report should include sections providing: • When and by whom the expert was contacted, what the assignment was, a short description of the date and accident scenario, scene layout and the persons and equipment involved • The case documentation (e.g., depositions, interrogatories, accident reports) provided and reviewed • Investigation activities conducted by the expert, e.g., review of the case documentation, review of other related literature, product and/or accident scene analysis, and analysis of accident data bases • A description of the findings and opinions regarding the human factors and ergonomics issues of the case, with a statement that the findings and opinions are based on the expert’s general professional experience; on training, knowledge, and education in human factors and ergonomics; and on the aforementioned analysis of the specific issues involved in the investigated accident • A copy of the expert’s curriculum vitae (CV), usually attached as an appendix


76

The formats for the declaration and the affidavit are nearly identical and very similar to a final report. A final report usually has a CV included with its submission; however, the affidavit and declaration are usually stand-alone documents and thus require a section detailing the present and past credentials and qualifications of the expert. Reports submitted to a federal court must follow specific guidelines. Under Rule 26 of the Federal Rules of Civil Procedure, filing of expert witness written reports is mandatory and must follow these guidelines for recording the analysis and opinions: • The report must be prepared by the expert witness. • The expert must sign the report. • The report must include a complete statement of all opinions to be expressed. • The basis and reasoning for each opinion must be included. • The data or other significant information considered by the expert in forming the opinions must be listed. • Exhibits to be used as a summary or as support for opinions must be included. • All publications authored by the expert within the past 10 years, regardless of relevance, must be listed; this is usually done with a copy of the expert’s CV attached as an appendix to the report. • The compensation to be paid for the testimony and analysis must be included; this can be provided by the expert’s schedule of fees as an attachment to the report. • A list of other cases in which the expert gave depositions or trial testimony within the past four years, without regard to relevance or relationship to the subject matter of the litigation, must be submitted with the report. This list may also be provided as an attachment to the report. The topics and activities to be considered for reports include: • The length of the report should be as short as possible, yet include all necessary information and data. Typically, three to four pages, plus attachments, is an acceptable length. • All written reports are evidence and therefore are discoverable; they probably will be used for examination during deposition and/or trial. • The report format should be reviewed with the attorney before beginning preparation, to ensure acceptance with the litigation venue, i.e., the local, county, state, or federal jurisdiction. • The report should be submitted on time. Report submission deadlines are very important for compliance with legal system requirements. Phase V. Deposition and Trial Testimony The overall purposes of deposition and trial testimony are similar in that they provide a forum for the expert witness to present opinions and the basis for the opinions. Deposition Testimony However, a clear distinction can be made in that the deposition is primarily for the opposing counsel through direct examination to discover the expert’s qualifications to be


77

a human factors and ergonomics expert, analyses and investigations that have been performed, concluded findings and opinions, and the specific basis for those opinions. Sometimes when opposing counsel is finished, the engaging counsel will ask clarifying questions before the deposition is concluded. In general, an expert’s deposition will follow the report phase; however, a report is not always required. Usually, the plaintiff’s expert witnesses will be deposed first in sequence, followed by the defense experts. Usually, the opposing attorney will call for, and pay for, the deposition, subject to jurisdiction rules and deadlines. Sometimes before deposition and usually before trial, the engaging attorney will want to conduct an examination practice session. This will occur less frequently as the expert becomes more experienced. At the conclusion of the deposition, the expert will be asked whether or not he would like to read and sign the deposition. It is suggested that the expert do this and indicate “form” only changes on the errata sheet. If possible, he should retain a copy of the deposition for use in preparation for, and at, trial. Trial Testimony In trial testimony, the expert’s credentials to be a human factors and ergonomics expert are first presented through direct examination by the employing counsel. Opposing counsel may challenge credentials and expertise at this point in the trial proceedings and question the expert. Assuming that the expert is acceptable to the judge or court as a human factors and ergonomics expert witness, direct examination will continue by engaging counsel until finished. The opposing counsel now begins cross-examination of the subject matter covered during direct examination. Cross-examination is followed by redirect and recrossexamination, continuing until all areas on a limiting basis have been covered. “Limiting basis” means that only issues included in testimony in the immediately preceding examination may be addressed. The judge or court will then excuse the expert witness from the stand and he should exit the courtroom. From time to time during cross-examination, the opposing counsel will present various challenges to the introduction of human factors and ergonomics testimony at trial. Some of the challenges human factors and ergonomics experts may encounter are: • Opposing counsel has chosen not to use a human factors and ergonomics expert for his own case and now realizes the likely effectiveness of the human factors and ergonomics professional’s testimony. • Opposing counsel challenges the utility of the human factors and ergonomics discipline and information in assisting judicial decisions (i.e., jurors would know based on their own expertise or it is “just common sense”). • Opposing counsel cites the Daubert decision challenging the acceptance of the human factors and ergonomics discipline and its scientific credibility. • Opposing counsel challenges that human factors and ergonomics on “accident causation” is based on opinions of human behavior and will “invade the province” of the jury for finding fault.


78

• In addition, the opposing counsel may challenge (as discussed previously) the qualifications of the human factors and ergonomics expert. The expert must be prepared to establish qualifications as a credible expert. For further discussion of these challenges and specific arguments supporting human factors and ergonomics testimony, see the “Position Paper Supporting Human Factors and Ergonomics Practitioners in Forensics,” produced by the Forensic Professional Group of the Human Factors and Ergonomics Society (Forensic Professional Group, 1996). The following general topics and activities should be considered in preparing for providing testimony in deposition and/or trial: • The three most important words in providing testimony are: preparation, preparation, and preparation. This activity cannot be overstressed. Remember that each accident litigation, regardless of its similarity to prior cases or the expert’s past experience, is uniquely different because the facts are never the same. The people are always different, and the law is probably different. Thus, each case involves new analysis, evaluation, and preparation. • Ensure a thorough understanding of, and basis for, the precise human factors and ergonomic issues in the accident and the case. • When giving testimony, an expert witness should: • Be honest, impartial, and always tell the truth. • Not speculate but, rather, be certain; avoid guessing (saying “I don’t know” if he does not know). • Work within the legal process, its practices, and procedures. • Look and act the part of a qualified professional. Remember that the expert is representing himself, possibly a company, the plaintiff or defense or court, and the profession of human factors and ergonomics. • In trial, face and talk to the jury. • Be articulate and not use sophisticated technical jargon. • Not be an advocate. An expert’s testimony is not intended to win a case or sell a point of view. Its purpose is to help the court reach a just decision. • Control his emotions. Remember that opposing counsel can be very challenging and emotional. This is his job, and if he can get the witness flustered or defensive, he will try to do it. It is all part of the system. • Answer questions responsively. As a general rule, do not volunteer information. Listen closely to the questions and answer what is asked to the best of his ability. If he does not fully hear or understand the question, he should say so and ask for the question to be repeated. • Always stay within his area of expertise. • The Human Factors and Ergonomics Society adopted in 1989 and amended in 1998 a code of ethics to promote and sustain the highest levels of professional and scientific performance by its members (Anonymous, annually). The seven basic principles of the forensic practice section of the code are particularly relevant during deposition and trial testimony. These principles are:


79

• Provide objective, unbiased testimony that is based on credible data and/or scientific principles. • Avoid impugning the integrity of other expert witnesses. • Do not accept fees on a basis contingent on the outcome of the litigation. • Accept that the client is the attorney who engaged you and not the client of that attorney. • Except when required by the Federal Rules of Evidence, avoid discussing litigation with others in a way that would disclose the caption of the suit or the parties involved without the permission of the engaging attorney, until the litigation is resolved. • Do not make public statements likely to influence or prejudice the proceedings. • After suit resolution, do not reveal information detrimental to the parties involved, except when there is evidence of a greater duty of protecting public health and safety. • The following is a general line of direct examination questions that an expert should be prepared to answer when giving testimony in a deposition and/or trial: • Introduction: name, title, age. How long have you been an expert witness? Have you worked for either counsel in this case before? If so, how many times? How much are you paid for giving testimony? • What is your specific field of expertise? • What is human factors and ergonomics and how does it relate to this accident/litigation? • Review of the expert’s resume: education, work experience, present job and daily duties and responsibility, honors, certifications, publications, professional societies, etc. • State your specific experience (including design) regarding the subject matter involved in the specific accident/litigation (to establish why the expert is qualified and credible to be an expert in the case). • How many times have you testified in depositions and trials for plaintiffs and for defendants? • What case materials have you reviewed regarding this accident/litigation? • What general reference documents (e.g., standards, recommended practices) have you reviewed? • What specific analyses have been performed during the investigation of the accident? • Was an accident scene investigation conducted? • Have you inspected the subject product/equipment/process and/or analyzed an exemplar? • Based upon your review of the case materials, reference documentation, and the analyses and inspections that have been conducted, do you have any opinions: yes or no? (If yes, the opinions are then stated, followed by the specific basis for each opinion.) • After case completion:


80

• Retain case file material until the case is settled plus one additional year allowing for appeal or other legal procedures. • Ask the engaging counsel for a critique of performance: testimony and effectiveness of the evidence and exhibits developed. This will help the expert to be a better expert witness in the future.

5.4 Example Cases Illustrating the Road Map This section presents two litigation cases—one plaintiff and one defense—that illustrate the phases of the road map. Plaintiff: Medical Equipment Phase I. Case Initiation A human factors and ergonomics expert was contacted by a plaintiff’s counsel with whom he had worked before on plaintiff and defendant accident litigations. Because of the previous experiences, the expert was very comfortable working with this attorney again. An updated resume and fee schedule were forwarded to the attorney. The initial conversation included a brief description of the accident scenario, activities of the persons involved, medical equipment (product) involved, and accident scene layout. Also discussed were: the case time schedule and critical dates, that the expert did not have any conflicts of interest, the basic role within the field of human factors and ergonomics that the expert would serve, and the general accident and case documentation available. The expert agreed to review the available documentation, begin an investigation of the accident, and set up a file. Phase II. Understanding the Accident Scenario The expert received and reviewed during the discovery phase the following documents from the plaintiff’s attorney • Photographs of the subject medical equipment, including the foot-operated controls • The plaintiff’s deposition • The defendant’s answers to the plaintiff’s interrogatories and motions to produce • The “Installation and Operation Manual” for the subject model equipment • Photographs of an older style of foot control and equipment from the same manufacturer • The defendant’s accident investigation report • The defendant’s sales brochures for the subject equipment • Underwriters Laboratory report on the subject equipment • Diagrams of the on-product labels • The defendant’s engineering change orders for the subject equipment from the older to the newer style • The defendant’s service manual and service bulletins for the subject equipment


81

• Depositions of the defendant’s manager of engineering and the manager of quality assurance • Deposition of the plaintiff’s supervising physician • The design manufacture drawings of the subject equipment The expert did not visit the accident scene or inspect the subject equipment. However, he did inspect and take measurements of the defendant’s exemplar equipment and a competitor’s similar equipment. No witnesses were interviewed. The following accident scenario was developed and used by the expert. While he was operating medical equipment, the technician received a crushing foot injury. The equipment employs a multiswitch foot control to control various functions, including equipment position. The foot control is tethered to the equipment and can be moved to any number of positions for operation. While he was assisting the physician in performing a medical procedure on a patient, the technician was operating the equipment using the foot control. The foot control was in a position under the equipment, so when the equipment was lowered to a set position, it made contact and trapped the operator’s foot on the continuing-to-function control. The weight and continuous force of the equipment on top of the operator’s trapped foot caused the injury. Phase III. Analysis, Findings, and Opinions Analyses included the following activities: • Review of all the documents listed in Phase II and summarization of the relevant information • Application of the forensic accident analysis model to the accident scenario and determination of potential human factors and ergonomics issues • Search for and review of evaluation criteria documents such as medical equipment industry and government regulations, standards, recommended practices and guidelines, human factors and ergonomics design principles, data, and methods from human factors and ergonomics handbooks and guides • Search for similar accidents • Review of defendant’s previous subject equipment designs and similar equipment designs • Competitors’ designs Based upon the expert’s analysis of the accident, the following findings and basic human factors and ergonomics opinions were reached. The first basic opinion was that the medical equipment and its foot-operated control were unreasonably dangerous based upon inadequate human factors and ergonomics design. Inadequate design and guarding issues included the unguarded, movable foot control and the single, two-function pedal built into the unshielded foot control. Inadequate warnings were found on the product and in the accompanying operator’s manual with regard to the accident scenario’s identified hazard and risk. Alternative designs were available on similar medical equipment that exemplified good human factors design criteria and principles. The second basic opinion was that if the design and manufacture of the subject medical equipment and foot control had been consistent with sound human factors and


82

ergonomics principles, the plaintiff would not have experienced the accident and corresponding injuries. Phase IV. Report No report was requested by the plaintiff’s attorney. Phase V. Deposition and Trial Testimony The human factors and ergonomics expert provided a 2-hour deposition based upon his investigation, findings, and opinions. The case settled prior to trial. Defense: Farm Equipment Phase I. Case Initiation A human factors and ergonomics expert was contacted by a defense counsel with whom he had worked before. During the conversation, the attorney provided: a brief description of the accident scenario, the activities of the persons involved, a description of the farm equipment involved and the defendant manufacturer, and the accident scene layout. Also discussed were: the case timeline, that no conflicts of interest existed for the expert, the basic human factors and ergonomics role that the expert would play, identities of the other defense experts and their expertise, and the availability of case documentation. The expert agreed to review the available documentation, begin an investigation of the accident, and set up a file. An updated resume and fee schedule were sent to the attorney. Phase II. Understanding the Accident Scenario The expert received and reviewed the following documents from the defense attorney: • Depositions of the plaintiff, the farm owner, the plaintiff’s supervisor, the equipment’s installer, and the plaintiff’s experts • The plaintiff’s medical records, including photographs of the foot and leg injuries • Photographs of the subject equipment and accident scene • The plaintiff’s answers to defendant’s interrogatories • Equipment “Assembly and Operating Instruction Manual” • The employer’s accident report of injury (signed by plaintiff) • The plaintiff’s notice to produce to defendant and the answers • The defendant manufacturer’s design and manufacture drawings of subject equipment • The defendant’s sales brochures • Competitors’ sales brochures The expert did not visit the scene of the accident or inspect the subject farm equipment. The expert did, however, rely on accident scene photographs taken by three different sources, as well as measurements of equipment sizes and distances made by plaintiff and defense investigators. The measurements proved to be particularly useful throughout the


83

investigation and during trial testimony. These measurements were also made into a scaled drawing of the equipment and accident scene. The expert requested and received, through defense counsel, certain anthropometric measurements of the plaintiff, such as foot and shoe size, ankle height, knee height, arm length, and stature. No witnesses were interviewed. Phase III. Analysis, Findings, and Opinions The following analyses were performed: • Review of all the documents listed in Phase II • Summaries of relevant information • The forensic accident analysis model applied to the accident scenario for evaluation of the human factors and ergonomics inadequacies identified by plaintiff’s experts • Biomechanical/anthropometric analysis of the plaintiff’s size in relation to the equipment size and operational characteristics • Review of the evaluation criteria documents identified by the plaintiff’s experts, such as American Society of Agricultural Engineers Standards. • Evaluation of the subject equipment’s warning labels and manuals for adequacy with regard to the specific accident scenario’s hazards and risks and the plaintiff’s training and experience • Evaluation of the alleged need for a remote control Based upon the expert’s analysis of the accident, the following findings and opinions were reached. The first opinion was that the accident did not happen in the way in which the plaintiff testified. Instead, the evidence and the expert’s analyses led to the conclusion that a different scenario had probably occurred. The second opinion was that the warnings on the subject equipment were adequate to advise anyone of the hazards and risks presented, especially someone with the plaintiff’s years of experience, awareness, and training in working with this type of equipment. The third opinion was that providing a remote control switch would not have prevented this accident.

TABLE 5.1 Checklist of Main Considerations by Phase Phases I. Case initiation

Considerations

Nature of case/time schedule Yes/no work on case Financial arrangements Available information/data II. Understand accident scenario Environmental Equipment Human Process Organizational III. Analysis, findings, and opinions Forensic accident analysis model Adapt model to accident scenario Analysis using model (five elements)


IV. Report

V. Deposition and trial testimony

84

Evaluate findings by criteria hierarchy Develop opinions Proper format Short length Timely Discoverable evidence Thorough preparation Examination questions Behavior during testimony Code of ethics

Phase IV. Reports No report was requested by the defense attorney. Phase V. Deposition and Trial Testimony The human factors and ergonomics defense expert gave a 2-hour deposition and testified at trial. The expert’s exhibits, including the equipment’s scale drawing and articulated anthropometric model of the plaintiff, were especially effective in demonstrating that the accident scenario did not occur in the way in which the plaintiff had testified. The trial resulted in a verdict for the defendant manufacturer. 5.5 Checklist of Main Considerations by Phase A summary checklist of the main considerations for the five phases is given in Table 5.1. 5.6 Technical Approach to Analysis, Findings, and Opinions An aid to performing an analysis and developing findings and opinions is a three-part technical procedure. The procedure includes: (1) a forensic accident analysis model, (2) applying the model to the specific type of accident and litigation, and (3) evaluating results of the analysis and developing findings and opinions about the human factors and ergonomics conditions of the accident scenario. Each of these topics is discussed in this section. Forensic Accident Analysis Model A product safety model advocated and used by Christensen (1981) served as the basis for the forensic accident analysis model. The evolved model has five steps that lead the analyst through the procedure to determine the adequacy or inadequacy of preaccident considerations of the human factors and ergonomics parameters relevant to the accident scenario. The preaccident concept is crucial because the analyst is attempting to


85

determine if efforts were made before the accident to provide for human performance and safety. The five steps of the analysis are: • Step 1: Design. Evaluate the design of the product, equipment, or process for its adequacy to have provided for effective and safe human interfaces such as installation, use, and maintenance. • Step 2: Remove. Assess the feasibility of having removed the human from a potentially unsafe condition, for example, a remote control station. • Step 3: Guard. Evaluate the adequacy of guarding to protect the human from potential hazards. • Step 4: Warn. Evaluate the adequacy of warnings to have informed the human about potential hazards and risks, to reinforce prior knowledge about hazards and risks, and to provide the steps to take to avoid the hazards. • Step 5: Train. Evaluate the adequacy of training about potential hazards and risks and the proper procedures to avoid the potential hazards and reduce the risks. This general forensic accident analysis model must now be applied to the specific type of accident under investigation and the unique litigation. Applying the Forensic Accident Analysis Model to the Accident Scenario All five elements of the analysis model might not necessarily apply to a given accident litigation. Therefore, the model needs to be used selectively, based on the characteristics of a given litigation accident. The tailoring generally would be of the type to select which one or more of the five elements apply to the specific accident case. This would be followed by an analysis and evaluation of the adequacy or inadequacy of the human factors and ergonomics considerations for the selected model elements. Information and data that can be used as criteria for the evaluations of the results of the analyses will be discussed later. For the remainder of this section, suggestions are provided for the usefulness of the five steps of the model for different types of accident cases. Potential usefulness of the five steps of the model is illustrated with four different types of cases: consumer product accidents; motor vehicle accidents; industrial machine or process accidents; and slip, trip, and fall accidents. Consumer Product Accidents Accidents of this type cover a wide range, from simple cans of spray paint or insect spray purchased off the shelf in stores to complex products such as pleasure boats and allterrain vehicles (ATV) purchased from dealers. The Consumer Product Safety Commission directory lists hundreds of consumer products, which have a great variety of human factors and ergonomics interface issues. The following suggestions are general information about applying the product safety model to consumer product accident cases; in no way can they reflect the full range and diversity of consumer products. Other chapters in this handbook provide additional information about consumer product accidents.


86

• Design. Design can certainly apply to consumer products. However, there is no single statement as to. the human factors and ergonomics design parameters for consumer products, given that such a broad range of product types is available. Design parameters can be as simple as the location and coding of a pushbutton on a spray can to the full range of operator and maintainer interface topics for a pleasure boat or an ATV. The intent during analysis is to understand the adequacy or inadequacy of the design as related to the accident. This may be quickly determined or may require extensive analysis. • Remove. The remove topic generally would not apply to consumer products, although it might be relevant to human tasks for some types of products. A historic example is the handle control start for lawn mowers, which reduces the risk of contact with the rotating blade while an individual is starting the mower and reduces the strain on the user’s body. • Guard. The guard topic would generally apply to the more complex consumer products that have mechanical or electrical hazards—for example, guards over switches to prevent inadvertent activation of a power source, or guards over moving parts to prevent accidental contact by a body part. • Warn. Warnings apply to all types of consumer products and have become one of the primary human factors and ergonomics issues in litigation. The simplest spray cans might require information about the hazardous nature of the propellant. The more complex products might need to warn about hazards for a variety of operating and maintenance procedures. Warnings can have many forms for consumer products, such as on-product labeling or on-packaging labeling, an information insert in the product packaging, information in the user manual, and information in advertising literature such as catalogues and brochures. Warnings may be appropriate for assembly, installation, operation, maintenance, and disposal of the product. Details on the subject of warnings may be found in other chapters in this handbook, as well as in Laughery et al. (1994) and Wogalter et al. (2001). • Train. For consumer products, the train topic applies to a wide range of situations, e.g., instruction labels on products, user guides accompanying the product, and customer training courses for the more elaborate products, such as pleasure boats or ATVs. Motor Vehicle Accidents The concern here is primarily with roadway vehicles under the regulation of the National Highway Traffic Safety Administration (NHTSA) and the Federal Highway Administration (FHA). Other chapters in this book provide additional information about motor vehicle accidents. • Design. Design is obviously very relevant to safe use of motor vehicles, but this is an extremely complicated topic involving the vehicle and the roadway and is best understood by reading publications produced by the SAE, FHA, NHTSA, and Peters and Peters (1993). The best that can be said here is that design, as it interfaces with human factors and ergonomics capabilities and limitations, can play an important part in proximal causes of motor vehicle accidents. • Remove. The remove factor is relevant and could apply to motor vehicles and to roadways. It generally is most relevant to roadway design, e.g., roadway overpasses to


87

avoid vehicle collisions at intersections and at railroad crossings. Useful references on this topic include publications provided by FHA and state Departments of Transportation. However, in some instances, vehicle design may remove the user from a hazard, e.g., the historical electric starter to replace the hand crank with its potential injury to the body. In the future, “removal” may become more intellectual by “removing” the driver from certain decisions regarding braking, steering, accelerating, etc. • Guard. Guarding is certainly relevant. Examples of the guard factor in vehicle design are seat belts, air bags, occupant interior padding, and guarding of the fan in the engine compartment. The guard factor is relevant also to roadway design, e.g., guardrails and slow-compressing containers on concrete posts. • Warn. Warnings would have a variety of applications, such as vehicle operator manuals, labels mounted on the vehicle (particularly in the occupant compartment), and roadway signs of potentially hazardous conditions, e.g., falling rocks and ice on bridges. • Train. Training could have a wide range of implications, from vehicle operator instruction manuals to formalized training programs for commercial drivers. Industrial Machine or Process Accidents The focus here is on the industrial manufacturing arena. There is a wide range of application of the accident analysis model in this area, from simple punch press and manual lathe equipment to very complex computer numerical control machines and automated assembly operations. Publications by OSHA, ANSI, NSC, and NIOSH are instructive on this topic. Other chapters in this book provide additional information about industrial accidents. • Design. Design for safety can range across the spectrum of industrial machines and processes, such as location of controls on basic machines to the layout and arrangement of work stations for complex process operations comprising many machines, conveyors, and workers. • Remove. Removing the human from contact with hazards by such techniques as remote manipulators, robots, and remote control panels is becoming an increasingly useful procedure for worker safety; see, for example, a NIOSH publication on the subject of robotic workstations (Etherton, 1988). When an industrial accident is investigated, the expert should ask, “Could the process have been designed to remove the worker from the hazard?” Certainly, economics and technical feasibility must be considered, but the question should be explored. • Guard. Guarding to prevent injury to workers is certainly relevant and has a long history of technical achievement with a variety of guarding techniques. A useful reference is Blundell (1983). The human factors and ergonomics expert should definitely look at the adequate use, or non-use, of guards for the specific accident. • Warn. There is a long history of the use of warnings in the industrial world, ranging from workplace signs and placards and on-machine labels to more elaborate auditory and visual signals. The use or non-use of warning should definitely be examined for a specific accident.


88

• Train. Training is certainly an important element in the industrial world. Generally, most industries provide training for their employees. For a specific accident case, questions should be asked about whether any training was provided and, if provided, whether it instructed about avoiding the conditions that led to the specific accident. Slip, Trip, and Fall Accidents Slip, trip, and fall accidents occur frequently, and the forensic accident analysis model also can be a useful approach for these types of accidents. Useful references to understanding the dynamics of slip, trip, and fall accidents are publications by NSC and ANSI. Other chapters in this book provide additional information about these types of accidents. • Design. Design is relevant, e.g., geometry of steps, slip resistance of the walking surface, handholds, and defects in the walking surface due to failure to maintain an adequate design. • Remove. The remove element is not really relevant, unless we visualize removing the plaintiff from the necessity of walking on a surface. • Guard. Guarding is certainly important, for example, the adequacy of railings, handholds, fences, grates, and temporary barricades. • Warn. Warning about hazards on or near a walkway is relevant, for example, signs about repairs in progress, water or ice on the surface, and edge drop-offs. A fairly recent type of warning is the auditory warning on moving walkways in airports that alerts the traveler that the end of the automated walkway is approaching. • Train. The train element is not generally applicable to slip, trip, and fall accidents. Evaluating Analysis Results and Developing Findings and Opinions Assuming that one or more of the accident analysis model elements were found relevant to the specific accident and litigation and that human factors and ergonomics information and data were compiled for the relevant model elements, the next step is to evaluate the adequacy or inadequacy of these findings as related to human performance and safety. Therefore, this section deals with the types of information and data that can provide the criteria against which to determine the adequacy or inadequacy of the design/remove/guard/warn/train findings for a given accident scenario and to develop opinions as to their effect on the accident. The information and data are grouped into categories, which are listed in a hierarchy that reflects the general legal acceptance of the information and data to support the development of opinions for forensic cases. The categories, in descending order of generally acceptable information and data, are: • Government codes, regulations, and law, e.g., OSHA, CPSC • Voluntary industry and association standards, e.g., ANSI, ASTM • Industry customs, guidelines, and recommended practices, e.g., SAE, ASAE • Professional handbooks, e.g., Human Factors Design Handbook • Historical accident databases, e.g., CPSC, NHTSA • Scientific literature, e.g., professional journals


89

• Empirical study, e.g., consumer use test, user surveys • Professional judgment, e.g., safety analysis of a tool or procedure As stated earlier, the significance of this hierarchy is that an evaluation of findings, as well as the development of opinions based on the information and data obtained from categories near the top of this hierarchy, generally carries more weight in the legal system. This is not to repute the lower-ranking categories because sometimes a wellconstructed empirical study focused on the accident case scenario in the absence of documented human factors and ergonomics data can provide powerful opinion data. Also, the opinions of a nationally recognized professional may be well received in the legal system. The opinions about the adequacy or inadequacy of the human factors and ergonomics parameters of the specific accident scenario to provide for safe human performance generally emerge during the course of the analysis and evaluation of the findings. As the analysis progresses through the five steps of the accident analysis model and the findings are tested against the evaluation criteria, the opinions often become self-evident. It should be remembered that this is an evolving, iterative procedure, with opinions gradually developed and refined over time before their presentation in the form of a written report or orally in deposition and/or trial. Defining Terms ANSI—American National Standards Institute ASAE—American Society of Agricultural Engineers ASTM—American Society of Testing and Materials ATV—All-terrain vehicle CPSC—Consumer Product Safety Commission CV—Curriculum vitae (aka resume) FAAM—Forensic Accident Analysis Model FHA—Federal Highway Administration NHTSA—National Highway Transportation Safety Administration NIOSH—National Institute of Occupational Safety and Health NSC—National Safety Council OSHA—Occupational Safety and Health Administration SAE—Society of Automotive Engineers References Anonymous, Annually, Code of ethics, Article V, forensic practice, in Directory and Yearbook , Santa Monica, CA: Human Factors and Ergonomics Society. Blundell, J.K., 1983, Machine Guarding Accidents , Columbia, MD: Hanrow Press, Inc. Christensen, J.M., 1981, Forensic human factors psychology, in ASME-ASNT-NBS Symposium on Non-destructive Evaluation: Reliability and Human Factors , Washington, D.C.: National Bureau of Standards. chap. 3.


90

Etherton, J.R., 1988, Safe maintenance guidelines for robotic workstations, Morgantown, WV: NIOSH Technical Report. Forensic Professional Group, 1996, Position paper supporting human factors and ergonomics practitioners in forensics, Santa Monica, CA: Human Factors and Ergonomics Society. Laughery, Sr., K.R., Wogalter, M.S., and Young, S.L., 1994, Human Factors Perspectives on Warnings , Santa Monica, CA: Human Factors and Ergonomics Society. Peters, G.A., 1971, Product Liability and Safety , Washington, D.C.: Coiner. Peters, G.A. and Peters, B.J., 1993, Automotive Engineering and Litigation , Vols. 1–6, New York: John Wiley & Sons. Poyntrer, D., 1987, Expert Witness Handbook , Santa Barbara, CA: Para Publishing Company. Weinstein, A.S., Twerski, A.D., Piehler, H.R., and Donaher, W.A., 1978, Product Liability and the Reasonably Safe Product , New York: John Wiley & Sons. Wolgalter, M.S., Young, S.L., and Laughery, Sr., K.R., 2001, Human Factors Perspectives on Warnings , Volume 2 , Santa Monica, CA: Human Factors and Ergonomics Society.

6 Can Training for Safe Practices Reduce the Risk of Organizational Liability? Katherine A.Wilson University of Central Florida Heather A.Priest University of Central Florida Eduardo Salas University of Central Florida C.Shawn Burke University of Central Florida 0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

6.1 Introduction to Human Factors Principles It has been well documented that human error contributes to accidents and incidents more than two thirds of the time (see Decker, 2001; Freeman and Simmon, 1991; Helmreich, in press; Kletz, 1985), leading to liability issues for the individual and even the organization. In an effort to reduce accidents, the Occupational Safety and Health Administration (OSHA) was created in 1970 and has since reduced the number of workplace deaths by 50%. However, the incidence of workplace accidents remains high. For example, it has been estimated that approximately 5.7 million worker-related injuries/illnesses and almost 6000 deaths occurred in private-sector firms in 2000 (http://www.osha.gov/). In other words, despite safety laws and regulations for organizations, a significant number of injuries continue to occur each year. Because many of these accidents and incidents are due to human error, organizations have turned to training as a way to teach employees how to mitigate and capture errors before they become detrimental. Training in today’s organizations is “big business” and is estimated to grow as organizational complexity increases the propensity for human error. It is estimated that organizations spend between $55.3 billion and $200 billion each year on training employees (Salas and Cannon-Bowers, 2001; Bassi and Van Buren, 1998). Although the purpose of training in organizations varies (e.g., improve safety, increase human capital, increase competencies, better product quality, error reduction), the overarching theme is to create a more productive and safe work environment for employees as well as for consumers.


92

Many organizations believe that by providing individual training as a means of reducing human error and improving safety will minimize the organization’s threat of legal liability. However, often the training provided by the organization is inadequate, thus opening the door to litigation against the organization. The most likely cause of inadequate training is its improper design, delivery, implementation (i.e., transfer of training), and evaluation. Training research has been around for several decades (see Campbell, 1971; Goldstein, 1980). Over the past decade, there has been an explosion in the science of training. The research conducted on training in the 1990s led to the expansion of theoretical frameworks, concepts, and constructs (Salas and Cannon-Bowers, 2001). The training field can now offer principles and guidelines based on sound theories to practitioners and instructional designers. In addition, new models to design training and better tools to evaluate it have been developed. Yet, despite this research, its impact on organizational practice has not been seen. Salas and colleagues (1999; in press) argue that the reason for this is the many myths and misconceptions about the science of training and its practice to which many designers fall prey. Thus, training is often designed based on assumptions that cannot be supported, and it is not always effective. Effective training, therefore, relies on incorporating sound principles of learning in the design and delivery, systematic evaluation of what was taught, and ensuring the transfer of the newly acquired skills to the job. In sum, when not designed and delivered correctly, training can represent an organizational liability. 6.2 Objective and Scope The purpose of this chapter is threefold. First, we provide an understanding of the principal issues that influence safe behaviors in organizations. Next, we discuss how training can be used to improve safe behaviors and the factors that must be considered when designing, delivering, evaluating, and ensuring transfer of training. Finally, we offer guidance (i.e., checklists) for organizations and training developers to help in the design, delivery, implementation, and evaluation of training to reduce the potential risk of organizational liability. 6.3 Discussion of Principal Issues What Is Human Error? Although often forgotten, human capabilities are limited. This, in turn with organizational and environmental complexity, makes error practically inevitable. Human error has been defined as any “occasion in which a planned sequence of mental or physical activities fails to achieve its intended outcome” as a result of an inadequate plan or of intended actions not going as planned (Reason, 1990, p. 9). Errors can originate from many different sources, including inadequate training and inadequate procedures (Helmreich and Merritt, 2000).

Can training for safe practices reduce the risk of organizational liability?

93

Reason (1990) has classified errors into three basic types: slips, lapses, and mistakes. Slips and lapses are unintentional actions that result when the execution and/or storage stage of an action plan breaks down (Reason, 1990; Shappell and Wiegmann, 1997). More specifically, a slip is a failure that can be observed (e.g., slip of tongue or action), whereas a lapse is typically not observable (e.g., lapse of memory) and thus may only be detected by the person who errs. Finally, a mistake is an inappropriate action that results when one’s judgment and/or inferential processes fail. Mistakes, like lapses, are often more difficult to detect because they are usually more complex, subtler, and less understood than other types of errors. Kletz (1985) and Reason (1990) argue that errors, specifically slips, can be minimized through improved training, instructions, and/or procedures. Reason (1994) argues that humans do not intentionally commit the acts that lead to errors; even highly qualified individuals commit them. On the other hand, some acts— violations—are not errors because they are intentional. Violations involve consciously bending the rules for something that at the time seemed reasonable and/or an acceptable risk. Yet, if organizations are not careful, these violations can become a normalization of deviance in which risk becomes the norm (i.e., maladaptive norms; Vaughan, 1996). Normalization of deviance can be characterized by a five-step process: (1) awareness of a problem, (2) unsafe behaviors seen deviating from the norm, (3) problem investigated, (4) working norm revised to “normalize” the problem, and (5) risk now deemed acceptable. An example of a seemingly harmless violation that led to a maladaptive norm and disastrous consequence involved the space shuttle Challenger (Pidgeon, 1998; Vaughan, 1996). In this case, erosion problems were identified with the rocket booster O-rings (steps 1 and 2). A work group was established to investigate the concerns (step 3). They estimated the probability that a hazardous condition would result to be unlikely and thus expanded the “normal” boundaries for the erosion (step 4). As a result, additional erosion incidents were deemed an acceptable risk (step 5). This deviance was also influenced by the organizational culture within NASA’s shuttle program, which was characterized by features such as interpretive flexibility, cost/safety compromises, and lack of appropriate guidelines. Who Is to Blame? Once an error has been committed and realized, typically the next question concerns where to place the blame. Using the Challenger explosion as an example, the following scenarios are possible. First, the equipment could be blamed because ultimately it was the equipment that failed and resulted in the accident. Next, the work group that investigated the erosion problems could be blamed because it created and enforced the deviant norms. Finally, the organization could be blamed because it provided a culture for its employees that tolerated behaviors that deviated from the norm. Ultimately, all three sources (i.e., equipment, individuals, and the organization) could potentially be to blame. The society in which we live today is one filled with blame, especially of the “fallible” human. It is argued that placing blame on the individual is a universal phenomenon that is emotionally satisfying (because it is human nature to want someone to blame) (Reason, 1994), and it may even be politically and legally convenient. In other words, it is easier,


94

natural, and more satisfying to blame a single individual than an entire organization. Although, from a legal perspective, it may be easiest to place the blame on an individual (Reason, 1994), all of the blame cannot be placed there. People and organizations are generally quick to place blame on the human and fail to recognize the important role that the organization may have played in the events leading up to an error (e.g., inadequate training, lack of safety culture, lack of safety procedures). However, this trend is beginning to change. Organizations are now being held accountable for improper training and unsafe behaviors of their workforce. For example, many law enforcement organizations are now held accountable for officers’ less-than-adequate performance on the job. In the case of Sampson v. City of Schenectady, two officers were charged with violating the defendant’s Constitutional rights when the officers took him to a remote location without justifiable reason. The officers claimed that they were not trained properly for their encounter with the defendant and that the organization (or city of Schenectady) had an “unofficial relocation” policy of individuals. This unofficial policy could ultimately put those being “relocated” in an unsafe situation. As such, the officers sought immunity by shifting some, if not all, of the liability to the city of Schenectady (also see Dunton v. County of Suffolk, 1984). This example shows how improper training and maladaptive procedures can lead to legal liability for the organization. (Refer to Section 6.4 for information on similar cases.) How Can Training Help? Training can be defined as the systematic acquisition of knowledge (i.e., what we think), skills (i.e., what we do), and attitudes (i.e., what we feel) (KSAs) that lead to improved performance in a particular environment (Salas et al, 1992). We recognize that much of the literature refers to the “A” as abilities (e.g., Goldstein, 1993); however, for the purpose of this chapter, we will refer to the “A” as attitudes. In addition, training is influenced by multiple factors prior to, during, and following training—all of which should be considered to ensure a training program’s effectiveness. When human error is combined with failures within the organization, such as training deficiencies and lack of safety culture, an accident or incident is likely to become imminent (Isaac, 1997). The previous example set in the law enforcement community stresses the importance of creating a training program that follows sound theoretical principles and guidelines. When designed and implemented correctly, training can help organizational members mitigate errors and increase safety (see aviation example, Section 6.4). However, improperly designed training can lead not only to unsafe organizational practices, but also legal liability for the organization (Goldman, 2000). Furthermore, simply providing an off-the-shelf, generic training program does not guarantee its effectiveness or protection from the law. Therefore, we will next present the steps that one would take to design a sound training program that targets safety, as well as additional factors to consider outside the training to help ensure desired outcomes and overall effectiveness. The development of a theoretically based and defensible training system should be done systematically (Salas and Cannon-Bowers, 2000b). Using sound theoretical underpinnings, instructional strategies are developed using available tools (e.g., needs analysis), and by incorporating delivery methods (e.g., simulation) and focused content


95

(i.e., competencies) specific to the needs of an organization (Salas and Cannon-Bowers, 1997, 2001). In other words, training creates an environment in which trainees learn necessary competencies (i.e., knowledge, skills, and attitudes—KSAs); to practice using the learned KSAs; and to receive constructive feedback on their performance. The steps for systematically developing a training program will be described next (see also Checklist 6.1). Training-Needs Analysis To deliver a defendable training system, training designers must first understand the tasks to be performed, the competencies that will be required of the worker, and worker/trainee characteristics (Salas and Cannon-Bowers, 2000b). Therefore, one of the earliest and most crucial steps is to conduct a training-needs analysis (Goldstein and Ford, 2002). The purpose of this analysis is to determine who needs to be trained for which tasks, and which KSAs need to be trained; this ultimately leads to the development of learning objectives. Three types of analyses need to be conducted: organizational analysis, job/task analysis, and person analysis. Organizational Analysis An organizational analysis is conducted to establish the organizational components (e.g., climate, norms), resources, and constraints that may affect how the training program is delivered (Salas and Cannon-Bowers, 2001; Goldstein, 1993). In addition, this analysis focuses on determining how well the training objectives (described later) fit with organizational factors (e.g., resources, goals, strategic focus, constraints). For example, an organization whose goal is safety should develop a training program that supports an environment for safe practices (e.g., error management). Finally, this analysis determines characteristics within the organization that will support the transfer of training. For instance, in order for the newly acquired skills to be transferred to the working environment, it is important that the organization provide a positive transfer climate that supports the behaviors that trainees learn (Rouiller and Goldstein, 1993; Tracy et al., 1995). Because we argue that transfer of training is an important factor influencing the outcomes of training, this will be described in more detail later (see “External Factors”). As such, it is important that organizations pay close attention to the organizational characteristics that may influence training because they may have a significant impact on the outcomes of training. Job/Task Analysis Another tool of the training needs analysis is the job/task analysis, the purpose of which is to analyze the job/task of those who are to be trained. More specifically, this analysis seeks to uncover the job description, task specifications, and task requirements (Goldstein, 1993). The first step is to determine the job/task description that needs to be trained, which generally consists of essential work functions of the job and the resources needed (e.g., materials, equipment). The next step is to determine the task specifications, which include the specific tasks to be performed and the conditions under which the job


96

is to be completed. Finally, the requirements of the task are determined. The requirements describe the competencies needed to perform the specific tasks effectively. Although it is relatively easy to determine the job description and task specifications, it is often more difficult to determine the needed KSAs. Training designers tend to focus on readily observable competencies (i.e., skills) and pay less attention to those that cannot be observed (i.e., knowledge and attitudes). However, as tasks become increasingly complex and require greater cognitive demands for decision-making and problem-solving skills, the need to understand the competencies required for the task can be seen. As such, one tool to help accomplish this is cognitive task analysis. Cognitive Task Analysis To determine these KSAs, training designers can use cognitive task analysis techniques (e.g., knowledge elicitation) to gain insight from subject matter experts (SMEs) on their cognitive processes and requirements (Cooke, 1999). Cognitive task analysis (CTA) has been defined as “a set of methods to elicit, explain, and represent the mental processes involved in performing a task” (Klein and Militello, 2001, p. 168). Some argue that CTA is most appropriate for analyzing tasks that are cognitively demanding in complex, dynamic environments (such as those in which safety is of concern) (Gordon and Gill, 1997). Using CTA, training designers attempt to construct a model of what the expert knows (i.e., his knowledge) (Cooke, 1999). This helps designers to determine not only which KSAs are needed to perform a task, but also the cues and cognitions that facilitate a trainee’s decision to apply the appropriate skills (Salas and Cannon-Bowers, 2001). Several knowledge elicitation methods can be used to obtain information from experts (e.g., interviews, verbal protocols, observation, and conceptual methods; see Cooke, 1999). Guidelines for conducting a cognitive task analysis are presented in Checklist 6.2. Klein and Militello (2001) argue that three criteria must be met for CTA to be successful: • An important discovery needs to be made. More specifically, the CTA should identify something new regarding judgments (e.g., patterns), decisions (e.g., strategies), and/or other cognitive demands (e.g., cue patterns). • Effective communication needs to take place. Success of this criterion means that those conducting the CTA must effectively communicate the findings in the discovery phase to the training designers so that the information may be utilized in training development. • The CTA must have impact, which is achieved when the findings of the CTA are successfully put into action (e.g., safety interventions). Person Analysis The third type of analysis that should be conducted as a part of the training-needs analysis is a person analysis, the purpose of which is to determine who needs to be trained and the training that each person needs (Tannenbaum and Yukl, 1992). This analysis ensures that the right people get the right training. For example, training for new employees has different needs and requirements than those of recurrent training for existing employees (Feldman, 1988). Additional research suggests that different-level managers require different skills; specifically, more administrative skills are needed by


97

lower-level managers (Ford and Noe, 1987). Therefore, an effective analysis must be conducted to determine the level of difficulty and job specificity needed for training to be adequate. The person analysis also determines whether trainees have the necessary KSAs and motivation to be trained. If trainees do not meet the training prerequisites, the training will not be as effective as it could be and may require additional training in the future; legal liability for the organization may result. Designing Training The information gained from the training needs analysis phase drives the objectives of the training, which should be specific, measurable, and task relevant so that they can be evaluated when training is completed. The training objectives, which serve to help guide training, have three general characteristics (Goldstein, 1993). Training objectives: • State what the trainee should be able to do as a result of the training (i.e., performance) • Describe the conditions during which the performance, stated previously, should occur • Provide a description of acceptable performance criterion, i.e., how the trainee should perform (at what level) to be judged acceptable In sum, training objectives state the competencies that trainees are expected to acquire and demonstrate once the training is complete. Once the objectives have been clearly defined, they are used to guide the training strategies that need to be implemented. The strategies are selected based on their effectiveness on promoting the task-relevant behaviors and competencies determined in the objectives. Instructional Strategies Once the training objectives have been established, the next step is to determine the instructional strategies to be used during training (i.e., how to train the requisite safe behaviors). Numerous instructional strategies have been developed over the last several decades that can prepare individuals and teams to increase their KSAs, reduce errors, and increase their expertise in performing their tasks—ultimately leading to safe behaviors. When individuals are trained to perform safe behaviors, we argue that there are three important issues. • Trainees should learn to be adaptable to changing situations and to recognize when things go wrong. By training flexible knowledge structures (i.e., cognitive representations), trainees can adjust their behavior to compensate for any changes. Rigid knowledge structures in a changing environment could lead to errors. • All training strategies must provide trainees with constructive feedback that focuses on the task. This feedback allows trainees to compensate for incorrect behaviors and readjust or correct their strategy to be more appropriate for a given situation. • Training needs to be dynamic (i.e., interactive). A recent report suggests that almost 84% of all companies use classroom-based and instructor-led training (Bassi and Van Buren, 1998). In addition, it was discovered that the most commonly used delivery methods (approximately 90% of the time) were videotapes and workbooks, compared to only 10% that used interactive, digital technologies. Computer-based or other technology-based training was used less than 35% of the time. Therefore, we argue


98

that training for safe practices must involve the trainees through practice (e.g., roleplay, simulations). Table 6.1 presents some frequently used instructional strategies that may be used to improve safe practices in the workplace. We focus on four: scenario-based training, error training, stress-exposure training, and team training. Scenario-Based Training Scenario-based training (SBT), also known as event-based training, is unique and especially helpful in training individuals and teams to exhibit safe behaviors (Fowlkes et al., 1998). This type of training provides the opportunity to embed learning events into scenarios (determined from critical incidents data), thus giving trainees a meaningful framework by which to learn (Salas and Cannon-Bowers, 2000a; Fowlkes et al., 1998). SBT provides trainers with valuable tools, including guidelines and steps to achieve training objectives, trigger events, measures of performance, scenario generation, exercise conduct and control, data collection, and feedback. Figure 6.1 provides a graphical representation of the components or cycle of SBT (Cannon-Bowers et al., 1998). The cycle begins with determining the tasks and competencies that the training will focus on (circle 1), followed by the development of training objectives (circle 2). Steps 1 and 2 serve to drive the development of the scenarios (i.e., events, exercises, curriculum) to be embedded into the training (circle 3). Next, performance measures must be developed (circle 4) to evaluate if the trained skills are being applied. Once performance data have been collected, feedback must be provided to trainees (circle 5). Finally, information pertaining to trainee performance (i.e., skill inventory, performance history) must be incorporated into future training programs so that it can build upon previous training (circle 6). A key reason why this method, when applied appropriately, is effective in improving safety is because it is practice-based. Trainees are able to practice in preplanned scenarios while their performance is evaluated. The embedded events serve as cues that trigger behaviors and competencies, which can then be observed, evaluated, and incorporated into feedback. Feedback can be given to trainees immediately and corrections in performance can be made if necessary. In addition, SET supplies trainees with instructional strategies that allow them to practice the trained skills in order to perform effectively on the job.

TABLE 6.1 Training Strategies Strategy Event-based training (EBAT) or Scenario-based Training (SET)

Definition

Level

Individual Provides trainees with a structure in dynamic, complex environments by embedding scenarios and team in work environments that trigger targeted behaviors in complex environments. Instructional strategy designed to structure training in complex, distributed environments. Provides guidelines and steps for training objectives, trigger events, measures of performance, scenario generation, exercise conduct and control, data collection, and feedback. Provides a

Sources Fowlkes et al, 1998; Oser et al., 1999; Salas and Cannon-Bowers, 2000a

Can training for safe practices reduce the risk of organizational liability? meaningful framework that supplies the opportunity to embed learning events into scenarios. Assertiveness Practice and feedback help create and reinforce Individual training assertiveness in trainees. Provides opportunities for practice and supplies feedback. Metacognitive Training that develops and reinforces cognitive Individual training skills, such as inductive and deductive reasoning, problem solving. Stress-exposure Provides information regarding links between Individual training (set) stressors, trainee affect, and performance. and team Provides coping strategies for trainees in dealing with stressors. Team training Team training provides trainees with the Team necessary competencies at individual and team levels to complete their assigned tasks safely and effectively, by providing interventions that facilitate: (1) information presentation, (2) demonstration of teamwork behaviors and skills, (3) opportunities to practice, and (4) diagnostic feedback. Cross training Team members receive practice in performing Team other team members’ roles and tasks. Leads to a better understanding of other team members’ responsibilities and taskwork. Leads to enhanced shared mental models and interpositional knowledge. Improves team coordination and communication Team Team (explicit and implicit), encourages backup coordination behavior, and provides practice opportunities for training other KSAs that lead to effective coordination. Individual On-the-job Provides an opportunity to practice actual and team training required behaviors needed to do task. Targets team members’ procedurally based cognitive skills and psychomotor development. Training is provided in the same environment in which they will be working. Self-correction Helps individuals correct and evaluate their own Individual and team training behavior to assess its effectiveness. Team members learn to assess other team members. Allows constructive feedback and correction of discrepancies. Simulation-based Provides opportunities for trainees to operate in a Individual training and realistic setting with lifelike terrain, interaction, and team games and dynamic situations. Range in fidelity, immersion, and cost. Widely used in business, the military, and research. Behavior Targets concrete behaviors that are optimal for a Individual modeling particular task. and team

99

Smith-Jentsch et al. 1996 Jentsch, 1997

Driskell and Johnston, 1998

Salas and CannonBowers, 2000a; Salas et al., 1997

Salas et al, 1997; Volpe et al., 2001

Entin and Serfaty, 1999; Bowers et al., 1998 Goldstein, 1993; Ford et al., 1997

Smith-Jentsch et al., 1998; Blickensderfer et al., 1997 Tannenbaum and Yukl, 1992; Marks, 2000

Tannenbaum and Yukl, 1992


100

FIGURE 6.1 Components of scenariobased training. (Adapted from CannonBowers, J.A. et al, 1998, in J.A. Cannon-Bowers and E.Salas, Eds., Making Decisions under Stress: Implications for Individual and Team Training, Washington, D.C.: American Psychological Association, 365–374.) Another benefit of SET is that scenarios can be varied in that they require different responses from trainees. This in turn creates templates for trainees of what to expect and how to react to many different situations (Richman et al., 1995). It is argued that these templates will allow for rapid recall from memory and the ability to make decisions quickly (Klein, 1997), which may be necessary in environments in which errors can have catastrophic consequences if not rectified promptly. This should ultimately affect the transfer of the training to the actual environment (more discussion on this later). A final benefit of SET is that it is a broad instructional strategy within which many competencies can be framed. SET has been used extensively in the military, especially with aircrews in simulators (Smith– Jentsch et al., 1998) and others who are required to operate in complex decision-making environments. An example of a situation in which SET could be used appropriately to encourage safe practices is through assertiveness training. It has been found in past accidents (e.g., 1978, United Airlines Flight 173 in Portland, OR; 1982, Air Florida Flight 90 in Washington, D.C.) that some lower-ranking individuals did not assert themselves to higher-ranking individuals (e.g., copilot to pilot) when they became concerned about their situation. Many of these situations have resulted in disastrous consequences. If SET were used, events could be embedded in a training scenario that would require a low-ranking individual to assert himself to a high-ranking individual. This training would serve to train lower ranking individuals to voice their concerns and also to make higher ranking individuals aware (e.g., attitudes) of the consequences of not listening to the concerns of others, regardless of rank. Additionally, metacognitive training could be trained using SET strategies. This type of training


101

improves trainees’ cognitive skills so as to optimize the decision-making and judgment processes (Jentsch, 1997) necessary for exhibiting safe practices. SET could be used to embed a dynamic, ambiguous situation into a training practice scenario requiring trainees to recognize, react, and continuously reevaluate the situation and their decisions. Error Training A second type of training that might be used to improve safe behaviors is error training, the purpose of which is to use errors as an information function that provides feedback to trainees (Frese and Altman, 1989; Lorenzet et al., 2003). This feedback then serves to help develop improved learning and strategies to transfer the learned knowledge to the actual task environment. Trainees are able to experience errors and thus see the consequences of such actions (Karl et al., 1993). Trainees are then taught strategies to correct these errors and are able to practice them. Research suggests that individuals given error training performed better than those who were given error-free training (e.g., Dormann and Frese, 1994; Ivancic and Hesketh, 1995). Error training is, however, more complex than it may appear. Two components must be considered when developing an error training program: error occurrence and error correction (see Lorenzet et al., 2003, for more detail). Error occurrence can be (1) avoided (i.e., training is designed to prevent errors), (2) allowed (i.e., training is designed to allow errors to occur but they are not triggered; Gully et al., 2002), (3) induced (i.e., training is designed so as to evoke errors; Dormann and Frese, 1994), or (4) guided (i.e., training is designed intentionally to guide trainees to errors to encourage learning). The second component to error training, error correction, has two subcomponents: (1) selfcorrection (i.e., trainees work through errors alone; Frese et al., 1991), or (2) supported correction (i.e., trainees are offered support to help them correct errors; Carlson et al., 1992). Based on the training literature available and research conducted using error training, Lorenzet and co-workers suggest that guided error training may improve the skill development of trainees best (2003). A brief description of their work is next. In a recent study by Lorenzet and colleagues (2003), error training utilizing a guided error occurrence and a supported correction approach was compared to an error-free training approach to determine its impact on trainee skill development and self-efficacy. The results of this study suggest that trainees who were given guided-error training performed better (i.e., more accurately) and faster than those given error-free training. We submit that providing trainees with an opportunity to make errors during training so that they can learn corrective strategies and practice them may improve safe practices on the actual job. Therefore, improving safe practices in employees will reduce the likelihood of errors on the job, resulting in fewer accidents and less chance of legal liability. Stress Exposure Training (SET) Stress exposure training (SET) is an instructional strategy for training individuals that seeks to give the trainee the tools and ability to maintain effective performance in environments with high stress (Driskell and Johnston, 1998). Because stress increases the likelihood of error, SET training is particularly important for maintaining safe behaviors on the job. SET has three phases of training: information provision, skill acquisition, and application and practice.


102

• Phase 1—information provision: trainees are provided with information regarding the types of stressors that may be encountered and the effects of stressors on performance. • Phase 2—skill acquisition: provides trainees with the opportunity to acquire the specific behavioral and cognitive skills necessary to manage and adapt to stressors. • Phase 3—application and practice: the knowledge and skills learned in earlier phases are applied and practiced within an environment that gradually approximates a highstress situation. At the end of this phase, trainees receive feedback on their performance. Two recent studies were conducted (see Driskell et al., 2001) to determine if stress training would generalize (1) from exposure to a specific stressor in training to a novel stressor after training (study 1) and (2) from exposure to stress while performing a specific task during training to a novel task after training (study 2). The results of these studies suggest that stress exposure training was beneficial and did generalize from stressor to stressor and task to task. Although these results are encouraging, additional research is needed to determine whether stress training would generalize to stressors and tasks dissimilar to those who were trained. Team Training The fourth instructional strategy that we argue to be effective for training safe behaviors is team training. Teams are increasingly used by organizations to perform in complex, ambiguous environments in which human error can have catastrophic consequences (e.g., aviation), so the need to understand the research involving teams is important. Whereas asking a group of qualified individuals to work together as a team will likely be ineffective, team training provides trainees with the necessary competencies at the individual and team levels to complete their assigned tasks safely and effectively (Salas et al., 1997). At the individual level, team members must possess the KSAs needed to perform their specific task roles within the framework of the team, while incorporating the individual competencies into an interdependent, coordinated unit at the team level. Research conducted in the aviation community concluded that errors were more likely to be based on failures in team communication and coordination, rather than the individual crewmember’s inability to fly the airplane or perform his duties (Murphy, 1980, as cited in Helmreich and Foushee, 1996). Team training seeks to provide team members with interventions that facilitate: (1) information presentation, (2) demonstration of teamwork behaviors and skills, (3) opportunities to practice, and (4) diagnostic feedback (Salas and Cannon-Bowers, 2000a). Additionally, Kozlowski (1998) proposed three conditions that may be necessary to promote effective team training: (1) training that involves knowledge and skill acquisition (i.e., instructions that link related concepts and goals that must be acquired), (2) the promotion of transfer and adaptation, and (3) training that involves maintenance (e.g., monitoring) and enhancement of the KSAs that the team learned during training. Finally, through the use of “team” instructional strategies (see Table 6.1), training can encourage and facilitate specific team competencies such as shared knowledge structures (e.g., cross training; see Salas et al., 1997) and coordination and communication (e.g., team coordination training; see Entin and Serfaty, 1999). These two instructional strategies will be described next.


103

Examples of Team Training Cross training, most effectively used in team training, involves exposing trainees to goals, roles, tasks, and responsibilities of other employees within the organization (Salas and Cannon-Bowers, 2000a). There are two advantages to this type of training. First, this strategy involves trainees learning and practicing the tasks required of others’ jobs, allowing them to gain some degree of proficiency needed by the organization. This is extremely helpful when an employee is unable to perform his job adequately—a “crosstrained” employee could provide assistance. The second advantage of this training strategy is that it creates and reinforces a common understanding (i.e., shared mental model) of the roles and responsibilities of others. In situations such as an emergency, this shared understanding will reduce the amount of explicit communication needed because each team members will know what his task is as well as that of others. In short, crosstraining not only improves a trainee’s ability to predict what others are doing, but also allows him to step into that role if needed. However, in training these shared cognitive frameworks, it is necessary to ensure that they are not rigid and therefore are flexible to change. Another example of an instructional strategy focused at the team level is team coordination training, shown in Table 6.1. In order to perform effectively in situations of high stress, teams must switch from explicit to implicit coordination (Entin and Serfaty, 1999; Kleinmen and Serfaty, 1989). Team coordination training has been found to be necessary to make this switch (Serfaty et al., 1998). Entin and Serfaty (1999) examined the effectiveness of team adaptation and coordination training (TACT). The first step in TACT involves teaching team members to recognize changes due to varying stress levels in order to give them the tools needed to become aware of increasing stress. Next, team members are given a set of coordination strategies to use in stressful situations. Last, they are given information regarding the conditions under which these strategies should be applied. The theoretical basis supporting coordination training is that when teams face increased stress, teams that have received coordination training will minimize their communication and coordination overhead while maintaining the level of their performance. The communication used will likely be necessary and effective for a safer performance. On-the-Job Training One final type of training that we argue might help to improve safe behaviors in the actual work environment is on-the-job training (OJT). OJT is one of the most widely used types of training in organizations and can be used in conjunction with an off-the-job instructional strategy (such as those mentioned earlier) or as the sole strategy (Goldstein, 1993). In this instructional strategy, an expert or supervisor trains specific skills related to the job (e.g., what to do, how to do it) to an employee in the actual physical and social environment in which the task is to be performed (Sacks, 1994; Wehrenberg, 1987; De Jong and Versloot, 1999). Many benefits of OJT may serve to improve safe practices. For example, transfer of the learned skills to the job will be more likely because trainees are trained in the actual physical and social environment to which the skills must be transferred. In addition, trainees are able to practice the actual behaviors that they will be expected to perform


104

under the supervision of an expert. As such, trainees can practice the behaviors necessary to conduct the task in a safe manner, thus making it more likely that these safe behaviors will be used when the training is completed. It should be mentioned, however, that often OJT programs are not designed and developed properly and are not as successful as they could be (Goldstein, 1993). Therefore, like any training program being implemented, sound principles of training design should be followed to ensure its effectiveness. We will next briefly discuss two types of OJT programs that might improve safe practices in organizations: apprenticeship training and mentoring. Examples of OJT When a task is performed, thinking about, knowing, and understanding the task are very important (Hendricks, 2001). One way to accomplish this is through apprenticeship training. Although apprenticeship training has most often been used to train trade skills, other nontraditional trade organizations are beginning to show an interest in it (Goldstein, 1993). A typical apprenticeship training program consists of classroom-based instruction followed by supervision on the job from an experienced employee. Following a predetermined time and if the necessary skills are met, the apprentice is promoted to a “journeyman” (Goldstein, 1993; Hendricks, 2001; Lewis, 1998) and is able to perform the duties on his own. The importance of this type of training for safety is that the trainee is provided first with the basic KSAs needed for the job (i.e., in the classroom) followed by the opportunity to practice the learned skills on the job. If safe behaviors are not demonstrated, the experienced supervisor can correct the behavior before it becomes a habit. In addition, a new employee will not be allowed to perform a task alone until the supervisor is certain that he can perform it safely. A second type of OJT, which is very similar to apprenticeship training, is mentoring— the process of creating a relationship between a less experienced individual and a more experienced one (Wilson and Johnson, 2001). As a part of this relationship, the more experienced worker will train and develop the trainee to perform his job properly. Research suggests that mentoring improves communication, job satisfaction, and success in an organization, which will likely influence safety (Mobley et al., 1994; Forret et al., 1996). Additionally, this type of training might benefit safety in that the trainee is guided by the mentor as to which behaviors are appropriate and which are not (Scandura et al., 1996). Therefore, it is important that the mentor be an individual who values and respects the cultures (e.g., safety), policies, and procedures within the organization and will encourage the trainee to do so also. Finally, mentoring may be beneficial not only to the trainee but also to the mentor by improving skills that may have diminished over time (Forret et al., 1996). Evaluating Training Once the instructional strategy has been chosen and the training program has been implemented, it is imperative that the training be evaluated. Few organizations conduct systematic evaluations of their training programs. We acknowledge that evaluation can be resource intensive; however, it is the only way to truly assess training’s effectiveness. Numerous methods of assessment can be applied to training. We argue for the use of a multilevel approach such as that suggested by Kirkpatrick (1976).


105

Kirkpatrick (1976) proposed a method of training evaluation that constituted a multilevel approach to evaluating the outcomes of training programs. He argued that training evaluation should include assessment at four levels: (1) reactions (i.e., what trainees think of the training), (2) learning (i.e., what trainees learned), (3) behavior (i.e., how trainees’ behavior changes), and (4) results (i.e., impact on the organization). Building on Kirkpatrick’s framework, Kraiger and colleagues (1993) outlined three similar outcomes: (1) affective (i.e., reactions), (2) cognitive (i.e., learning), and (3) skillbased (i.e., behavior) outcomes. The first level, reactions, evaluates how well trainees liked the program (i.e., affect), as well as how useful they think it is (i.e., utility). Similarly, Kraiger et al. referred to affective outcomes as attitudinal outcomes and motivational states (i.e., motivational disposition, self-efficacy, and goal setting) that training has produced.

TABLE 6.2 Kirkpatrick’s Multilevel Evaluation (1976) Step

Guidelines/requirements

Measurement

1. Determine what you want to know. Use a Self-report Reaction written survey to get at what you want to know. survey The form should be quantifiable. The forms should be anonymous to get honest reactions. Additional written comments by trainees should be allowed. 2. Each trainee should be measured. Pre- and post- Self-report Learning test evaluations should be used. Measurement survey should be as objective as possible. A control group should be used, if possible. Measurement of learning should be analyzed statistically.

3. Trainees must want to improve. Trainees must Observation Behavior be aware of their own weaknesses. The work environment must be permissive. Someone who is skilled and interested must provide help. An opportunity to try out ideas must be provided. 4. Although results may appear to be Longitudinal Results straightforward, it is difficult to determine how data much is due to training.

Example survey questions How well did the trainer state the objectives? How helpful and friendly was the trainer? Did you like the training? T/F. Well trained employees are a reflection of a large training department. T/F. The supervisor is closer to his employees than management is. Is employee more observant? Is employee more patient? Does employee demonstrate learned behaviors? What changes have been observed since training in the following areas: quantity of production; safety; employee attitudes; employee turnover?

The next level, learning, evaluates the principles, skills, and knowledge that the trainees gained from the training. This level seeks to determine if the trainees learned the KSAs presented in the training (i.e., training validity) and not whether the trainees exhibit the trained behaviors. The third level of evaluation, behavioral, determines the changes in behavior of employees after the training intervention. The behavioral level evaluates how


106

the learned KSAs transfer to the actual work environment (i.e., transfer validity). Finally, the last and highest level, results, evaluates the organizational outcomes of the training program (e.g., improved safety, reduced costs, improved quality). Evaluation at this level can help determine intraorganizational validity (i.e., are the performances of multiple groups of trainees consistent?) and interorganizational validity (i.e., will the training program in one organization or department be effective in another?). Unfortunately, due to the necessity of and difficulty in collecting longitudinal data, this level is not often evaluated. See Table 6.2 for more information on utilizing Kirkpatrick’s (1976) multilevel approach (e.g., guidelines, measurement). This multilevel approach has been shown to be effective and is still used in a number of studies (Cohen and Ledford, 1994; Field, 1995). For example, Salas and colleagues (2001b) examined the success of crew resource management (CRM) training in aviation using this approach by examining the literature available (58 studies total). Although 41% of the studies collected information at multiple levels of Kirkpatrick’s framework, a majority of those only collected information pertaining to two of the levels—usually, reaction/learning or reaction/behavior. Although research generally suggests that CRM has been successful in improving safety in the aviation community, the results are still unclear. As such, the results of this review further emphasize the importance of organizations evaluating their training programs at all levels. However, regardless of an organization’s goal of training (for our purposes, improving safe behaviors), any multilevel evaluation should revolve around two central issues: (1) establishing

TABLE 6.3 Augmented Kirkpatrick Training Taxonomy Augmented framework 1. Reactions

2. Learning

3. Transfer

Components

Definitions

Affective Emotionally based opinions given immediately with little, if any, reactions forethought Utility judgments Trainees’ opinions about the transferability and utility of training; more behaviorally evaluative Immediate Trainees required to indicate how much they know about what knowledge they were trained in (i.e., multiple-choice tests, open-ended questions, listing of facts) Knowledge Essentially the same tests as immediate knowledge, but retention administered much later; can be used instead of or in combination with immediate knowledge Behavior/skill Indicators of behavioral proficiency exhibited within training, as demonstration opposed to on the job; measures include simulations, behavioral role-plays, behavioral reproduction, scores in performancecentered classes, and ratings of training performance Includes measures that indicate on-the-job performance taken some time after training that seeks to assess a measurable aspect of job performance; measures on-the-job performance, outputs, outcomes, and work samples


107

4. Organizational impact of training assessed, typically by measuring productivity gains, Results customer satisfaction, any change in cost, an improvement in employee morale, and profit margin; results often difficult to measure because of organizational limitations and because they are the most distal from training; results often regarded as the basis for judging training success, but judgments are often based on false expectations Source: Alliger, G.M. et al., 1997, Personnel Psychol., 50, 341–358.

measures of success and (2) using experimental and nonexperimental methods to determine what has changed due to training (Goldstein, 1993). Although Kirkpatrick’s approach was a breakthrough for training evaluation and is still widely used (see Cohen and Ledford, 1994; Field, 1995), the original multilevel framework has also been revised or modified based on more recent research (Alliger et al., 1997; Kraiger et al., 1993). Alliger and colleagues (1997) reviewed Kirkpatrick’s approach through meta-analysis of 34 articles used in an earlier study (Alliger and Januck, 1989). The authors suggested an augmented framework, shown in Table 6.3, based on their analysis of the results. Their framework further clarified Kirkpatrick’s original method of evaluation by classifying different types of reactions and learning and by focusing on transfer of training to the work environment, instead of the more vague behavior. A significant problem with training evaluation is that it is often conducted using selfreport measures of trainees’ reactions to the training. Although these measures are an easy method of evaluation, they fail to provide evaluators with valuable information regarding the training’s effectiveness. Whether a trainee “likes” the training intervention generally does not reflect how well the training was conducted or if the training addressed the appropriate competencies. Ultimately, researchers found that “liking does not equate to learning or to performing” (Alliger et al., 1997, p. 344). In part to combat this fallacy, another dimension of reactions, similar to that suggested by Kraiger and colleagues (1993), was added to the framework: utility judgments, which refer to trainees’ opinions of whether what they learned will help them in their jobs. Affective reactions continue to measure how much the trainees liked the training. As noted earlier, liking does not translate into desired outcomes, but the trainees’ feelings toward training can tell researchers about organizational factors (i.e., organizational support of training programs). Alliger and colleagues (1997) divided the learning phase into three categories: • Immediate knowledge involves trainees indicating how much they know about the training content (i.e., what they were taught during training). Immediate knowledge assessment, under Alliger and colleagues’ (1997) framework, consists of multiple methods of evaluation (i.e., multiple-choice tests, open-ended questions, listing of facts). • Knowledge retention involves the same types of tests given in immediate knowledge assessment, except that they are given after more time has elapsed between training and testing. Knowledge retention can be used alone or in combination with immediate knowledge assessment. • Behavior/skill demonstration involves the trainees’ performances within training, as opposed to their performances on the job. Measures of behavioral and skill


108

demonstration include simulations, behavioral role plays, and ratings of training performance by SMEs. All three components are designed to evaluate what trainees learn. When combined with a trainee’s reactions, they strengthen the reactions’ predictive power of performance, but even alone, learning assessment is a strong indicator of performance (Alliger et al., 1997). The last two steps in the augmented framework are transfer and results. Transfer measures seek to evaluate on-the-job performance and are taken some time after training. Transfer differs from knowledge retention because transfer assessment is more behaviorally based. Transfer measures focus on on-the-job performance, outputs, outcomes, and work samples. Finally, results refer to the “bottom line” variables or what management typically assesses for training outcomes. Results include measurement of productivity gains, customer satisfaction, changes in cost of production, employee morale, and profits. These variables can be difficult to measure and may be misleading because they are not always directly related to training. Additionally, problems are often encountered when evaluating results due to falsified or inflated expectations of training’s impact on organizational results by management. Researchers recommend caution when dealing with result variables. We will next present a tool that can be utilized to evaluate performance following training. Line Operations Safety Audit (LOSA) One tool that can be adapted to help evaluate an organization’s training program is based on the line operational safety audit (LOSA) developed for the aviation industry. LOSA was developed by Helmreich in 1995 as a way to evaluate how CRM was working in the cockpit (Croft, 2001). This audit provides training and performance evaluation with a framework for collecting data. Although used primarily with flight crews, the framework can be applied to evaluate the performance of trainees in a variety of settings. Traditionally, LOSA uses nonintrusive, trained observers who ride in the cockpit jumpseat and examine how pilots respond to threats (i.e., weather, airway congestion, or unexpected events resulting from errors). The observers complete a questionnaire regarding the crew’s performance during a flight and assign one of four ratings: (1) poor (observed performance had safety implications), (2) marginal (observed performance was barely adequate), (3) good (observed performance was effective), or (4) outstanding (observed performance was truly noteworthy). The observers also interview the pilots to gain more insight into what happened and what their thought process was in the course of making decisions. One of the most important features of LOSA is that, in addition to recording errors, it evaluates how crews deal with or adapt to the consequences of the errors. See Checklist 6.4 for an example of a training evaluation form adapted from the LOSA program. The implementation of LOSA ideally results in the discovery of weaknesses or critical incidents that may exist within the training structure that, in turn, can be incorporated into future training programs. Croft (2001) provides an example of LOSA’s success. Pilots at a major airline were found to have difficulty flying according to company standards. The LOSA program found that, during training, pilots were provided with ambiguous guidelines as to how to fly within company standards; this led to their difficulties. The


109

error was corrected for the future training interventions and the pilots showed a 59% improvement in that area of performance. External Factors Until this point, we have provided a brief methodology by which to develop and evaluate an effective program for training safe behaviors systematically. Although this may ensure a training program that produces desired learning outcomes, it may not guarantee exclusion from legal liability. Factors outside training also need to be considered because they will impact the program’s outcomes over and above the content and strategies used (see Table 6.4). A few of these—specifically, the pretraining environment, organizational and individual characteristics, training motivation, and post-training environment—will be briefly described next.

TABLE 6.4 Components of Training Effectiveness

Source: Adapted from Salas, E. et al., 1999, Personnel Hum. Resour. Manage., 17, 123–161.


110

Pretraining Environment An additional key to training’s success is the pretraining environment, of which two main characteristics need to be considered: (1) prepractice conditions and (2) pretraining climate (Salas and Cannon-Bowers, 2001). Prepractice conditions are elements in the pretraining environment that help prepare trainees for practice during training. Research suggests that practice is more than mere repetition of a task; rather, it is a complex process that leads to skill acquisition (Ehrenstein et al., 1997; Shute and Gawlick, 1995). Cannon-Bowers and colleagues (1998) suggest that trainees can be prepared by using such interventions as preparatory information and advance organizers; however, this is just beginning to be investigated empirically. The second characteristic of the pretraining environment to be considered is the pretraining climate. Characteristics of the climate that will influence the outcomes of training include framing of the training, attendance policies, and previous training experience. More specifically, research has shown that whether the training is framed as remedial or advanced will influence trainees’ motivation and learning (see Quinones, 1995, 1997). In addition, the attendance policy of the training program (i.e., voluntary vs. mandatory) is also believed to influence the outcomes of training (Baldwin and Magjuka, 1997). This research suggests that the ability of trainees to choose whether or not to attend training will influence the training outcomes. Finally, it is suggested that trainees with previous training experience, whether positive or negative, will influence the trainees’ learning and retention (see Smith-Jentsch et al., 1996). Organizational Characteristics In addition to characteristics of the pretraining environment, organizational characteristics (i.e., those present within the organization to which the newly acquired KSAs must be performed) may also influence the outcomes of training. Examples of these characteristics are selection and notification, situational constraints (e.g., improper equipment), organizational climate (e.g., perceived organizational support, safety culture, and policies), and opportunities to practice (see Salas et al., 1995, for more detail). We do, however, want to expand on two organizational characteristics that greatly influence the transfer of safe behaviors from training to the actual work environment—safety culture and policies. Safety Culture An important factor influencing training within an organization is its culture. The necessity of a positive organizational culture has been argued to be of importance for the effectiveness and transfer of a training program (Rouiller and Goldstein, 1993). More importantly—especially when designing training for safe behaviors—is the organization’s safety culture. A safety culture can be defined as the end result of the values, beliefs, attitudes, behaviors, perceptions, and competencies of individuals and groups within an organization that determine the proficiency and style of, and commitment to, the organization’s safety management (HSC, 1993, as cited in Glendon, 2001). Pidgeon (1991) and others (Turner, 1991; Pigeon and O’Leary, 1994) suggest that a positive safety culture is encouraged through four factors: (1) commitment to safety from upper-level management, (2) shared attitudes regarding hazards and safety, (3)


111

flexible norms and rules to deal with hazardous situations, and (4) practice and organizational learning. Others also argue that a positive safety culture is characterized by effective communications, mutual trust, and efficacy (Glendon, 2001). Hofmann and colleagues (1995) addressed a related issue. Specifically, the authors emphasized the need for management in organizations to make safety an observable issue. Management must, as they say, “talk the talk and walk the walk” when safety is concerned. In order for the four factors (presented here) to translate into a positive safety culture (see Pidgeon, 1991; Pidgeon and O’Leary, 1994; Turner, 1991), it is argued that the organization, through management, must promote them through actions and words. In order to accomplish this, some researchers recommended two guidelines gleaned from research with coal miners and their managers (Fiedler et al., 1984). First, in order to promote a positive safety climate, employees must perceive management as devoting time, energy, and attention to safety issues and training interventions. This will reinforce the support of the organization and the importance of safety within the work environment. Second, it is argued that positive safety cultures develop when employees are allowed to take an active role in the safety training intervention. By giving workers a degree of control over the interventions, their investment in and dedication to the maintenance of a positive safety culture will increase. In sum, research suggests that organizational support and encouragement of employee involvement are necessary for success of a safety culture. Safety Policies and Procedures Underlying the safety culture philosophy within an organization are its safety policies and procedures. Policies can be described as broad requirements that management has set to let employees know about what is expected of them (Degani and Wiener, 1997). For example, a safety policy may encourage employees to demonstrate safe behaviors and procedures and then provide guidance to employees on how to meet these expectations. Thus, in order for employees to demonstrate safe behaviors, management may require that checklists be used for certain tasks or following standard operating procedures. Trainees can be made aware of these policies and procedures through a well-designed training program. According to Hofmann and Stetzer (1996), however, a problem develops when social pressures are more influential than the formalized rules and procedures. These authors argue that, regardless of an organization’s established safety policies and procedures, if employees perceive that deviations from them are appropriate and encouraged (i.e., normalization of deviance), the rules and procedures will be less influential. For example, if the organizational climate promotes faster production at all costs, the rules promoting safety will likely be ignored. Wright (as cited in Hofmann et al, 1995) observed this phenomenon in his examination of workers on oil rigs, which are expensive to build, are expensive to maintain, and depend heavily on interdependence. In addition, this is a dangerous job in which small errors could lead to significant consequences. Due to the expense and resources used, management often focuses on increasing production times at the cost of sacrificing worker safety. Wright’s research uncovered an underlying theme for a majority of the accidents examined (which were in the thousands): an attitude of “get the job done, while placing less emphasis on safety issues” (Hofmann et al., 1995, p. 134). Therefore, safety rules and procedures must be developed and promoted with


112

consideration of the powerful effect of the organizational climate on adherence to policies and procedures by employees. Error Reporting Some argue that underlying organizational issues (i.e., culture) influence errors and even implicitly encourage and support individuals to commit unsafe acts (Hoffman et al., 1995). Error reporting is one tool that can help organizations to organize and learn from these errors. Reason (1998) argues that error reporting is an essential element of a safety culture because it allows the organization to be informed about unsafe practices in the workplace. Therefore, a positive safety culture that encourages safe behaviors, adheres to safety policies and procedures, and encourages error reporting will likely reduce human error. Although many of the human errors that occur do not result in an accident, numerous “near misses” (i.e., incidents) do result. As a part of creating a positive safety culture, organizations need to realize that great lessons can be learned from these incidents (Klair, 2000) and should encourage employees to report them. By taking a proactive approach to promoting safe behaviors, organizations may be able to correct these errors (e.g., through an effective training program) and prevent an accident from occurring in the future (Helmreich and Merritt, 2000). Error reporting can be organized in two ways: (1) critical incidents technique and (2) voluntary reporting systems. Critical incidents technique. This technique, now often used as a job analysis tool (i.e., a trained analyst interviews workers about their work duties and responsibilities), began as a technique to gather information to help develop training needs and performance appraisal (Gatewood and Field, 1998). The critical incidents technique involves the observation of good and bad examples of behavior performed on the job. In essence, it consists of a description of the task, the behaviors observed, and the consequences of those behaviors. Although it does not illustrate every possible dimension of a job, the report provides a range of behaviors providing examples of what not to do, what should be done, and what results either elicited. For an example of the critical incidents technique, see Prince and Salas (1993). Voluntary reporting systems. Most individuals are not likely to report errors due to fear of retribution; thus, anonymous, voluntary reporting systems are a viable alternative. The aviation safety reporting system (ASRS) was developed by the National Aeronautics and Space Administration (NASA) for the aviation industry. Using the ASRS database, pilots and/or crewmembers are able to report errors and unsafe acts that occurred during the flight without providing discernible information about themselves. The aviation industry has been extremely successful since its inception and receives more than 32,000 reports each year (Orlady and Orlady, 1999). The data collected from ASRS have allowed the aviation community to react to errors proactively (Sexton et al., 2000) by incorporating critical incidents that occur frequently into training. The data not only are useful for training purposes but also provide an awareness to other aviation professions via publication in periodicals and on the Internet. The success of ASRS has led to the development of similar systems in other organizations, for example, the medical, nuclear, and petrochemical domains (see Köhn et al., 1999; Helmreich, 2000). Köhn and colleagues (1999) offer some guidance for developing an error-reporting system. First, it is suggested that the system be overseen (if possible) by an organization


113

separate from that at which those reporting the errors work. This will likely improve the participation rate by employees because they will have less fear of retribution. It is understandable that organizations may not have the resources to work with another organization to oversee their reporting system. We advocate that an error-reporting system can be a valuable tool for obtaining information regarding safe behaviors, and therefore encourage organizations to implement the system regardless of which organization may oversee it. Next, employees need to be confident that the information they provide to the system will remain anonymous and that they will not be punished. If employees fear retribution from the organization for unsafe behaviors, they will be less likely to report an error that went “undetected” by the organization. Finally, it is suggested that the data be accessible so that it can be analyzed. Based on the data, critical incidents can be identified and incorporated into the organization’s training programs. Individual Trainee Characteristics Training outcomes are also influenced by characteristics beyond the organization. Characteristics that the trainee brings to the training program will influence its outcomes and may do so before it begins or after it is completed (i.e., transfer). Individual characteristics reported to influence the outcomes of training are • Cognitive abilities (i.e., “g” or general intelligence). Cognitive ability has been shown to influence trainees’ attainment of knowledge about the job (see Ree et al., 1995; Colquitt et al., 2000). In addition, cognitive ability has been shown to be a strong determinant of success in training (Ree and Earles, 1991). • Self-efficacy. Research that has looked at self-efficacy (i.e., belief in one’s own ability) indicates that high self-efficacy leads to better performance (see Quinones, 1995; Ford et al., 1997; Martocchio and Webster, 1992). In addition, it is suggested that selfefficacy is influenced by cognitive ability (see Hunter, 1986). • Expectations. Trainees’ expectations regarding training will influence training outcomes. In a study conducted by Tannenbaum and associates (1991), it was found that trainees’ whose expectations were met demonstrated better post-training commitment, greater motivation, and higher self-efficacy. • Goal orientation (i.e., mastery vs. performance). Goal orientation will influence training outcomes (Dweck, 1986; Dweck and Leggett, 1988). Individuals high in mastery orientation aim to acquire new skills and master novel situations, and research suggests that this is positively related to metacognitive activity and self-efficacy (Ford et al., 1997; Phillips and Gully, 1997). On the other hand, those high in performance orientation aim to achieve high performance ratings and to avoid negative ones. See Salas and colleagues (1995, 2001b) for more detail as to how these factors may influence training. Trainee Motivation Trainee motivation is influenced by individual (e.g., self-efficacy) and organizational (e.g., notification) characteristics. It consists of variables that influence trainees’ willingness to participate in and learn from the training. As with individual trainee characteristics, motivation may also influence the outcomes of training prior to or after


114

training. Training motivation influences the amount of time and effort the trainee will invest and the behaviors the trainee will exhibit (Naylor et al., 1980, as cited in Goldstein, 1993). Research suggests that higher trainee motivation prior to training results in greater learning and positive training reactions (see Mathieu et al., 1992; Williams et al., 1991; Tannenbaum et al., 1991; Baldwin et al., 1991). Finally, it is suggested that training is more effective when trainees believe that the outcomes of the training are relevant to their job performance (Noe, 1986). For example, if the purpose of training is to promote safe behaviors, the trainee must believe that the behaviors will be applicable to and improve his performance on the job. Recently, Colquitt and colleagues (2000) conducted a meta-analysis of the literature related to training motivation. The results of their analysis suggest that individual and situational characteristics influence motivation to learn in training. Specifically, the individual characteristics related to motivation included self-efficacy, valence, anxiety, cognitive ability, and age. Situational characteristics included supervisor and peer support, positive climate, and organizational support. Although it was previously believed that cognitive ability alone influenced learning outcomes, the results of this meta-analysis suggest that training motivation explained additional variance beyond just cognitive ability. These results indicate the importance of including multiple individual differences when conducting the needs assessment because they will likely influence training outcomes. Post-Training Environment The post-training environment is important in determining whether the competencies learned during training will transfer to the job. Regardless of how well the training program was developed, without an environment that encourages the transfer of the learned competencies, it will not be effective. Research suggests that several characteristics of the work environment are essential for the transfer of training: (1) supervisor support, (2) organizational transfer climate, and (3) continuous-learning culture, although supporting empirical evidence is limited (see Baldwin and Ford, 1988; Rouiller and Goldstein, 1993; Tracy et al., 1995; Ford and Weissbein, 1997). Additionally, it is argued that some elements of the transfer climate may facilitate (e.g., rewards, positive transfer climate) or hinder (e.g., lack of peer or supervisor support, lack of resources) the transfer of training (Tannenbaum and Yukl, 1992; Rouiller and Goldstein, 1993). We will briefly discuss how supervisor support and organizational transfer climate can promote the transfer of training. Supervisor support has been argued to influence transfer of training. Research conducted suggests that discussions with supervisors prior to and following training, as well as supervisor sponsorship, result in transfer of learned skills to the job (Huczynski and Lewis, 1980; Brinkerhoff and Montesino, 1995). Additional research concluded that opportunities to perform learned skills provided by supervisors encourage transfer of training (Ford et al., 1991). However, Baldwin and Ford (1988) argue that there is a lack of understanding regarding the behaviors that lead trainees to perceive support. Tannenbaum and Yukl (1992) state that supervisor support could include goal-setting activities (e.g., minimize number of accidents), reinforcement (e.g., error reporting), and modeling of trained behaviors (e.g., safe behaviors). Although these results are


115

encouraging, additional research is needed to determine the true impact of supervisor support on transfer of training. The research evidence supporting the role of the organizational transfer climate on transfer of training is slightly larger than that for supervisor support. Organizational climate can be defined as the interaction between observable elements within the organizational setting as well as those that are perceived by trainees (Hellreigel and Slocum, 1974; James and Jones, 1974). Research suggests that trainees who perceive a positive organizational climate (e.g., organizational support, rewards, safety policies, nonpunitive error-reporting systems) apply learned competencies on the job (Baumgartel et al., 1984; Rouiller and Goldstein, 1993; Tracy et al., 1995). In addition, we argue that a positive safety culture (a part of the organizational climate) that supports the demonstration of safe behaviors by trainees will encourage the transfer of these behaviors back to the work environment (see subsection on External Factors). Finally, a continuous-learning culture (i.e., work environment) is important for the transfer of learned competencies to the job. A continuous-learning work environment is one in which the acquisition of knowledge and skills is supported and opportunities are provided; achievement is reinforced; and innovation and competition are encouraged (Dubin, 1990; Rosow and Zager, 1988; Tracy et al., 1995). Members of organizations with these characteristics understand that learning is a part of their daily work environment. Research conducted by Tracy and colleagues (1995) suggests that trainees who perceived a continuous-learning environment demonstrated more post-training behaviors. In organizations in which errors are likely to occur, we argue that a continuous-learning environment may be necessary to encourage employees continuously to demonstrate safe behaviors learned during training and to learn from the errors that may occur. 6.4 Examples A number of liability cases have addressed training and the repercussions of inadequate training (i.e., resulting in unsafe behaviors) in deciding legal culpability. Going as far back as 1978, the courts have made landmark decisions regarding the responsibility of an organization or municipality in providing adequate training. Cases that have come before the courts cover numerous training issues, including the use of deadly force (Gilligan v. Morgan, 413 U.S. 1, 1973), training for high-stress situations such as a car chase (Springfield v. Kibbe, 480 U.S. 257, 1987), training to ensure safety for employees (Collins v. Harker Heights, 503 U.S. 115, 1992), and the need for single-sex environments with certain types of training (United States v. Virginia et al., 518 U.S. 515, 1996). One notable case was that of Monell v. Department of Social Services of the City of New York, 436 U.S. 658 (1978). As a result of this case, the Supreme Court justices upheld the decision that municipality officials could be held legally responsible for policy making, which has proved to be unfavorable for some organizations.


116

Law Enforcement In Oklahoma City v. Tuttle, 471 U.S. 808 (1985), an innocent man was shot and killed by a police officer outside a bar where a robbery in progress had been reported. The city was held liable in the original trial due to negligence in training of the officer. The precedent had been set by Monell v. New York City Department of Social Services, and the court found that the training enforced through city policy had resulted in the inappropriate behavior of the officer and, ultimately, in one man’s death. Upon appeal, the court upheld the decision and indicated that even a single incident can represent negligence on the part of the policy makers (i.e., those making policies regarding training). Although the Supreme Court eventually overturned the decision, it was due to their disagreement with the lower courts that a single incident could represent negligence. The justices agreed with the culpability of the city in providing appropriate training. Industrial In 2002, a construction company was cited and ordered to pay more than $100,000 when a worker was electrocuted and killed while on the job (OSHA, 2002). The company was cited for improper procedures, inadequate training of workers, failure to conduct safety briefings, and improper equipment. In a similar instance, a trucking company was ordered to pay more than $70,000 when a worker was killed in an explosion. The company was cited for improper training, failure to follow safety precautions, improper equipment (e.g., lack of fire extinguishers), and failure to inspect the work environment. Aviation Beyond the examples presented earlier, the aviation industry has generally been successful in combating unsafe practices through the use of team training, specifically crew resource management (CRM) training (Helmreich and Merritt, 2000). It is argued that this industry is the number one advocate of team training (Helmreich et al., 1993). CRM was developed as a means to train aviation crewmembers to utilize all resources available to them (e.g., people, information, equipment) through coordination and communication (Wiener et al., 1993; Salas et al., 2001a). Although CRM’s focus has changed over the past two decades (i.e., from changing interpersonal style to assertiveness training for junior crewmembers), its current focus is on error management. Since the inception of CRM training, the aviation safety record has improved greatly. The risk of death when traveling by commercial airliners has dramatically decreased from 1 in 2 million to 1 in 8 million over the past two decades as a result of CRM training (Kohn et al., 1999). The most notable success of CRM training was demonstrated during the approach to landing of United Airlines Flight 232, which had suffered a catastrophic engine and hydraulic failure (Roberts and Bea, 2001; www.airdisaster.com). Crewmembers worked together as a team to maneuver the nearly uncontrollable aircraft to a crash landing in Sioux City, Iowa. Approximately 200 of the 300 passengers and crew on that flight survived as a result of the crew’s teamwork and CRM training. The overwhelming


117

success of CRM training in aviation has led to its adaptation and implementation in many other organizations (e.g., medical, oil, nuclear power; see Salas et al., 200la). Medical For many years, the medical industry has ignored the role that human error has played in the field (Pietro et al., 2000), in part due to the risk of legal actions that could be brought against not only individual medical professionals but also medical organizations. It has been estimated that human error leads to approximately 100,000 patient deaths and many more injuries in U.S. hospitals each year and results in costs of up to $9 billion annually (Barach and Small, 2000). At least 50% of these errors are not reported and 70% of those reported were deemed preventable (Leape, 1994). However, medical errors are often justified and rationalized in this domain due to the complex and subjective nature of medicine (Pietro et al., 2000). To further complicate the issue, medical organizations and their personnel are not obligated to report errors. Reasons provided for not disclosing or further investigating errors are risk of negative publicity and legal actions, high costs, lack of standards to define an unacceptable error, and lack of justification to conduct such an investigation. The medical domain’s resistance to accepting human error makes it difficult to train for safe practices; however, there is some good news. Recently, the medical domain began taking advantage of the lessons learned from the aviation industry and implemented CRM training to medical teams (e.g., surgical; Helmreich, 2000). Specifically, surgical and anesthetic teams have been provided with CRM training to improve communication within the operating room (Helmreich, 1997). The training programs embed critical scenarios using a “dummy patient” that require the team members to make decisions together to perform the operation successfully. Following the training, the teams are provided feedback on their performance. Helmreich suggests that other medical domains (e.g., emergency rooms, ambulance teams) could also benefit from this training, as well as other complex, dynamic environments. 6.5 Conclusion Human error is unavoidable, and training employees to trap and mitigate the consequences of these errors and unsafe behaviors is a welcome start in the right direction. However, the threat of organizational liability for improper training resulting in an accident or incident exists. The purpose of this chapter was to provide an understanding of issues that influence safe practices in organizations as well as how training can be utilized to reduce errors and improve safe behaviors. We also offered guidance regarding factors to consider during the design, delivery, and evaluation/transfer of training. In this chapter, we have argued that training must be designed and delivered systematically by taking into consideration key factors such as the pretraining environment, organizational and trainee characteristics, trainee motivation, and the posttraining environment. We emphasized the importance for organizations to understand that training is more than just a program. The human factors field, as well as the


118

industrial/organizational psychology field, has helped to develop a science behind it that needs to be exploited to better ensure its intended outcomes. Therefore, to minimize the risk of legal liability, organizations need to rely on the science of training when developing a training program. By incorporating sound learning principles; conducting systematic, multilevel evaluations; and ensuring the transfer of trained skills to the job, organizations can appropriately determine that the desired outcomes were produced and training’s overall effectiveness was achieved. Until then, inadequate training will lead to unsafe practices and the threat of legal liability will always be on the horizon. Can training for safe practices reduce the risk of organizational liability? Yes, when human factors principles are incorporated into the design, delivery, implementation, and evaluation of training as is suggested in this chapter. Although there is no guarantee that a training program that does not utilize human factors principles will lead to legal liability, the risks involved and potential consequences do not make this a chance worth taking.

CHECKLIST 6.1 Steps to Develop an Effective Training Program Systematically Primary factors

Considerations

1. Conduct training needs analysis Provides information used to develop instructional objectives and training criteria • Where is training needed? a. Organizational analysis • When is training needed? b. Task analysis (cognitive task • Task characteristics analysis—see Checklist 6.2) • Task competencies (KSAs) • Who needs to be trained? c. Person analysis • In what do they need to be trained? 2. Consider external factors Determine impact of organizational (e.g., safety culture), individual (e.g., self-efficacy), and motivation characteristics on training a. Organizational characteristics • Selection and notification • Situational constraints (e.g., improper equipment) • Organizational culture (e.g., perceived organizational support, safety culture, and policies) • Organizational resources and opportunities to practice b. Safety culture • Safety policies • Error reporting system c. Individual characteristics • Cognitive ability • Self-efficacy • Goal orientation d. Motivation • Direction • Effort • Intensity • Persistence 3. Establish training objectives (see Are objectives: Checklist 6.3) • Measurable? • Specific?


4. Determine the methods to use

5. Determine the strategies to use

6. Develop scenarios for training

Primary factors 7. Evaluate training

8. Ensure transfer of training

119

• Relevant to task? • Information based • Demonstration based • Practice based • Combination of above methods • Event-based training • Assertiveness training • Metacognitive training • Stress-exposure training • Cross training • Represent critical incidents? • Represent actual task environment?

Considerations Questions to answer: • Was training effective? • Were training objectives accomplished? Methods to use: • Multilevel evaluations (see Table 6.2 and Table 6.3) • Experimental methods • Nonexperimental methods • Supervisor support • Organizational climate • Continuous-learning culture

CHECKLIST 6.2 Development and Implementation of Cognitive Task Analysis Primary factors Select experts

Develop scenarios based on task analysis Choose a knowledge elicitation method: • Interviews • Verbal protocols • Observations • Conceptual methods Implement knowledge elicitation method chosen: • Interviews

Considerations • Experience with domain • Experience with task • Number to include • Develop relevant task scenario(s) with problem statement • Pretest scenarios to determine completeness • What information are you trying to elicit? • Ask experts about cognitive processes needed to complete task scenarios • Ask experts to think aloud while completing task scenarios • Observe experts completing task scenarios • Inferences made based on experts’ relatedness judgments • Decide the number of sessions to be recorded • Obtain consent and provide task scenario to expert • Ask experts questions, for example: – What rules/strategies used to complete scenario?


• Verbal protocols

• Observations

• Conceptual methods

Organize and analyze data

120

– What knowledge is needed to complete scenario? – What cognitive skills are needed to complete scenario? – If this happens, then what would you do? • Record comments provided by experts via video, audio, pen and paper • Repeat each scenario for each expert • Ask expert to “think aloud” while working through scenario • Record comments provided by experts via video, audio, pen and paper • Repeat each scenario for each expert • Be as unobtrusive as possible • Ask for clarification as necessary • Record comments via video, pen and paper • Repeat each scenario for each expert • Present pairs of tasks to experts • Ask experts to judge how related the tasks are • Present each pair of tasks to each expert • Record similarities between experts • Interview additional experts if questions arise from data • Identify rules and strategies applied to the scenario • Identify knowledge required • Identify cognitive skills required • Generate list of task requirements • Verify list with additional experts

CHECKLIST 6.3 Developing Training Objectives Primary factors Review existing documents to determine job tasks and competencies required

Translate identified competencies into training objectives

Considerations Sources to examine: • Essential task lists • Previous training objectives • Organization performance standards • Instructors, based on their knowledge and experience Objectives should include: • Specification of behavior to be observed – Use “action” verbs (e.g., “provide,” “prepare,” “locate,” and “decide”) – State the specific behavior(s) that demonstrates the skill or knowledge – State in a way that can be plainly understood by everyone • Conditions under which they are to be exhibited • Standards to which they will be performed or demonstrated – Are standards realistic?


Organize training objectives

Implement training objectives

121

– Are standards clearly stated? – Do standards include completeness, accuracy, timeliness, and performance rates? Categorize: • General objectives specify the end state that trainee should attain • Specific objectives identify tasks that must be performed to meet the general objectives Use training objectives to: • Design exercise training events (e.g., scenarios) – Use events to provide opportunities to evaluate how well trainees performed training objectives • Develop performance measurement tools (e.g., checklists) • Brief trainees on training event

CHECKLIST 6.4 Evaluating Training (Metacognitive training example) Task being observed Narrative

Narrative should provide a context and justify the behavioral ratings assigned 1 2 3 4 Poor: observed Marginal: observed Good: observed Outstanding: observed performance had safety performance was performance was performance was truly implications barely adequate effective noteworthy Execution behavioral markers Behavior to be observed Definition of behavior Evaluator comments: to of behavior Rating: justify rating assigned 1–4 Problem Identified Problem to be solved and objectives to meet were correctly identified Communication Correct information was communicated to the appropriate individuals Tasks Delegated Tasks were delegated to appropriate individuals Monitor Individuals monitored own behaviors and those of others Task Completion Tasks were completed in the correct order and at the appropriate times Review/modify behavioral markers Behavior to be observed Definition of behavior Evaluator Rating: comments: to 1–4 justify rating assigned Evaluation of procedures Procedures were reviewed and modified when appropriate


122

Inquiry Individuals sought help/information when uncertain Source: Adapted from University of Texas at Austin, 2002, LOSA Observation Form Example. Austin, TX: The University of Texas Human Factors Research Project.

Defining Terms Critical incidents technique—“A job analysis technique in which SMEs provide many specific behavioral descriptions of incidents resulting in effective or ineffective performance” (Chmiel, 2000, p. 441). Error reporting—Reporting of errors that resulted in unsafe behaviors but did not lead to an accident. Human error—“…Occasion in which a planned sequence of mental or physical activities fails to achieve its intended outcome” (Reason, 1990, p.9). Job/task analysis—“…Carried out by interview or questionnaire, with generic questions, relying on the incumbent, supervisor or subject matter expert to provide job specific information…[the worker] on answering the questions,…is drawing on an implicit analysis of the job and its knowledge and skill requirements, without which, training needs cannot be determined” (Lees and Cordery, 2000, p. 47). Motivation—“…Refers to psychological processes that cause arousal, direction, and persistence of behavior. Arousal and persistence focus on the time and effort that an individual invests, and direction refers directly to the behaviors in which the investment of time and effort are made” (Goldstein, 1993, p. 90). Organizational analysis—“…Begins with an examination of the short- and long-term goals of the organization, as well as of the trends that are likely to affect goals…focuses on training program and support systems—for example, selection, human-factors engineering, and work procedures” (Goldstein, 1993, pp. 20–21). Person analysis—“…To identify on an individual basis the skills to be learned by each trainee” (Patrick, 2000). “…concerned with how well a specific employee is carrying out the tasks that comprise the job” (Goldstein, 1993, p. 22). Safety climate—“…Can be determined by examining structural properties of organizations such as system complexity and leadership style…often measured by eliciting worker perceptions about organizational commitment to safety, such as the perceived importance of safety training, management’s attitude to safety and so on” (Chmiel, 2000, p. 270). Training—“The systematic acquisition of skills, rules, concepts, or attitudes that result in improved performance in another environment” (Goldstein, 1993, p. 3). Training effectiveness—“…Why training did or did not achieve its intended outcomes…identifying and measuring the effects of individual, organizational, and training-related factors on training outcomes such as learning and transfer of training” (Kraiger et al., 1993, p. 312). Training evaluation—“Refers to a system for measuring whether trainees have achieved learning outcomes…concerned with issues of measurement and design, the accomplishment of learning objectives, and the attainment of requisite knowledge and skills” (Kraiger et al., 1993, p. 312).


123

Training objectives (instructional objectives)—“From information obtained in the assessment of instructional needs, a blueprint emerges that describes the objectives to be achieved by the trainee upon completing the training program” (Goldstein, 1993, p. 23). “…are vital ingredients in the development of training, and influence or determine the content, design and evaluation of training (Patrick, 2000, p. 110). Training outcomes—“Outcomes…are defined as their name implies, as the outcome of the various task processes” (Cannon-Bowers and Salas, 1997, p. 51). Outcomes “may be regarded as a combination of behavior and results criteria” (Kozlowski et al., 2000, p. 192) and/or as “changes in individual behavior” (p. 193). Examples of outcomes include: productivity, time, number of errors, and number of safety behaviors exhibited. References Alliger, G.M. and Januck, E.A. (1989). Kirkpatrick’s levels of training criteria: thirty years later. Personnel Psychol. , 42, 331–342. Alliger, G.M., Tannenbaum, S.I., and Bennett, W., Jr. (1997). A meta-analysis of the relations among training criteria. Personnel Psychol. , 50, 341–358. Baldwin, T.T. and Ford, J.K. (1988). Transfer of training: a review and directions for future research. Personnel Psychol. , 41, 63–105. Baldwin, T.T. and Magjuka, R.J. (1997). Training as an organizational episode: pretraining influences on trainee motivation. In J.K.Ford et al. (Eds.), Improving Training Effectiveness in Work Organizations (99–127). Mahwah, NJ: Lawrence Erlbaum Associates. Baldwin, T.T, Magjuka, R.J., and Loher, B.T. (1991). The perils of participation: effects of choice of training on trainee motivation and learning. Personnel Psychol. , 44, 51–65. Barach, P. and Small, S.D. (2000). Reporting and preventing medical mishaps: lessons from nonmedical near miss reporting systems. Br. Med. J. , 320, 759–763. Bassi, L.J. and Van Buren, M.E. (1998). The 1998 ASTD state of the industry report. Training Dev. , January, 21–43. Baumgartel, H., Reynolds, M., and Pathan, R. (1984). How personality and organizational-climate variables moderate the effectiveness of management development programs: a review and some recent research findings. Manage. Labour Stud. , 9, 1–16. Blickensderfer, E.L., Cannon-Bowers, J.A., and Salas, E. (1997). Theoretical bases for team selfcorrection: fostering shared mental models. In M.Beyerlein, D.Johnson, and S.Beyerlein (Eds.,), Advances in Interdisciplinary Studies in Work Teams Series (Vol. 4, 249–279). Greenwich, CT: JAI Press. Bowers, C.A., Blickensderfer, E.L., and Morgan, B.B. (1998). Air traffic control specialist team coordination. In M.W.Smolensky and E.S.Stein (Eds.), Human Factors in Air Traffic Control (215–236). San Diego, CA: Academic Press. Brinkerhoff, R.O. and Montesino, M.U. (1995). Partnership for training transfer: lessons from a corporate study. Hum. Resour. Dev. Q. , 6, 263–274. Campbell, J.P. (1971). Personnel training and development. Annu. Rev. Psychol. , 22, 565–602. Cannon-Bowers, J.A. and Salas, E. (1997). A framework for measuring team performance measures in training. In M.T.Brannick, E.Salas, and C.Prince (Eds.), Team Performance Assessment and Measurement: Theory, Methods, and Applications (45–62). Hillsdale, NJ: Lawrence Erlbaum Associates. Cannon-Bowers, J.A., Burns, J.J., Salas, E., and Pruitt, J.S. (1998). Advanced technology in scenario-based training. In J.A.Cannon-Bowers and E.Salas (Eds.), Making Decisions under Stress: Implications for Individual and Team Training (365–374). Washington, D.C.: American Psychological Association.


124

Carlson, R.A., Lundy, D.H., and Schneider, W. (1992). Strategy guidance and memory aiding in learning problem-solving skill. Hum. Factors , 34, 129–145. Chmiel, N. (2000). Safety at work. In N.Chmiel (Ed.), Introduction to Work and Organizational Psychology: a European Perspective (255–273). Maiden, MA: Blackwell Publishers. Cohen, S.G. and Ledford, G.E. (1994). The effectiveness of self-directed teams. Hum. Relations , 47, 13–43. Collins v. Marker Heights . (1992). Case No. 503 U.S. 115. Retrieved online May 31, 2002, from: http://caselaw.lp.findlaw.com/scripts/getcase.pl?navby=caseandcourt=USandvol=503andinvol= 115. Colquitt, J.A., LePine, J.A., and Noe, R.A. (2000). Toward an integrative theory of training motivation: a meta-analytic path analysis of 20 years of research. J. Appl. Psychol. , 85(5), 678– 707. Cooke, NJ. (1999). Knowledge elicitation. In F.T.Durso, R.S.Nickerson, R.W.Schvaneveldt, S.T.Dumais, D.S.Lindsay, and M.T.H.Chi (Eds.), Handbook of Applied Cognition (479–509). New York: John Wiley & Sons. Croft, J. (2001). Research perfects new ways to monitor pilot performance. Aviation Week Space Technol. , 155(3), 76–81. De Jong, J.A. and Versloot, B. (1999). Structuring on-the-job training: report of a multiple case study. Int. J. Training Dev. , 3(3), 186–199. Decker, S. (2001). The Field Guide to Human Error Investigations . Aldershot, Hampshire: Ashgate. Degani, A. and Wiener, E.L. (1997). Philosophy, policies, procedures and practices: the four Ps of flight deck operations. In N.Johnston, N.McDonald, and R.Fuller (Eds.), Aviation Psychology in Practice (44–67). Aldershot: Avebury. Dormann, T. and Frese, M. (1994). Error training: replication and the function of exploratory behavior. Int. J. Hum-Computer Interaction , 6, (4), 365–372. Driskell, J.E. and Johnston, J.H. (1998). Stress exposure training. In J.A.Cannon-Bowers and E.Salas (Eds.), Making Decisions under Stress: Implications for Individual and Team Training (191–217). Washington, D.C.: American Psychological Association. Driskell, J.E., Johnston, J.H., and Salas, E. (2001). Does stress training generalize to novel settings? Hum. Factors , 43(1), 99–110. Dubin, S.S. (1990). Maintaining competence through updating. In S.L.Willis and S.S.Dubin (Eds.), Maintaining Professional Competence (9–43). San Francisco: Jossey-Bass. Dunton v. County of Suffolk (1984). Case number 729 F.2d 903. Second Circuit. Dweck, C.S. (1986). Motivational processes affecting learning. Am. Psychol. , 41, 1040–1048. Dweck, C.S. and Leggett, E.L. (1988). A social-cognitive approach to motivation and personality. Psychol. Rev. , 95, 256–273. Ehrenstein, A., Walker, B., Czerwinski, M., and Feldman, E. (Eds.) (1997). Some fundamentals of training and transfer: practice benefits are not automatic. In Training for a Rapidly Changing Workplace: Applications of Psychological Research (31–60), Washington, D.C.: American Psychological Association. Entin, E.E. and Serfaty, D. (1999). Adaptive team coordination. Hum. Factors , 41(2), 312–325. Feldman, D. (1988). Managing Careers in Organizations . Glenview, IL: Scott Foresman. Field, L. (1995). Organizational learning: basic concepts. In G.Foley (Ed.), Understanding Adult Education and Training . Sydney: Allen-Undwin. Fiedler, F.E., Bell, C.H., Chemers, M.M., and Patrick, D. (1984). Increasing mine productivity and safety through management training and organization development: a comparative study. Basic Appl. Social Psychol. , 5(1), 1–18. Ford, J.K. and Weissbein, D.A. (1997). Transfer of training: an updated review and analysis. Performance Improvement Q., 10, 22–41. Ford, J.K. and Noe, R.A. (1987). Self-assessed training needs: the effects of attitudes toward training, managerial level, and function. Personnel Psychol. , 40, 39–53.


125

Ford, J.K., Kozlowski, S., Kraiger, K., Salas, E., and Teachout, M. (Eds.) (1997). Improving Training Effectiveness in Work Organizations . Mahwah, NJ: Laurence Erlbaum Associates. Ford, J.K., Quinones, M., Sego, D., and Speer, J. (1991). Factors affecting the opportunity to use trained skills on the job. Paper presented at the 6th Annu. Conf. Soc. Ind. Organ. Psychol., St. Louis, MO. Forret, M.L., Turban, D.B., and Dougherty, T.W. (1996). Issues facing organizations when implementing formal mentoring programmes. Leadership Organ. Dev. J. , 17(3), 27–30. Retrieved January 17, 2003, from InfoTrack OneFile database. Fowlkes, J., Dwyer, D.J., Oser, R.L., and Salas, E. (1998). Event-based approach to training (EBAT). Int. J. Aviation Psychol. , 8(3), 209–221. Freeman, C. and Simmon, D.A. (1991). Taxonomy of crew resource management: information processing domain. In R.S.Jensen (Ed.), Proceedings of 6th Annual International Symposium on Aviation Psychology (391–397). Columbus, OH: The Ohio State University. Frese, M. and Altman, A. (1989). The treatment of errors in learning and training. In L.Bainbridge and S.A.Quintanilla (Eds.), Developing Skills with Information Technology (65–86). Chichester, U.K.: John Wiley & Sons. Frese, M., Brodbeck, E, Heinnbokel, T., Mooser, C., Schleiffenbaum, E., and Thiemann, P. (1991). Errors in training computer skills: on the positive function of errors. Hum.-Computer Interaction , 6, 77–93. Gatewood, R.D. and Field, H.S. (1998). Human Resource Selection . Fort Worth, TX: The Dryden Press/ Harcourt Brace College Publishers. Gilligan v. Morgan (1973). Case No. 413 U.S. 1. Retrieved online May 31, 2002 from: http://www.usscplus.com/online/index.asp?case=4130001. Glendon, A.I. (2001). Safety culture. In W.Karwowski (Ed.), International Encyclopedia of Ergonomics and Human Factors (1337–1340). New York: Taylor & Francis. Goldman, D. (2000). Legal landmines to avoid in employment training. Retrieved on June 14, 2002, from http://startribune.hr.com/. Goldstein, I.L. (1980). Training in work organizations. Annu. Rev. Psychol , 31, 229–272. Goldstein, I.L. (1993). Training in Organizations (3rd ed.). Pacific Grove, CA: Brooks/Cole Publishing. Goldstein, I.L. and Ford, J.K. (2002). Training in Organizations: Needs Assessment, Development, and Evaluation (4th ed.). Belmont, CA: Wadsworth Publishing. Gordon, S.E. and Gill, R.T. (1997). Cognitive task analysis. In C.E.Zsambok and G.Klein (Eds.), Naturalistic Decision Making (131–140). Mahwah, NJ: Lawrence Erlbaum Associates. Gully, S.M., Payne, S.C., Kiechel Koles, K.L., and Whiteman, J.K. (2002). J. Appl Psychol , 87, (1), 143–155. Hellreigel, D. and Slocum, J.W. (1974). Organizational climate: measures, research, and contingencies. Acad. Manage. J. , 17, 255–280. Helmreich, R.L. (in press). Culture, threat, and error: assessing system safety. In Safety in Aviation: the Management Commitment: Proceedings of a Conference . London: Royal Aeronautical Society. Helmreich, R.L. (2000). On error management: lessons from aviation. Br. Med.J. , 320, 781–785. Helmreich, R.L. (1997). Training and evaluation through simulation in aviation and medicine. Proc. 1997 AERA Annu. Meeting . Chicago, IL. Helmreich, R.L. and Foushee, H.C. (1996). Why crew resource management? Empirical and theoretical bases of human factors training in aviation. In E.L.Wiener, E.C.Kanki, and R.L.Helmreich (Eds.), Cockpit Resource Management (3–45). San Diego, CA: Academic Press. Helmreich, R.L. and Merritt, A.C. (2000). Safety and error management: the role of crew resource management. In B.J.Hayward and A.R.Lowe (Eds.), Aviation Resource Management (107– 119). Aldershot, UK: Ashgate.


126

Helmreich, R.L., Wiener, E., and Kanki, E.C. (1993). The future of Crew Resource Management in the cockpit and elsewhere. In E.Wiener, B.Kanki, and R.L.Helmreich (Eds.), Cockpit Resource Management (479–501). San Diego, CA: Academic Press. Hendricks, C.C. (2001). Teaching causal reasoning through cognitive apprenticeship: what are the results from situated learning? J. Educ. Res. , 94(5), 302–311. Hofmann, D.A. and Stetzer, A. (1996). A cross-level investigation of factors influencing unsafe behaviors and accidents. Personnel Psychol , 49, 307–339. Hofmann, D.A., Jacobs, R., and Landy, F. (1995). High reliability process industries: individual, micro, and macro organizational influences on safety performance. J. Saf. Res. , 26 , 131–149. Huczynski, A.A. and Lewis, J.W. (1980). An empirical study into the learning transfer process management training. J. Manage. Stud. , 17, 227–240. Hunter, J.E. (1986). Cognitive ability, cognitive aptitudes, job knowledge, and job performance. J. Vocational Behav. , 29, 340–362. Isaac, A. (1997). The cave creek incident: a REASONed explanation. Austr. J. Disaster . Article retrieved online on June 12, 2002, from http://www.massey.ac.nz/~trauma/issues/1997– 3/isaacl.htm. Ivancic, K. and Hesketh, B. (1995). Making the best of errors during training. Training Res. J. , 1, 103–125. James, L.R. and Jones, A.P. (1974). Organizational climate: a review of theory and research. Psychol. Bull , 81, 1096–1112. Jentsch, F. (1997). Metacognitive training for junior team members: solving the “copilot’s catch22.” Unpublished doctoral dissertation, University of Central Florida, Orlando. Karl, K.A., O’Leary-Kelly, A.M., and Martocchio, J.J. (1993). The impact of feedback and selfefficacy on performance in training. J.Organ. Behav. , 14, (4), 379–394. Kirkpatrick, D.L. (1976). Evaluation of training. In R.L. Craig (Ed.), Training and Development Handbook: a Guide to Human Resource Development (2nd ed., 1–26). New York: McGraw Hill. Klair, M.B. (2000). The mediated debrief of problem flights. In R.K.Dismukes and G.M.Smith (Eds.), Facilitation and Debriefing in Aviation Training and Operations (72–92). Aldershot: Ashgate. Klein, G.A. (1997). Developing expertise in decision making. Unpublished draft. Klein, G. and Militello, L. (2001). Some guidelines for conducting a cognitive task analysis. In E.Salas (Ed.), Advances in Human Performance and Cognitive Engineering Research (Vol. 1, 161–199). Amsterdam: Elsevier Science. Kleinmen, D.L. and Serfaty, D. (1989). Team performance assessment in distributed decisionmaking. In R.Gibson, J.P.Kincaid, and B.Goldiez (Eds.), Proc. Interactive Networked Simulation Training Conf. (22–27). Orlando, FL: Naval Training System Center. Kletz, T.A. (1985). An Engineer’s View of Human Error . Warwickshire, England: The Institution of Chemical Engineers. Köhn, L.T., Corrigan, J.M., and Donaldson, M.S. (Eds.) (1999). To Err Is Human: Building a Safer Health System . Washington, D.C.: National Academy Press. Kozlowski, S.W.J. (1998). Training and developing adaptive teams: theory, principles, and research. In J.A.Cannon-Bowers and E.Salas (Eds.), Making Decisions under Stress: Implications for Individual and Team Training (115–153). Washington, D.C.: American Psychological Association. Kozlowski, S.W.J., Brown, K., Weissbein D., Cannon-Bowers J., and Salas, E. (2000). A multilevel approach to training effectiveness: enhancing horizontal and vertical transfer. In K.Klein and S.W.J. Kozlowski (Eds.), Multilevel Theory, Research and Methods in Organization (157–210). San Francisco, CA: Jossey—Bass. Kraiger, K., Ford, J.K., and Salas, E. (1993). Application of cognitive, skill-based, and affective theories of learning outcomes to new methods of training evaluation. J. Appl. Psychol. , 78(2), 311–328.


127

Leape, L.L. (1994). The preventability of medical injury. In M.S.Bogner (Ed.), Human Error in Medicine (13–25). Hillsdale, NJ: Lawrence Erlbaum Associates. Lees, C.D. and Cordery, J.L. (2000). Job analysis and design. In N.Chmiel (Ed.), Introduction to Work and Organizational Psychology: a European Perspective (45–68). Maiden, MA: Blackwell Publishers. Lewis, M. (1998). What workers need to succeed. Am. Machinist , 142(3), 88–98. Lorenzet, S.J., Salas, E., and Tannenbaum, S.I. (2003). The impact of guided errors on skill development and self-efficacy. Manuscript submitted for publication. Marks, M.A. (2000). A critical analysis of computer simulations for conducting team research. Small Group Res. , 31(6), 653–675. Martocchio, J.J. and Webster, J. (1992). Effects of feedback and cognitive playfulness on performance in microcomputer software training. Personality Psychol , 45, 553–578. Mathieu, J.E., Tannenbaum, S.I., and Salas, E. (1992). An influence of individual and situational characteristics on training effectiveness measures. Acad. Manage. J. , 35, 827–847. Mobley, G.M., Jaret, C., Marsh, K., and Lim, Y.Y. (1994). Mentoring, job satisfaction, gender, and the legal profession. Sex Roles: J. Res. , 31, 79–98. Retrieved January 17, 2003, from InfoTrac OneFile database. Monell v. Department of Social Services of the City of New York . (1978). Case No. 436 U.S. 658. Retrieved May 31, 2002, from http://www.k9fleck.org/k9cases.htmtfMONELLvNEwYORK. Noe, R.A. (1986). Trainee attributes and attitudes: neglected influences on training effectiveness. Acad. Manage. Rev. , 4, 736–749. OSHA (Occupational Safety and Health Administration). (2002). Fatal explosion and fire lead to over $71,000 in proposed fines for Worcester Company. OSHA Regional News Release. Retrieved May 31, 2002, from http://www.osha.gov/pls/oshaweb/owadisp.show_document?p_table=NEWS_ RELEASES&p_id=651 &p_text_version=FALSE. Oklahoma City v. Tuttle . (1985). Case No. 471 U.S. 808. Retrieved May 31, 2002, from http://case-law.lp.findlaw.com/scripts/getcase.pl?navby=case&court=US&vol=471&invol=808. Orlady, H.W. and Orlady, L.M. (1999). Human Factors in Multicrew Flight Operations . Aldershot: Ashgate. Oser, R.L., Gualtieri, J.W., Cannon-Bowers, J.A., and Salas, E. (1999). Training team problem solving skills: an event-based approach. Computers Hum. Behav. , 15(3–4), 441–462. Patrick, J. (2000). Training. In N.Chmiel (Ed.), Introduction to Work and Organizational Psychology: a European Perspective (100–124). Maiden, MA: Blackwell Publishers. Phillips, J.M. and Gully, S.M. (1997). Role of goal orientation, ability, need for achievement, and locus of control in the self-efficacy and goal setting process. J. Appl Psychol. , 82, 792–802. Pidgeon, N.F. (1991). Safety culture and risk management in organizations. J.Cross-Cultural Psychol. , 22, 129–140. Pidgeon, N.F. (1998). Safety culture: key theoretical issues. Work Stress , 12, 202–216. Pidgeon, N.F. and O’Leary, M. (1994). Organizational safety culture: implications for aviation practice. In N.A.Johnston, N.McDonald and R.Fuller (Eds.), Aviation Psychology in Practice (21–43). Aldershot: Avebury Technical. Pietro, D.A., Shyavitz, L.J., Smith, R.A., and Auerbach, B.S. (2000). Detecting and reporting medical errors: why the dilemma? Br. Med.J. , 320, 794–796. Prince, C. and Salas, E. (1993). Training and research for teamwork in the military aircrew. In E.L.Wiener, E.C.Kanki, and R.L.Helmreich (Eds.), Cockpit Resource Management (337–366). New York: Academic Press. Quinones, M.A. (1995). Pretraining context effects: training assignment as feedback. J. Appl Psychol , 80, 226–238. Quinones, M.A. (1997). Contextual influencing on training effectiveness. In M.A.Quinones and A. Ehrenstein (Eds.), Training for a Rapidly Changing Workplace: Applications of Psychological Research (177–200). Washington, D.C.: American Psychological Association.


128

Reason, J. (1990). Human Error . New York: Cambridge University Press. Reason, J. (1994). Foreword. In M.S.Bogner (Ed.), Human Error in Medicine . Hillsdale, NJ: Lawrence Erlbaum Associates. Reason, J. (1998). Achieving a safe culture: theory and practice. Work Stress , 12, 293–306. Ree, M.J. and Earles, J.A. (1991). Predicting training success: not much more than G. Personality Psychol. , 44, 321–332. Ree, M.J., Carretta, T.R., and Teachout, M.S. (1995). Role of ability and prior job knowledge in complex training performance. J. Appl. Psychol. , 80, 721–730. Richman, H.B., Staszewski, J.J., and Simon, H.A. (1995). Simulation of expert memory using EPAM IV. Psychol. Rev. , 102(2), 305–330. Roberts, K.H. and Bea, R. (2001). When systems fail. Organ. Dynamics , 29(3), 179–191. Rosow, J.M. and Zager, R. (1988). Training the Competitive Edge . San Francisco: Jossey-Bass. Rouiller, J.Z. and Goldstein, I.L. (1993). The relationship between organizational transfer climate and positive transfer of training. Hum. Resour. Dev. Q ., 4, 377–390. Sacks, M. (1994). On-the-Job Learning in the Software Industry . Westport, CT: Quorum Books. Salas, E. and Cannon-Bowers, J.A. (1997). Methods, tools, and strategies for team training. In M.A. Quinones and A.Ehrenstein (Eds.), Training for a Rapidly Changing Workplace: Applications of Psychological Research (249–279). Washington, D.C.: American Psychological Association. Salas, E. and Cannon-Bowers, J.A. (2000a). The anatomy of team training. In S.Tobias and J.D.Fletcher (Eds.), Training and Retraining: A Handbook for Business, Industry, Government, and the Military (312–335). New York: Macmillan Reference. Salas, E. and Cannon-Bowers, J.A. (2000b). Designing training systems systematically. In E.A.Locke (Ed.), The Blackwell Handbook of Principles of Organizational Behavior (43–59). Maiden, MA: Blackwell Publisher. Salas, E. and Cannon-Bowers, J.A. (2001). The science of training: a decade of progress. Annu. Rev. Psychol. , 52, 471–499. Salas, E., Bowers, C.A., and Edens, E. (Eds.). (2001a). Improving Teamwork in Organizations: Applications of Resource Management Training . Hillsdale, NJ: Lawrence Erlbaum Associates. Salas, E., Burke, C.S., Bowers, C.A., and Wilson, K.A. (2001b). Team training in the skies: does crew resource management (CRM) training work? Hum. Factors , 43(4), 641–674. Salas, E., Burgess, K.A., and Cannon-Bowers, J.A. (1995). Training effectiveness techniques. In J.Weiner (Ed.), Research Techniques in Human Engineering (439–471). Englewood Cliffs, NJ: Prentice Hall. Salas, E., Cannon-Bowers, J.A., and Johnston, J.H. (1997). How can you turn a team of experts into an expert team? Emerging training strategies. In C.E.Zsambok and G.Klein (Eds.), Naturalistic Decision Making (359–370). Hillsdale, NJ: Lawrence Erlbaum Associates. Salas, E., Cannon-Bowers, J.A., Rhodenizer, L., and Bowers, C.A. (1999). Training in organizations: myths, misconceptions, and mistaken assumptions. Personnel Hum. Resour. Manage. , 17, 123–161. Salas, E., Dickenson, T.L., Converse, S.A., and Tannenbaum, S.I. (1992). Toward an understanding of team performance and training. In R.J.Swezey and E.Salas (Eds.), Teams: Their Training and Performance (3–29). Norwood, NJ: Ablex. Salas, E., Wilson, K.A., Burke, C.S., and Bowers, C.A. (2002). Implementing crew resource management (CRM) training: myths to avoid. Ergonomics in Design , 10, 20–24. Scandura, T.A., Tejeda, M.J., Werther, W.B., and Lankau, M.J. (1996). Perspectives on mentoring. Leadership Organ. Dev. J. , 17, 50–56. Retrieved January 17, 2003, from InfoTrac OneFile Database. Serfaty, D., Entin, E.E., and Johnston, J.H. (1998). Team coordination training. In J.A.CannonBowers and E.Salas (Eds.), Making Decisions under Stress: Implications for Individual and Team Training (221–246). Washington, D.C.: American Psychological Association.


129

Sexton, J.B., Thomas, E.J., and Helmreich, R.L. (2000). Error, stress, and teamwork in medicine and aviation: cross sectional surveys. Br. Manage.J. , 320 , 745–749. Shappell, S.A. and Wiegmann, D.A. (1997). A human error approach to accident investigation: the taxonomy of unsafe operations. Int. J. Aviation Psychol , 7(4), 269–291. Shute, V.J. and Gawlick, L.A. (1995). Practice effects on skill acquisition, learning outcome, retention, and sensitivity to relearning. Hum. Factors , 37, 781–803. Smith-Jentsch, K., Salas, E., and Baker, D.P. (1996). Training team performance-related assertiveness. Personnel Psychol , 49, 909–936. Smith-Jentsch, K.A., Zeisig, R.L., Acton, B., and McPherson, J.A. (1998). Team dimensional training: a strategy for guided team self-correction. In J.A.Cannon-Bowers and E.Salas (Eds.), Making Decisions under Stress: Implications for Individual and Team Training (271–297). Washington, D.C.: American Psychological Association. Springfield v. Kibbe (1987). Case No. 480 U.S. 257. Retrieved online May 31, 2002 from: http://case-law.lp.findlaw.com/scripts/getcase.pl?navby=case&court=US8cvol=480&invol=257. Tannenbaum, S.I., Mathieu, J.E., Salas, E., and Cannon-Bowers, J.A. (1991). Meeting trainees’ expectations: the influences of training fulfillment on the development of commitment, selfefficacy, and motivation. J. Appl. Psychol , 76, 759–769. Tannenbaum, S.I. and Yukl, G. (1992). Training and development in work organizations. Annu. Rev. Psychol , 43, 399–441. Tracy, B.J., Tannenbaum, S.I., and Kavanagh, M.J. (1995). Applying trained skills on the job: the importance of the work environment. J. Appl Psychol , 80, 239–252. Turner, B.A. (1991). The development of a safety culture. Chem. Ind. , 1, 241–243. United States v. Virginia et al (1996). Case No. 518 U.S. 515. Retrieved May 31, 2002, from http://%20supct.law.cornell.edu/supct/html/94–1941.ZS.html. University of Texas at Austin. (2002). LOSA observation form example. Austin, TX: The University of Texas Human Factors Research Project. Vaughan, D. (1996). The Challenger Launch Decision . Chicago: The University of Chicago Press. Volpe, C.E., Cannon-Bowers, J.A., Salas, E., and Spector, P.E. (2001). The impact of cross-training on team functioning: an empirical investigation. In R.W.Swezey and D.H.Andrews (Eds.), Readings in Training and Simulation: a 30-Year Perspective (115–128). Santa Monica, CA: Human Factors and Ergonomics Society. Wehrenberg, S.B. (1987). Supervisors as trainers: the long-term gains of OJT. Personnel I. , 66(4), 48–51. Wiener, E.L., Kanki, E.C., and Helmreich, R.L. (Eds.) (1993). Cockpit Resource Management . New York: Academic Press. Williams, T.C., Thayer, P.W., and Pond, S.B. (1991). Test of a model of motivational influences on reactions to training and learning. Paper presented at the meeting of the Society for Industrial and Organizational Psychology, St. Louis. Wilson, P.F. and Johnson, W.B. (2001). Core virtues for the practice of mentoring. J. Psychol Theol. , 29(2), 121–130.

Further Information Barling, J., Weber, T., and Kelloway, E.K. (1996). Effects of transformational leadership training on attitudinal and financial outcomes: a field experiment. J. Appl Psychol , 81(6), 827–832. Helmreich, R.L., Merritt, A.C., and Wilhelm, J.A. (1999). The evolution of crew resource management training in commercial aviation. Int. J. Aviation Psychol. , 9(1), 19–32. Reason, J. (1990). Human Error . New York: Cambridge University Press.


130

Salas, E. and Cannon-Bowers, J.A. (2000). Designing training systems systematically. In E.A.Locke (Ed.), The Blackwell Handbook of Principles of Organizational Behavior (43–59). Maiden, MA: Blackwell Publisher. Salas, E. and Cannon-Bowers, J.A. (2001). The science of training: a decade of progress. Annu. Rev. Psychol. , 52, 471–499. Salas, E., Bowers, C.A., and Edens, E. (Eds.) (2001). Improving Teamwork in Organizations: Applications of Resource Management Training . Hillsdale, NJ: Lawrence Erlbaum Associates. Wiener, E.L., Kanki, B.C., and Helmreich, R.L. (Eds.) (1993). Cockpit Resource Management . San Diego, CA: Academic Press.

7 The Influence of Daubert on Expert Witness Testimony—The Human Factors Context Jone McFadden papinchock SHL Litigation Support Frank J.Landy SHL Litigation Support 0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

The role of the human factors expert witness in Federal Court was dramatically changed with the United States Supreme Court decision in Daubert v. Merrell Dow Pharmaceuticals in 1993. 1 This chapter provides basic information about the general impact of Daubert on legal proceedings and illustrations as to the specific impact in human factors cases. In particular, it provides guidance for human factors experts in preparing for and delivering testimony in the litigation context. 7.1 Some Rationale The reason for providing “safeguards” in the case of expert testimony is to prevent the “finder of fact” (judge in the case of a bench trial, or jury in a nonbench trial) from being unduly swayed by the mantle of “expert” testimony. By definition, expert testimony deals with areas outside the knowledge domain of 1

Jason Daubert and Eric Schuller sued Merrell Dow, alleging that their birth defects were the result of their mothers’ use of a prescription antinausea drug (Bendectin) during pregnancy.the judge or jury. As a result, there may be an inclination simply to accept anything said by an expert as “truth” in a general sense. The purpose of guidelines such as those in Daubert is to prevent the finder of fact from even hearing expert testimony that lacks scientific foundation. In practice, many judges feel sufficiently confident to admit questionable testimony if the trial is a bench trial. When objections are raised by opposing counsel, they often respond by saying that they will admit the testimony but may not place much weight on it. 2

Judges tend to be more cautious, however, with what a jury may hear, believing that judges and attorneys are better able to discern flawed scientific findings than jurors (Kovera and McAuliff, 2000). When the testimony of an expert may be difficult for a jury to evaluate through common sense, the standard of admissibility maybe increased (Mark, 1999). The 11th Circuit Court of Appeals specifically addressed this issue in Allison v. McGhan Medical Corporation (1999) in the statement that


132

while meticulous Daubert inquiries may bring judges under criticism for donning white coats and making determinations outside their field of expertise, the Supreme Court has obviously deemed this less objectionable than dumping a barrage of questionable scientific evidence on a jury, who would be even less equipped than the judge to make reliability 3 and relevance determinations and more likely than the judge to be awestruck by the expert’s mystique (at 1310). 4 Because human factors litigation invariably involves a jury as the finder of fact, the discussion that follows will assume a jury-trial scenario rather than a bench trial. 7.2 Superceding of Frye by Federal Rules of Evidence in Daubert The Supreme Court provided a ruling in Daubert that had a direct impact on the guidelines for admissibility of scientific evidence through expert testimony in federal courts. The ruling established that the Frye test 5 set in 1923 had been superceded by the Federal Rules of Evidence (U.S. House of Representatives, Committee on Judiciary, 2002), which were originally enacted in 1975 to direct the use of evidence in federal courts on civil as well as criminal cases. 6 It should be noted that as a result of Daubert, the Federal Rules of Evidence have also been accepted by some, but not all, state courts. The State of Florida, for example, still relies on the Frye test. The Frye test was stated in the following way: …just when a scientific principle or discovery crosses the line between the experimental and demonstrable stages is difficult to define. Somewhere in this twilight zone the evidential force of the principle must be recognized, and while courts will go a long way in admitting expert testimony deduced from a well recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs (at 1014). The Frye test had been criticized because of the de facto waiting period that it implied before the “general acceptance” of new theories and their impact on emerging science (Black et al., 1994; Miller et al., 1994; Price and Kelly, 1998; Sanders et al., 2002). 2

One trial judge has indicated that reserving a ruling during a bench trial occurs because it “buys time for study and reflection” that does not exist in a jury trial (Gless, 1995, p. 263). 3 The Court’s use of “reliability” is more commonly referred to as validity by scientists (Sanders, 1994). 4 It should be noted that the U.S. 11th Circuit Court of Appeals made these comments in the context of a case with a 3-day Daubert hearing by the district court. 5 Frye v. United States, 54 App. D.C. 46, 293 E 1013 (1923). 6 Available at www.house.gov/judiciary/Evid2002.pdf

The influence of daubert on expert witness testimony-the human factors context

133

In contrast, the Federal Rules of Evidence (FRE) 402 states that “all relevant evidence is admissible, except as otherwise provided by the Constitution of the United States, by Act of Congress, by these rules, or by other rules prescribed by the Supreme Court pursuant to statutory authority. Evidence which is not relevant is not admissible” (U.S. House of Representatives, Committee on Judiciary, 2002) A refinement to the inclusion of “relevant evidence” is presented in 403, which states that, “although relevant, evidence may be excluded if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence” (U.S. House of Representatives, Committee on Judiciary, 2002). Article VII of the Federal Rules of Evidence includes specific requirements for admissibility of expert testimony in FRE 702, which prescribes that 7

if scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skills, experience, training, or education, may testify thereto in the form of an opinion or otherwise, if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case (U.S. House of Representatives, Committee on Judiciary, 2002). In Daubert, the U.S. District Court in California granted summary judgment for the defendant, specifically ruling that the testimony of the plaintiffs’ experts did not meet the “general acceptance” standards under Frye and that the studies performed by the plaintiffs’ experts were not published and had not been peer-reviewed. Specifically, the court excluded evidence provided by the plaintiffs’ experts in rebuttal of Merrill Dow’s experts about the association between Bendectin and birth defects in humans. When heard by the U.S. Court of Appeals (Ninth Circuit), the ruling was affirmed. The U.S. Supreme Court reversed and remanded the case, indicating that the lower courts had erroneously relied on the Frye test of “general acceptance.” The Court indicated that the Federal Rules of Evidence did not include reference to Frye or to a “general acceptance” standard. The Court assigned what has come to be known as “gatekeeping responsibility” (Daubert at 2800) to the trial judge by indicating that the “trial judge must ensure that any and all scientific testimony or evidence admitted is not only relevant, but reliable” (at 2795). In his opinion, Judge Blackmun further indicated some of the important factors that have come to be recognized as “Daubert factors.” It should be noted that in his opinion, Judge Blackmun wrote that these were “general observations” and that the Court did “not presume to set out a definitive checklist or test” (at 2796). Furthermore, in Kumho Tire Co. v. Carmichael (1999) following the Daubert decision, Justice Breyer wrote, The conclusion in our view, is that we can neither rule out, nor rule in, for all cases and for all time the applicability of the factors mentioned in Daubert, nor can we now do so for a subset of cases categorized by


134

category of expert or by kind of evidence. Too much depends upon the particular circumstances of the particular case at issue. Thus, the following factors should be interpreted in that light: Ordinarily, a key question to be answered in determining whether a theory or technique is scientific knowledge that will assist the trier of fact will be whether it can be (and has been) tested. Scientific methodology today is based on generating hypotheses and testing them to see if they can be falsified; indeed, this methodology is what distinguishes science from other fields of human inquiry (at 2796). 7

“Relevant evidence” is defined in 401 as “evidence having any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence.”

Another pertinent consideration is whether the theory or technique has been subjected to peer review and publication. Publication (which is but one element of peer review) is not a sine qua non of admissibility; it does not necessarily correlate with reliability…and in some instances well grounded but innovative theories will not have been published…. Some propositions, moreover, are too particular, too new, or of too limited interest to be published. But submission to the scrutiny of the scientific community is a component of “good science,” in part because it decreases the likelihood that substantive flaws in methodology will be detected (at 2797). Additionally, in the case of a particular scientific technique, the court ordinarily should consider the known or potential rate of error…and the existence and maintenance of standards controlling the technique’s operation (at 2797). Finally, “general acceptance” can yet have a bearing on the inquiry. A reliability assessment does not require, although it does permit, explicit identification of a relevant scientific community and an express determination of a particular degree of acceptance within that community…. Widespread acceptance can be an important factor in ruling particular evidence admissible, and a known technique that has been able to attract only minimal support within the community…may properly be viewed with skepticism (at 2797).

7.3 Impact of Daubert More than 10 years have passed since the Daubert ruling and during that period dozens of articles have been written on the impact of the ruling. The popular press and many legal tomes have concentrated on the demise of “junk science” in the face of Daubert (Holden, 1998; Price and Kelly, 1998; Riley, 1996; Vondrak, 1998; Worthington et al., 2002). Junk science is commonly defined as “speech or information which claims to have scientific basis but, is false” (http://home.comcast.net/~bkrentzman/glossary.html). “Junk


135

science is the mirror image of real science, with much of the same form but none of the substance. There is the astronomer on one hand, and the astrologist on the other. The chemist is compared with the alchemist…” (Huber, 1991, p. 2). Given this definition, the full impact and complexity of Daubert might be overlooked. The role of the judge is far more straightforward in making a decision on the exclusion of outright “junk science” compared to the nuances associated with determining whether scientific findings apply in a given case or whether expert opinions have been derived using the scientific method. This latter condition most often will affect human factors litigation. Within this type of litigation, “junk science” has probably been a lesser concern than the misapplication or misinterpretation of scientific findings intentionally or unintentionally by experts who lacked the necessary training, experience, or animus. Thus, substantial emphasis is placed on the judge’s role as an interpreter and gatekeeper in human factors cases. Daubert has provided a nomenclature (relevance, reliability, error rate, falsifiablity, peer review, scientific method) upon which judges can determine the appropriateness of expert witness testimony. However, it is unclear how well judges understand all of this nomenclature. The ability of judges to perform the gatekeeper role, regardless of their best efforts, has been a source of concern since the Daubert ruling. It is presumed that many do not have backgrounds necessary to provide the theoretical structure for evaluating the information needed to determine the reliability and relevance of scientific evidence (Dixon and Gill, 2002; Faigman, 1995a; Krauss and Sales, 1998; Saxe and BenShakhar, 1999). This issue was even voiced by Chief Justice Rehnquist and Justice Stevens in their partially dissenting opinion in Daubert when they cautioned against the role of judges as “amateur scientists” (Rehnquist at 2800). In a survey of state court judges, 48% (or 191 judges) reported they were not fully equipped through their education to evaluate the myriad of scientific issues in their courts (Gatowski et al., 2001). The vast majority of the survey respondents (241 of 251) indicated that they had not received training in general scientific methods and principles. Demonstrating this lack of foundation, only a fraction (6%) of the survey participants (23 of 400) gave accurate responses to a question written to assess their level of understanding of the scientific meaning of the term “falsifiability.” In fact, 35% gave responses that highlighted a complete lack of understanding such as “I would want to know if the evidence was falsified” or “I would look at the results and determine if they are false” (Gatowski et al., 2001, p. 445). These results are particularly striking in light of the fact that, in 1994, the Federal Judicial Center issued a copy of the Reference Manual on Scientific Evidence (Federal Judicial Center, 1994) to all federal judges (Walker and Monahan, 1996). This manual contains sections on the scientific method and falsification theory. On this point, in his introduction to the second edition to the Reference Manual on Scientific Evidence, Associate Justice Breyer wrote that “most judges lack the scientific training that might facilitate the evaluation of scientific claims or the evaluation of expert witnesses who make such claims. Judges typically are generalists, dealing with cases that vary widely in subject matter. Our primary objective is usually process related…” (Federal Judicial Center, 2000, p. 4).


136

7.4 Affirmative Role of the Human Factors Expert after Daubert The survey data of Gatowski and colleagues (2001) reported in the preceding section might lead one to consider whether extensive changes in court decisions have occurred since Daubert and what, if any, impact they should have on the role of the human factors expert. In fact, there is some question as to how extensively the Daubert test is applied by judges (even in federal courts), particularly as relates to experts in the social sciences (Dixon and Gill, 2002; Lipton, 1999; Shuman and Sales, 1999; Tenopyr, 1999). In contrast, there seems to be more evidence of the impact of Daubert in the exclusion of experts in other areas of law such as toxic tort cases (Clark, 1996). Furthermore, it has been suggested that fewer Daubert proceedings will take place than expected for a number of reasons, including the possibility that a trial lawyer may not raise the issue of the opposing expert’s credibility because he does not recognize the technical weaknesses, his expert used similar methods and procedures, he recognizes the weaknesses but considers it a strategic advantage to bring the weaknesses directly to the jury’s attention through cross-examination, or he simply wants to control costs (Shuman and Sales, 2001). Any question about the applicability of Daubert to human factors testimony was clarified in Kumho Tire Company v. Carmichael (1999), when the U.S. Supreme Court ruled that the Federal Rules of Evidence govern all expert testimony. In Kumho, the Supreme Court found that the testimony of an expert in a tire failure case (opining about whether a tire failure on a minivan resulted from a defect in the tire’s manufacture or design) was subject to Daubert consideration. Specifically, the Court indicated that Federal Rules of Evidence 702 does not distinguish between scientific and nonscientific knowledge. In further support of the applicability of Daubert to human factors testimony, in a 1998 survey, of the 1281 expert witnesses identified in 297 federal civil trials, 21.8% were designated as “engineering/process/ safety”; these were further deconstructed into “accident reconstructionists” (2%), “products engineers” (2%), and “other engineering/process/safety” (8.3%) (Krafka et al., 2002). This 1998 survey also reported that experts addressed the cause of injury or damage in 63.5% of the cases, the reasonableness of the party’s actions in 34.1% of the cases, and industry standards or state of the art in 30.3%—all issues that human factors experts commonly handle. Moreover, some preliminary evidence suggests that Daubert is affecting case outcomes (Dixon and Gill, 2002). Of particular note is the finding that, since the Daubert ruling, challenges to the relevance and reliability of expert witness opinions have resulted in more summary judgments, with 90% of the judgments against the plaintiffs, “making it likely that challenges to plaintiffs’ evidence increasingly resulted in case dismissal” (Dixon and Gill, 2002, pp. 298–299). Also, review of opinions pre- and post-Daubert indicates a “rapid rise in the frequency with which judges addressed the clarity and coherence of the expert’s explanation of the theory, methods, and procedures underlying the evidence” (Dixon and Gill, 2002, p. 299). In their survey of judges, Krafka and colleagues (2002) reported that pretrial scrutiny of experts had increased between 1991 and 1998 and that judges were less likely to admit the testimony in 1998. There seems little doubt that human factors experts are, and will continue to be, challenged on Daubert standards.


137

This means that the human factors expert should carefully consider his or her role in the context of expert testimony. Preparing one’s testimony as if it will be challenged on the basis of Daubert is never a waste of time for an expert. It ensures that he will locate relevant research, use a scientific approach to gathering and evaluating evidence (“through the methods and procedures of science”; Daubert at 2795), and be prepared for trial testimony and cross-examination (Shuman and Sales, 2001). Shuman and Sales liken preparing opinions as if there will be a Daubert challenge to “Pascal’s wager,” which recommends that little risk is attached to choosing to believe in God but greater potential risk is associated with not believing. Minimally, the expert will serve as an educator and resource to the judge in his or her gatekeeping capacity and as an interpreter of the applicability of technical or scientific research to a given case. The expert should carefully craft the expert report because a substantial proportion of judges (63%) have indicated that expert reports assist them in focusing on case issues (Krafka et al., 2002). From this perspective, it is clear that it is insufficient simply to present theories, particularly sophisticated or complex theories, without explaining them in lay terms and drawing parallels to the facts of the current case. This role of the testifying expert has been described as “assisted sense making” (Mark, 1999). In the role of educator, the human factors expert must competently present sophisticated concepts not only in his specialty area but also in basic scientific methods. The expert witness cannot assume that the judge or jury will have sufficient background in the scientific method to proceed immediately to more complex theories. The language of the expert should be easily understandable. The human factors expert must also demonstrate that he has relied on the scientific method in reviewing case information, performing tests, analyzing results, and arriving at conclusions. “One of the basic lessons of Daubert should not be lost: Daubert exhorts scientists to do good science and expects them to be scientists first and expert witnesses (and advocates) second” (Faigman, 1995b, p. 975). That is, the expert demonstrates that “through the use of proper methods we can know with some degree of certainty which of our judgments about the world are correct” (Sanders et al., 2002, p. 151). The human factors expert must also carefully explain the linkages between the theories presented and the facts of the case (Thornton and Wingate, in press). As described in the Daubert decision, “‘fit’ is not always obvious, and scientific validity for one purpose is not necessarily scientific validity for other, unrelated purposes Rule 702’s ‘helpfulness’ standard requires a valid scientific connection to the pertinent inquiry as a precondition to admissibility” (at 2796). It is insufficient simply to state that a particular theory has been published and is well grounded from a scientific basis. It must be tied to the facts of the case. One must demonstrate that the theory proposed is applicable to the facts of the case. Some preliminary research indicates that, particularly when testimony is nonscientific in nature, demonstrating the applicability of the Daubert factors may be less important than demonstrating the witness’s capacity to assist the trier of facts and that the testimony is relevant and nonprejudicial (Groscup et al., 2002).


138

7.5 Rebuttal Role of the Human Factors Expert after Daubert Given the potential lack of skill among judges (and, to a greater extent, the jury) to evaluate the use of the scientific method and determine reliability and relevance, it is also important to recognize and describe clearly for them the flaws in the opposing expert’s arguments (Kovera et al., 2002). In fact, it has been recognized that the Daubert decision was predicated on the assumption that a safeguard is built into the judicial process by the use of opposing expert evidence that is part of the adversarial process. Specifically, Daubert assumes that “presentation of opposing expert testimony, which contradicts flawed expert testimony, will sensitize jurors to the unreliability of the original expert evidence” (Kovera et al., 2002, p. 184). Given this significant responsibility, it is essential that human factors experts carefully evaluate the testimony of the opposing expert and educate the court on flaws in design, analysis, or interpretation. There are at least four major methods of rebuttal based on flaws in the opposing expert opinions. Each of these will be described next. Rebuttal of Mischaracterization of the Literature Demonstration of the mischaracterization of literature addresses the reliability and relevance of the opposing experts’ opinions in addition to consideration of the known error rate of the published research upon which they rely. This can occur in an expert rebuttal report in several ways: • It is common to read in expert reports that “a scientific body of knowledge” supports a given opinion. Many times no references are provided to delineate the body of knowledge and often no such body of knowledge exists. In the educator role, the rebutting expert should explain such discrepancies. In United States v. Rincon (1994), the Ninth Circuit upheld the trial court’s decision to exclude the proposed testimony of an expert in eyewitness identification. The declaration provided by counsel indicated that “there is a wealth of research supporting this point,” “the research is clear,” and “the research suggests.” Initially, the district court excluded the proposed testimony, indicating that “no showing has been made that the testimony relates to an area that is recognized as a science.” The appellate court indicated that “none of the research was submitted or described so that the district court could determine if the studies were indeed scientific on the basis the Court explained in Daubert? thus providing an indication that Daubert is interpreted as meaning that simply indicating that research exists is insufficient support for an expert opinion. When an article “General Acceptance of Psychological Research on Eyewitness Testimony” (Kassin et al, 1989) 8 was submitted (during the hearing in district court, which was remanded by the Ninth Circuit on remand from the U.S. Supreme Court), the Ninth Circuit found insufficient detail for the district court to determine the scientific validity. Furthermore, the Court of Appeals indicated that it based its support of the district court’s ruling on the lack of information provided by this expert in this case and not a general predisposition against expert testimony on eyewitness identification. The Court of Appeals stated, “Our conclusion does not preclude the admission of such testimony when the proffering party satisfies the standard established in Daubert by showing that the


139

expert opinion is based upon ‘scientific knowledge’ which is both reliable and helpful to the jury in any given case.” • Often, references are provided but inspection reveals that they are nonscientific sources (popular magazines, newspapers, or thought papers) that have not been peer-reviewed or subjected to scientific testing. Thus, such statements may be extremely misleading to judges and juries who assume that the expert is relying on scientific sources. It is important to highlight this factor. As the Supreme Court stated in Daubert, peer review is not a prerequisite but it should be considered, particularly if an expert is presenting concepts from popular sources as if they were based on empirical research. • Even when scientific literature is presented by the opposing expert, it should be carefully reviewed to determine whether it supports the interpretations and conclusions for which it is proffered. It is possible for an expert to “cherry pick” sections of an article or book (for use in an expert report) that are not at all representative of the totality of the source. For example, in a product liability case concerning breast implants, the testimony of the plaintiff’s expert was excluded because of the unreliability of the two published studies that were introduced. In one study of the relationship between breast implants and Sjogren’s syndrome, the statistical significance level was 0.067 and the authors indicated that further research was necessary to confirm conclusions; in the second study, the authors indicated that there were limitations on their findings because of possible bias in their sampling techniques (Kelley v. American Heyer-Schulte Corporation, 1997). These qualifications of results were not presented by the expert who cited the research. • It is also common for experts to overstate the reliability of findings in the research. For example, the research in a given area may be equivocal and as yet undetermined with findings both supporting and opposing a given theory. It is important that the rebutting expert present the 8

This article reported the survey of 63 experts’ opinions about scientific acceptance of research on various topics.

missing perspective from the literature and/or demonstrate the flaws in the studies presented by the opposing expert. One might consider presenting information about the number of times a particular study has been cited based on listing in the Social Science Citation Index. It may also be useful to search for meta-analytic studies that provide insight into differences across individual studies that establish scientific reliability in the face of equivocal findings (Penrod et al., 1995). This approach, of course, introduces its own complexity in terms of explaining the statistical methodology to the judge or jury; nevertheless, it is becoming more common. Rebuttal of Misapplication of Scientific Literature to Case Facts The most difficult rebuttal to prepare is often the explanation of why a body of scientific literature does or does not apply to the facts in a particular case. Clearly, this is an important aspect of the Daubert factors as visualized by the Daubert court. The trier of fact makes the determination of “whether the reasoning or methodology underlying the


140

testimony is scientifically valid and of whether that reasoning or methodology properly can be applied to the facts at issue” (Daubert at 2796). An example of this might be an expert who references a “significant body of literature on the relationship between driver fatigue and accidents,” suggesting that this literature supports his or her opinion that the performance of a truck driver involved in an accident was affected by fatigue. One cannot simply apply a body of literature to a given case without careful consideration of the research base for the literature and the facts of a case. If, upon inspection, it is determined that the research base of the literature proposed by the expert was measurement of fatigue and vehicle accidents among long-distance drivers, this may not be appropriate for application to local-haul drivers. In this situation, for example, the literature on fatigue and vehicle accidents specifically references the impact on long-haul drivers of irregular hours of work, significant periods of night driving, interrupted sleep patterns from sleeping in the truck cab, and potential use of stimulants. The research postulates that any or all of these factors may have an impact on vehicle accidents. As such, the research would not apply in a case of an accident involving a local-haul driver who did not work irregular hours, who did not perform night driving, who slept in his or her own home and did not have signs of interrupted sleep patterns, and who had not used any forms of stimulants. No matter how strong the relationship in the literature is between fatigue and vehicle accidents, if the research has been performed on long-haul drivers, it is of little value in the case at hand. A finding in the literature (based on long-haul drivers) that illustrates the inappropriateness of application to local-haul drivers is that “most instances in which drivers were judged to be drowsy occurred during the nighttime hours (between 7:00 p.m. and 6:59 a.m.)” (Hanowski et al., 1999, p. 378). These are not hours commonly worked by local-haul drivers and, as such, these results would be meaningless if applied to localhaul drivers. According to Daubert, “scientific validity for one purpose is not necessarily scientific validity for other purposes…. (FRE) Rule 702’s ‘helpfulness’ standard requires a valid scientific connection to the pertinent inquiry as a precondition to admissibility” (at 591). This example might also be considered more generally from the perspective of whether this literature has a known error rate. To quote from a keynote address to the Association of Professional Sleep Societies relative to these studies: Although…we have investigated many accidents in which sleep loss, sleep disorders, fatigue and circadian factors are clearly implicated, I don’t think we have the foggiest notion of the true prevalence of these factors in transportation system accidents. One of the most perplexing problems our accident investigators face is how to determine what role, if any, fatigue played in a specific accident. Unlike metal fatigue, human fatigue generally leaves no telltale signs and we can only infer its presence from circumstantial evidence (Brown, 1994, p. 309). One might also consider the applicability of this literature more generally from the perspective of the adequacy of the scientific methodology:


141

There is broad, consensual agreement among transport professionals and the driving public alike about what constitutes unsafe or undesirable driving. It is common sense to pronounce that sleepy, intoxicated or sick drivers constitute a safety hazard to themselves and to other road users. However, there are theoretical problems surrounding the diagnosis of driver impairment beyond this anecdotal level. In the simplest case, the diagnosis of impairment due to alcohol intoxication is relatively straightforward. The amount of alcohol consumed may be measured by the blood alcohol concentration (BAC); this variable has an exponential relationship with accident likelihood…. There is no equivalent index for other categories of driver impairment such as fatigue. This absence may simply be a conceptual limitation. The individual often spontaneously perceives and can report changes in energetic state from a purely phenomenological perspective. In the absence of an anchor scale (such as BAC in the case of alcohol…), the multidimensional character of driver impairment renders the concept susceptible to an undesirable level of indeterminancy [sic]. Ambiguity at the conceptual level inevitably creates practical problems of measurement and interpretation. The limitation is particularly striking when attempting to measure the impact of multivariate energetic states on a complex skill such as driving (Brookhuis et al., 2003, p. 434). Rebuttal of Testimony Provided by Experts Testifying outside Their Area of Expertise Testifying outside one’s area of expertise can occur in any field. Many judges (45% compared to 24% of attorneys) believe that expert reports discourage such testimony (Krafka et al., 2002). However, within human factors, many experts in accident cases, for example, testify without experience or training in areas such as employee training or employee compensation. They will make statements related to a workplace accident resulting in injury such as “the worker was not properly trained in lockout procedures” or the “worker was not adequately trained in driver safety.” Similarly, it is common for experts to make statements such as “if the company had trained the drivers on specific conditions on rural roads with farm equipment, the accident would not have occurred.” These statements are made on the basis of common sense or intuition and may go unnoticed by a judge or jury unless challenged. “Common sense, like science, provides knowledge about natural phenomena, but it does not systematically establish connections between occurrences that are not obviously related” (Black et al., 1994, p. 754). This is a situation in which the opposing expert should identify the qualifications that are required for this testimony, such as education and/or experience in developing, evaluating, researching, or modifying employee training programs. It may not otherwise occur to a judge or jury that a body of science surrounds such issues. It is also useful, again as an educator, for the opposing expert to present the research literature on the identified issues. The same types of statements as “the worker was not properly trained in lockout procedures” can also be rebutted on the basis of the expert’s not relying on a scientific


142

approach to arrive at this conclusion. As the Daubert court indicated, “Science is not an encyclopedia body of knowledge about the universe. Instead it represents a process for proposing and refining theoretical explanations about the world that are subject to further testing and refinement.” However, in order to qualify as “scientific knowledge,” an inference or assertion must be derived by the scientific method. Proposed testimony must be supported by appropriate validation—i.e., “good grounds,” based on what is known. In short, the requirement that an expert’s testimony pertains to “…‘scientific knowledge’ establishes a standard of evidentiary reliability” (Daubert court). Often experts will not only opine that “the worker was not properly trained” but also conclude that the “company was negligent in its training practices.” This latter point should be, but is almost never, based on a scientific approach. One cannot conclude from the behavior of one employee in one incident that training practices were inadequate. One must consider how probable it is that any training program would be 100% effective. That probability would be low. Thus, the expert would not have employed the necessary scientific rigor to arrive at such a conclusion by only considering the actions of the employee involved in an accident. Rebuttal of Misinterpretation of Scientific Findings Again, often without reference to a particular body of literature, experts may misinterpret scientific findings based on common-sense interpretations instead of empirical evidence. These opinions may make intuitive sense to the lay person, so it is essential that the opposing expert present the correct interpretations and concrete examples to demonstrate the error. For example, estimating reaction time is often an element in reconstructing a vehicular accident. This is an area about which jurors may believe (based on a cursory understanding of reaction time gained from reading a driver’s handbook or some other brief exposure to the concept) that an expert is presenting an accurate recreation of some accident or event. From this exposure, the jurors may readily agree with an expert in a vehicle accident case who opines that it took several seconds for the driver to begin responding. The expert will often establish this by breaking a behavior like stopping a car into multiple steps such as noticing an oncoming train, making a decision on what action to take, deciding to stop the car, moving the foot to the brake, and applying the brake. Calculated in this way, the response time is often “estimated” to be 2 sec or more (before adding time for the car to come to a stop based on speed and distance after application of the brake). In terms of the scientific literature on reaction time, numerous studies indicate that traditional simple and choice reaction time estimates are far shorter than experts independently calculate. Sanders and McCormick (1987) offer information on the approximate reaction time for situations with from 1 to 10 choices. Even if there are 10 choices of what behavior to take, the approximate reaction time is 0.65 sec. The estimated reaction time is only 0.35 sec with two choices and 0.20 sec with go/no go choice. Furthermore, studies that specifically consider driver behavior provide even more direct rebuttal evidence. Hills (1980) found that the mean time for 215 subjects to “transfer their foot from the accelerator pedal to the brake in response to a large stop signal” was 0.63 sec. Approximately 85% of the drivers in that study accomplished this


143

task within 0.70 sec (p. 207). These types of studies can be used to rebut such overestimates of reaction time. However, the opposing expert must take the initiative to locate the appropriate studies and present them. Another area often handled inappropriately by experts is motion detection. Specifically, an expert may erroneously indicate that a vehicle driver was not looking in the direction of the train so did not recognize that a train was approaching. The expert will then begin calculations of reaction time at the point at which the driver oriented to the train (when the engineer blew the whistle, for example), implying that motion detection is a complex information-processing task. In fact, the literature indicates that the process of motion detection is almost instantaneous and that motion perception is actually more sensitive in the periphery of the visual field than the central retina (McBurney and Collings, 1984). Thus, one does not need to be looking directly at the train in order to notice it and begin a response. 7.6 Conclusion Daubert provided the courts with criteria for considering whether an expert’s testimony should be admitted or excluded. At the same time, this ruling provided experts with notice of what aspects of their work would be reviewed and evaluated by the court. It also offered insight into how opposing experts could “debunk” the opinions of the first expert. As such, Daubert offers experts a road map for how to perform the duties of an expert witness. This direct link between court procedures and expert witness behavior might be problematic if the U.S. Supreme Court had chosen other parameters of evaluation. No scientist wants to be directed in his or her work by a legal system that is presumably a nonscientific institution. However, because the legal system specified that it would carefully consider good scientific method, reliability, and relevance, there should be relatively few conflicts between the work that the expert would perform based on the expectations of his or her specialty field and that expected by the court. In fact, the U.S. Supreme Court specified that the expert’s methods should mirror those he or she would use outside the context of litigation. An expert can rarely, if ever, go wrong by following good scientific method and relying on well-developed research and literature in his or her field. References Allison v. McGhan Medical Corp., 184 F.3d 1300 (11th Cir. 1999). Black, B., Ayala, F.J., and Saffran-Brinks, C., 1994, Science and the law in the wake of Daubert a new search for scientific knowledge. Tex. Law Rev. , 72, 715–802. Brookhuis, K.A., De Waard, D., and Fairclough, S.H., 2003, Criteria for driver impairment. Ergonomics , 46, 433–445. Brown, I.D., 1994, Driver fatigue. Hum. Factors , 36, 298–314. Clark, M.W., 1996, The impact of Daubert on the admissibility of expert opinion. Advocate , April, 10–17.


144

Daubert v. Merrell Dow Pharmaceuticals, Inc., 951 F.2d 1128 (9th Cir. 1991), vacated, 113 S.Ct. 2786 (1993). Dixon, L. and Gill, B., 2002, Changes in the standards for admitting expert evidence in Federal civil cases since the Daubert decision. Psychol, Public Policy, Law , 8(3), 251–308. Faigman, D.L., 1995a, Mapping the labyrinth of scientific evidence. Hastings Law J. , 46, 555–579. Faigman, D.L., 1995b, The evidentiary status of social science under Daubert is it “scientific,” “technical,” or “other” knowledge? Psychol., Public Policy, Law , 1, 960–979. Federal Judicial Center, 1994, Reference Manual on Scientific Evidence , 1st ed., Washington, D.C.: Federal Judicial Center. Federal Judicial Center, 2000, Reference Manual on Scientific Evidence , 2nd ed., Washington, D.C.: Federal Judicial Center. Federal Rules of Evidence, H.R. Rep. No. 8, 107th Cong., 2d Sess. (2002), Washington, D.C.: U.S. Government Printing Office. Gatowski, S.I., Dobbin, S.A., Richardson, J.T., Ginsburg, G.P., Merlino, M.L., and Dahir, V., 2001, Asking the gatekeepers: a national survey of judges on judging expert evidence in a postDaubert world. Law Hum. Behav. , 25, 433–458. Gless, A.G., 1995, Some post-Daubert trial tribulations of a simple country judge: behavioral science evidence in trial courts. Behav. Sci. Law , 13, 261–291. Groscup, J.L., Penrod, S.D., Studebaker, C.A., Huss, M.T., and O’Neil, K.M., The effects of Daubert on the admissibility of expert testimony in state and federal criminal cases. Psychology, Public Policy, and Law , 8, 339–372. Hanowski, R.J., Wierwille, W.W., Gellatly, A.W., Dingus, T.A., Knipling, R.R., and Carrol, R., 1999, Safety concerns of local/short haul truck drivers. Transp. Hum. Factors , 1(4), 377–386. Hills, B.L.,1980, Vision, visibility, and perception in driving. Perception , 9 , 183–216. Holden, C., 1998, Supreme Court clarifies junk science stance. Science , 279, 35. Huber, P.W., 1991, Galileo’s Revenge: Junk Science in the Courtroom , New York: Basic Books. Kassin, S.M., Ellsworth, P.C., and Smith, V.L., 1989, The “general acceptance” of psychological research on eyewitness testimony. Am. Psychologist , August, 1089–1098. Kelley v. American Heyer-Schulte Corp., 957 F. Supp. 873 (W.D. Tex. 1997). Kovera, M.B. and McAuliff, B.D., 2000, The effects of peer review and evidence quality on judge evaluations of psychological science: are judges effective gatekeepers? J. Appl Psychol , 85, 574–586. Kovera, M.B., Russano, M.B., and McAuliff, B.D., 2002, Assessment of the common sense psychology underlying Daubert legal decision makers’ abilities to evaluate expert evidence in hostile work environment cases. Psychol, Public Policy, Law , 8, 180–200. Krafka, C., Dunn, M.A., Johnson, M.T., Cecil, J.S., and Miletich, D., 2002, Judge and attorney experiences, practices, and concerns regarding expert testimony in federal civil trials. Psychol, Public Policy, Law , 8, 309–332. Krauss, D.A. and Sales, B.D., 1998, The problem of “helpfulness” in applying Daubert to expert testimony: child custody determinations in family law as an exemplar. Psychol, Public Policy, Law , 5, 78–99. Kumho Tire Co. v. Carmichael, 526 U.S. 137 (11th Cir. 1999). Lipton, J.P., 1999, The use and acceptance of social science evidence in business litigation after Daubert . Psychology, Public Policy, and Law , 5, 59–77. Mark, M.M., 1999, Social science evidence in the courtroom: Daubert and beyond? Psychol., Public Policy, Law , 5, 175–193. McBurney, D.H. and Collings, V.B., 1984, Introduction to Sensation/Perception , 2nd ed., Englewood Cliffs, NJ: Prentice Hall. Miller, P.S., Rein, B.W., and Bailey, E.O., 1994, Daubert and the need for judicial scientific literacy. Judicature , 77, 254–260.


145

Penrod, S.D., Fulero, S.M., and Cutler, B.L., 1995, Expert psychological testimony on eyewitness reliability before and after Daubert the state of the law and the science. Behav. Sci. Law , 13, 229–259. Price, J.M. and Kelly, G.G., 1998, Junk science in the courtroom: cause, effects and controls. Hamline Law Rev. , 19, 395–407. Riley, S.E., 1996, The end of an era: junk science departs product liability. Defense Counsel J., 63, 502–508. Sanders J., 1994, Scientific validity, admissibility, and mass torts after Daubert . Minn. Law Rev. , 78, 1387–1441. Sanders, M.S. and McCormick, E.J., 1987, Human Factors in Engineering and Design , 6th ed., New York: McGraw-Hill. Sanders, J., Diamond, S.S., and Vidmar, N., 2002, Legal perceptions of science and expert knowledge. Psychology, Public Policy, and Law , 8, 139–153. Saxe, L. and Ben-Shakhar, G., 1999, Admissibility of polygraph tests: the application of scientific standards post-Daubert . Psychol., Public Policy, Law , 5, 203–223. Shuman, D.W. and Sales, B.D., 1999, The impact of Daubert and its progeny on the admissibility of behavioral and social science evidence. Psychol., Public Policy, Law , 5, 3–15. Shuman, D.W. and Sales, B.D., 2001, Daubert’s wager. J. Forensic Psychol. Pract. , 1, 69–77. Tenopyr, J.L., 1999, A scientist-practitioner’s viewpoint on the admissibility of behavioral and social scientific information. Psychol., Public Policy, Law , 5, 194–202. Thornton, G.C. and Wingate, P.H., (in press), Industrial and organizational psychologists as expert witnesses: Employment discrimination litigation post Daubert . In Landy, F.J. (Ed.), Employment Discrimination Litigation , San Francisco: Jossey-Bass. United States v. Rincon, 921 F.3d 28 (9th Cir. 1994). Vondrak, A., 1998, Junk science finally gets slammed. The Washington Times , 18 October, B5. Walker, L. and Monahan, J., 1996, The future of science in law. Va. Law Rev ., 82, 837–857. Worthington, D.L., Stallard, M.J., Price, J.M., and Goss, P.J., 2002, Hindsight bias, Daubert, and the silicone breast implant litigation: Making the case for court-appointed experts in complex medical and scientific litigation. Psychology, Public Policy, and Law , 8, 154–179.

II Human Performance in the Legal Context

8 Reconstructing Situated Performance in Human Error Investigations Sidney W.A.Dekker Lund University 0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

Every scripture is entitled to be read in the light of the circumstances that brought it forth. To understand the choices open to people of another time, one must limit oneself to what they knew; see the past in its own clothes, as it were, not in ours. Barbara Tuchman, Practicing History: Selected Essays, 1981 (p. 75)

8.1 Introduction: How to Make Sense out of Puzzling Performance How could it have made sense for the pilot to continue his approach despite (what we now know to have been) increasingly bad weather? How could the shift operator have missed the valve setting that, in hindsight, appeared so obviously necessary? How could it have made sense for the surgeon to revert to an open procedure despite (what we now know as) indications that this would be more dangerous for the patient than the laparoscopic one? Forensic human factors experts can be called on to make sense out of such puzzling performances. However, making sense out of other people’s puzzling performances is incredibly difficult. It is difficult, in fact, for two different reasons. The first is that whenever we look at past performance, all kinds of biases start to cloud our ability to understand the reasons for people’s behavior. Even the human factors expert is not immune to this. William James called it the “psychologist’s fallacy”: we easily mistake our own reality for the one that surrounded people at the time. “Standards of care” become difficult to assess objectively, even by experts, when one is faced with the rubble of human failure. Even the expert can see, in all retrospective clarity, the choices that were open to people at a different time and even the expert can feel befuddled at how these people ended up picking the wrong ones. This hindsight bias is persistent. It will confound anybody’s attempt at making sense of past puzzling performance if it is not recognized and controlled. This is discussed in the subsection “Recognizing and Controlling the Hindsight Bias.”

Reconstructing situated performance in human error investigations

149

The second reason why it is difficult to make sense of other people’s puzzling performance is a technical one. Technical, in this sense, refers to the human factors theories, issues, or concepts that could shed light on why people did what they did. Scientifically, it refers to the problem of how to make a credible, traceable connection between the behavioral sequence (which played out in a particular setting) and a human factors concept that covers what went on in there. Matching the context-specific particulars of an accident sequence with a human factors explanation may seem straightforward. For example, the human factors expert could note how, in a cockpit voice recording, pilots were asking each other questions about direction or the location of navigational way-points before crashing into a mountain. The expert then proclaims that the pilots “lost situation awareness.” Yet this only conveys the illusion of understanding. Large labels such as this may sound appealing in a courtroom, but if no verifiable connection is shown between the label and the data it supposedly covers, then it explains nothing. It amounts to folk- or pseudoscience and can easily be attacked by the opposite side on credibility issues alone. Folk modeling fools the advised team members into believing that they have learned something of value. The traps here are severe and deep and need to be discussed (see the subsection “Reconstructing Behavioral Sequences”). The two difficulties are not separate, nor are the ways to deal with each. In fact, much of what human factors calls protocol analysis, or process tracing, is aimed at reconstructing the situation as it evolved from the point of view of the people involved in it at the time. Reconstructing the situation provides a relatively unambiguous scaffolding upon which to hang people’s puzzling assessments and actions. People assessed and acted inside that situation: their performance was “situated.” Thus, if the situation is understood in detail, maybe their assessments and actions will be, too. Much recent human factors work and cognitive theory back this up (see subsection “Reconstructing Behavioral Sequences”). Such reconstruction also persuades the analyst not to judge the situation from a position as omniscient 20/20 retrospective outsider, but from the perpetually incomplete and unfolding perspective of the people caught up in the situation. Such reconstruction continually pulls the analyst away from the lofty hindsight outlook where he enjoys the vista of a completely evolving situation and its disastrous outcome. It compels the analyst back into the trenches where assessments and decisions were actually made with a restricted outlook, under time pressure, with limited knowledge, and under uncertainty— to see how the world looked from there and what it would have been like to assess, decide, and act under those circumstances. In the end, whether standards of care were adhered to can honestly be assessed only when the locally limited perspective of the protagonist is taken and any reference to outcome is erased. Indeed, the quality of decisions cannot be evaluated on the basis of their outcome because the people making those decisions at the time did not know the outcome. The subsection “Recognizing and Controlling the Hindsight Bias” is reserved for more discussion on this.


150

8.2 Objective and Scope of the Chapter This chapter deals with the two difficulties mentioned previously. The subsection “Recognizing and Controlling the Hindsight Bias” shows human factors experts and those who consult them how to recognize and control the hindsight bias. The subsection “Reconstructing Behavioral Sequences” discusses how to reconstruct situated performance technically without getting trapped in pseudoscience. Section 8.4 offers some examples. The chapter then concludes with a checklist for quick reference, defining terms, and references. 8.3 Discussion of Principal Issues Recognizing and Controlling the Hindsight Bias Human factors experts and attorneys know more about the incident or accident than the people who were caught up in it, thanks to hindsight (Fischoff, 1975), which: • Means being able to look back, from the outside, on a sequence of events that led to an outcome they already know about • Provides almost unlimited access to the true nature of the situation that surrounded people at the time (where they actually were vs. where they thought they were or what state their system was in vs. what they thought it was in) • Allows them to pinpoint what people missed and should not have missed and what they did not do but should have done From the perspective of the outside and hindsight, people can oversee the entire sequence of events—the triggering conditions, its various twists and turns, the outcome, and the true nature of circumstances surrounding the route to trouble. This is in stark contrast with the perspective from the inside of the situation. To people inside, neither the outcome nor the entirety of surrounding circumstances was known. Time to decide and act was pressurized. Resources (to think, to ask around, to reflect) were limited. These people contributed to the direction of the sequence of events on the basis of what they knew and saw on the inside of the unfolding situation then, not on the basis of what the expert knows now, looking back from the outside. Any human factors expert who takes the retrospective position and pretends to explain people’s behavior from there is not a real human factors expert. However, although critical for determining whether standards of care were met, it is very difficult to attain the insider perspective. The mechanisms by which hindsight operates are powerful and mutually reinforcing. Together they continually pull in the direction of the position of the retrospective outsider. The ways in which even human factors experts retrieve performance evidence from the rubble of an accident, represent it, and retell it typically sponsor this migration of viewpoint. I discuss three of those mechanisms here so that the various stakeholders can be aware of how they operate.


151

Making Tangled Histories Linear by Cherry-Picking and Regrouping Evidence One effect of hindsight is that “people who know the outcome of a complex prior history of tangled, indeterminate events, remember that history as being much more determinant, leading ‘inevitably’ to the outcome they already knew” (Weick, 1995, p. 28). Hindsight allows us to change past indeterminacy and complexity into order, structure, and oversimplified causality (Reason, 1990). In trying to make sense of past performance, it is always tempting to group individual fragments of human performance, rear-range them linearly, and then make them point to some prima facie condition or mindset. For example, “hurry” to land is a leitmotif extracted from the evidence in the following investigation, and that haste in turn is enlisted to explain the errors that were made: Investigators were able to identify a series of errors that initiated with the flightcrew’s acceptance of the controller’s offer to land on runway 19… The CVR indicates that the decision to accept the offer to land on runway 19 was made jointly by the captain and the first officer in a 4-second exchange that began at 2136:38. The captain asked, “Would you like to shoot the one nine straight in?” The first officer responded, “Yeah, we’ll have to scramble to get down. We can do it.” This interchange followed an earlier discussion in which the captain indicated to the first officer his desire to hurry the arrival into Cali, following the delay on departure from Miami, in an apparent [attempt] to minimize the effect of the delay on the flight attendants’ rest requirements. For example, at 2126:01, he asked the first officer to “keep the speed up in the descent”…[This is] evidence of the hurried nature of the tasks performed (Aeronautica Civil, 1996, p. 29). The fragments used to build the argument of haste come from over half an hour of extended performance. This excerpt treats the record as if it were a public quarry to pick stones from, and the accident explanation as the building that needs erecting. The problem is that each fragment is meaningless outside the context that produced it; each fragment has its own story, background, and reasons for being, and when it was produced it may have had nothing to do with the other fragments with which it is now lumped. Also, behavior takes place in between the fragments. These intermediary episodes contain changes and evolutions in perceptions and assessments that separate the excised fragments not only in time, but also in meaning. Thus, the condition and the constructed linearity in the story that binds these performance fragments arise not from the circumstances that brought each of the fragments forth; it is not a feature of those circumstances. It is an artifact of hindsight. In the case described here, “hurry” is a condition identified after the fact—one that plausibly couples the start of the flight (almost 2 hours behind schedule) with its fatal ending (on a mountainside rather than an airport). “Hurry” is a retrospectively invoked leitmotif that guides the search for evidence about itself. It produces a linear, plausible story that, nevertheless, is founded on tautologies, not findings.


152

Finding What People Could Have Done to Avoid the Accident Tracing the sequence of events back from the outcome, which we already know about, we invariably come across junctures at which people had opportunities to revise their assessment of the situation but failed to do so; people were given the option to recover from their route to trouble, but did not take it. These are counterfactuals and are quite common in forensic analysis. For example, “The airplane could have overcome the windshear encounter if the pitch attitude of 15 degrees nose-up had been maintained, the thrust had been set to 1.93 EPR [engine pressure ratio] and the landing gear had been retracted on schedule” (NTSB, 1995, p. 119). Counterfactuals prove what could have happened if certain minute and often Utopian conditions had been met. However, saying what people could have done in order to prevent a particular outcome does not explain why they did what they did. Counterfactuals circumvent the human factors expert’s difficult problem: finding out why people actually did what they did. Counterfactuals are a powerful tributary to hindsight bias. They help impose structure mostly in the form of easy, binary choices. For example, people could have perfectly executed the go-around maneuver, but did not; they could have denied the runway change, but did not. As the sequence of events rewinds from its outcome, the story of missed opportunities and failed perceptions and wrong decisions rolls into time in reverse: all these options to do the right thing and they were ignored! Attorneys have a field day. How could they have been so stupid…so ignorant…so negligent? The role of the human factors expert here is to counsel caution: human work in complex, dynamic worlds is seldom about simple dichotomous choices (as in: to err or not to err). Bifurcations are rare—especially those that yield clear previews of the respective outcomes at each end. In reality, people’s choice moments (such as there are) typically offer multiple possible pathways that stretch out, like cracks in a window, into the ever-denser fog of futures that were not yet known. Their outcomes are indeterminate, hidden in what is still to come. In reality, actions need to be taken under uncertainty and under the pressure of limited time and other resources. What from the retrospective outside may look like a discrete, leisurely, two-choice opportunity not to fail is, from the inside, just one fragment caught up in a stream of surrounding actions and assessments. In fact, from the inside it may not look like a choice at all. These are often choices only in hindsight. To the people caught up in the sequence of events, there was perhaps not any compelling reason to reassess their situation or decide against anything (or else they probably would have) at the point that an attorney has now found significant or controversial. They were likely doing what they were doing because they thought they were right, given their understanding of the situation and their pressures. The challenge for the human factors expert becomes to understand how this may not have been a discrete event to the people whose actions are under investigation. The expert needs to make clear how other people’s “decisions” to continue were likely nothing more than continuous behavior—reinforced by their current understanding of the situation, confirmed by the cues on which they were focusing, and reaffirmed by their expectations of how things would develop.


153

Judging People for What They Did Not Do but Should Have Done When counterfactuals are used, even as explanatory proxy, they often require explanations as well. After all, if an exit from the route to trouble stands out so clearly, how was it possible for other people to miss it? If there was an opportunity to recover, not to crash, kill, or die, then failing to grab that opportunity demands an explanation. The place where attorneys often look is the set of rules, professional standards, and available data that surrounded people’s operation at the time, and how people did not see or meet that which they should have seen or met. Mismatches between what the rules said and people did or available data that were not noticed are easily chalked up as negligence. For the human factors expert, it is important to recognize that pointing to such mismatches is merely judging people for not doing what they (in perfect hindsight) could or should have done. It does not explain anything. Procedures Were Not Followed When fragments of behavior are contrasted with written guidance found to have been applicable in hindsight, actual performance is often found wanting; it does not live up to procedures or regulations. For example, “One of the pilots…executed [a computer entry] without having verified that it was the correct selection and without having first obtained approval of the other pilot, contrary to procedures” (Aeronautica Civil, 1996; p. 31). Attorneys invest considerably in organizational archeology so that they can construct the regulatory or procedural framework within which the operations took place or should have taken place. Inconsistencies between existing procedures or regulations and actual behavior are easy to expose when organizational records are excavated after the fact and rules uncovered that would have fit a particular situation. This is not, however, very informative. There is virtually always a mismatch between actual behavior and written guidance that can be located in hindsight (Suchman, 1987; Woods et al., 1994). Pointing out that there is a mismatch sheds little light on the why of the behavior in question. For that matter, mismatches between procedures and practice are not unique to mishaps (Degani and Wiener, 1991), and some of the most dangerous, yet safest, work is carried out entirely without procedures (Rochlin, 1999). The role of the human factors expert is not to parrot that procedural guidance was not adhered to because this, again, explains nothing. The challenge is to elucidate how a procedure is never the job: that situated work requires the skillful application and sometimes adaptation of written guidance, which is substantive cognitive work. Available Data Were Not Noticed Another route to constructing a world against which attorneys hold individual performance fragments is to find all the cues in a situation that were not picked up by the practitioners, but that, in hindsight, proved critical. Take the turn toward the mountains on the left that was made just before an aircraft collided with one of them (Aeronautica Civil, 1996). What should the crew have seen in order to notice the turn? They had plenty of indications:


154

Indications that the airplane was in a left turn would have included the following: the EHSI [electronic horizontal situation indicator] map display (if selected) with a curved path leading away from the intended direction of flight; the EHSI VOR display, with the GDI [course deviation indicator] displaced to the right, indicating the airplane was left of the direct Cali VOR course, the EaDI indicating approximately 16 degrees of bank, and all heading indicators moving to the right. Additionally the crew may have tuned Rozo in the ADF and may have had bearing pointer information to Rozo NDB on the RMDI (Boeing, 1996, p. 13). This is a standard response after mishaps: point to the data that would have revealed the true nature of the situation. Knowledge of the “critical” data comes only with the omniscience of hindsight, of course. However, if data can be shown to have been physically available, it is assumed that this information should have been picked up by the practitioners in the situation. The problem is that pointing out that it should have does not explain why it was not, or why it was interpreted differently back then (Weick, 1995). The role of the human factors expert here is to explain that in complex, dynamic work, cues and indications about the true nature of the situation emerged over time, in the midst of other, competing cues, tasks, priorities, and pressures. Attorneys can be fond of presenting a shopping bag full of epiphanies (“Look at all these indications! How could they have missed all of this?”), but this is far from how cues and indications reached people inside the situation at the time. Indeed, there is a dissociation between data availability and data observability (Woods et al., 1994)—between what can be shown to have been physically available and what would have been observable given the multiple interleaving tasks, goals, attentional focus, interests, and, as Vaughan (1996) shows, culture of the practitioner. The role of the human factors expert is to explain why there is a mismatch between data availability and data observability: how it would have made sense for people to see what they saw and to interpret it the way that they did. Professional Standards Were Not Met There are also less obvious or not-documented standards. These are often invoked when a controversial fragment (e.g., a decision to accept a runway change [Aeronautica Civil, 1996] or the decision to go around or not [NTSB; 1995]) knows no clear preordained guidance but relies on local, situated judgment. For these cases there are always “standards of good practice” based on convention and putatively practiced across an entire industry. One such standard in aviation is “good airmanship,” which will explain the variance in behavior not yet accounted for, if nothing else can. Judging Instead of Explaining Performance It is always easier to judge than to explain performance, even for human factors experts. A number of recent labels make it even easier to judge. They appear to be explanations of human error, but are just reinventions of it, stating what people should have done but did not do, or should not have done but did do, for example: • Loss of crew resource management (CRM)—judging people for failing to invest in common ground, to share data that proved operationally significant in hindsight


155

• Loss of situation awareness—judging people for their failure to notice things that, in hindsight, turned out to be critical. Loss of situation awareness says only that you now know more about the situation than the people who were caught up in it at the time. This explains nothing, but only judges them for not knowing what you know now. • Complacency—a judgment of people’s failure to recognize the gravity of a situation as you now know it, or to follow procedures or standards of good practice that you match in hindsight with their circumstances Judgments, by whatever (modern) label, frame people’s past assessments and actions in a world that is invoked or built retrospectively. The problem is that this after-the-fact world has very little relevance to the actual world that produced the behavior in question. The behavior is contrasted against a retrospective, complete reality, not the unfolding, incomplete reality surrounding (and producing) the behavior at the time. Judging people for what they did not do relative to some rule or standard does not explain why they did what they did. Saying that people failed to take this or that pathway (only in hindsight the right one) judges other people from a position of outcome knowledge that they did not have. It does not explain a thing; it does not shed any light on why people did what they did given their surrounding circumstances. The professional commitment of the human factors expert is not to judge behavior (that should be left to others), but to explain behavior. Human factors experts should understand that human error by whatever label (“loss of SA”) is not an explanation. Instead, human error demands an explanation. Resisting the urge to judge instead of explain can be challenging. Most people, in order to explain failure, will seek failure. In order to explain dead bodies on a hillside (failure), they want to lay out the prior missed opportunities and bad choices (failures). In order to explain those, they in turn seek flawed analyses, inaccurate perceptions, violated rules, or negligence (failures), even if none of these was influential or obvious or flawed or even true at the time (Starbuck and Milliken, 1988). This search for people’s failures is another well-documented effect of the hindsight bias: knowledge of outcome fundamentally influences how people see a process. If people know the outcome was bad, they can no longer objectively look at the behavior leading up to it—it must also have been bad (Fischoff, 1975; Woods et al., 1994; Reason, 1997). Given this, the human factors expert must recognize that accuracy or, really, truthfulness is not a criterion for a successful legal forensic explanation; however, plausibility is, even if, in the words of Weick (1995), it all makes “lousy history.” Local Rationality: Nobody Comes to Work to Do a Bad Job What is striking about many accidents in complex systems is that people were doing exactly the sorts of things they would usually be doing—the things that usually lead to success and safety. Mishaps are more typically the result of everyday influences on everyday decision making than they are isolated cases of erratic individuals behaving unrepresentatively (e.g., Perrow, 1984; Woods et al., 1994; Reason, 1997; Amalberti, 2001). People are doing what makes sense given the situational indications, operational pressures, and organizational norms existing at the time. Accidents are seldom preceded by bizarre behavior. Because of the hindsight bias, this is very difficult to accept for most people. It is frightening, too. Rather than changing their beliefs about the system in which the


156

accident took place (“Hey, it can actually happen and it is not that abnormal”), people will change their beliefs about the players who took part in creating the accident (“Look at how bad, how uncharacteristically negligent these operators were”). They will often embrace an account that singles out limited groups of people (departments, ethnic clans) or individuals (e.g., operators) for their responsibility because it is easier to grasp. This is called the fundamental surprise error (Lanir, 1986), which limits the need for any systemic doubt and changes the accident into a merely local “hiccup” (due to erratic, nonrepresentative individuals) that temporarily ruffled an otherwise smooth operation. The reassurance is that the system is basically safe and only some people or other parts in it are unreliable. In the end, it is not often that an existing view of a system gives in to the reality of failure. Instead, the event or the players in it are changed to fit existing assumptions and beliefs about the system, rather than the other way around. Good attorneys know this and capitalize on it. In reality, however, human errors are seldom just about the people who committed them. People’s errors and mistakes (such as can be found in any objective sense) are systematically coupled to their circumstances, tools, and tasks. Indeed, a most important empirical regularity of human factors research since the mid-1940s is the local rationality principle. What people do makes sense to them at the time; it must or they would not do it. People do not come to work to do a bad job; they are not out on crashing cars or airplanes or grounding ships. The local rationality principle says that people do things that are reasonable, or rational, based on their limited knowledge, goals, and understanding of the situation and their limited resources at the time (Woods et al., 1994). Avoiding the mechanisms of the hindsight bias means acknowledging that failures are baked into the nature of people’s work and organization and are symptoms of deeper trouble or by-products of systemic brittleness in the way business is done. It means needing to find out why the things people did at the time actually made sense, given the situation that surrounded them. Reconstructing Behavioral Sequences To understand what went on in people’s minds, we first must understand the situation in which those minds found operated. Basic findings from cognitive science and related fields keep stressing how human performance is fundamentally embedded in, and systematically connected to, the situation in which it takes place (Neisser, 1976; Gibson, 1979; Winograd and Flores, 1987; Varela et al., 1995; Clark, 1997). Human actions and assessments can be described meaningfully only in reference to the world in which they are made; they cannot be understood without intimately linking them to details of the context that produced and accompanied them (Orasanu and Connolly, 1993; Woods et al., 1994; Hutchins, 1995; Klein, 1998). Some human factors experts favor traditional information-processing approaches for explaining people’s performance difficulties. To understand what went on in the mind, they will try to look into the mind. For example, they will assert that particular data were not noticed because of an overloaded working memory (which can only contain seven plus or minus two chunks of information). This approach explains performance shortcomings by reference to constraints imposed by hypothesized internal mental mechanisms (like working memory). The problem is that such assertions are difficult to


157

verify because we still do not know whether something like a working memory actually exists and, if it exists, how it really works. Looking into the mind is singularly difficult. The alternative approach, often called the ecological approach, tries to understand performance problems by reference to constraints imposed by the world on people’s goaldirected behavior. The commitment of the ecological approach is to look first at the evolving situation in which the mind found itself and then reconstruct how features of that situation shaped assessments and actions. There is a lot of merit in human factors experts taking this approach also for credibility reasons. Past situations can be objectively reconstructed to a great extent and are often documented in detail; there are tight and systematic connections between situations and behavior—between what people did and what happened in the world around them. These connections between situations and behavior work both ways. People change the situation by doing what they do, by managing their processes. However, the evolving situation also changes people’s behavior. An evolving situation provides changing and new evidence; it updates people’s understanding, presents more difficulties, and forecloses or opens pathways to recovery. Human factors experts can uncover the connections between situation and behavior; investigate, document, and describe them; and represent them graphically. Other people can look at the reconstructed situation and how they related it to the behavior that took place inside it. Other people can actually trace the explanations and conclusions. Starting with the situation brings a human error investigation out into the open. It does not rely on hidden psychological structures or processes, but instead allows verification and debate by those who understand the domain. When a human error investigation starts with the situation, it sponsors its own credibility. A large part of human factors analysis, then, is not at all about the human behind the error. It is about the situation in which the human was working, the tasks he or she was carrying out, and the tools that were used. It is about reestablishing the connections between: • How the process, the situation, changed over time • How people’s assessments and actions evolved in parallel with their changing situation • How features of people’s tools and tasks and their organizational and operational environment influenced their assessments and actions inside that situation This is what the following five steps discuss. Step 1: Lay Out the Sequence of Events in Context-Specific Language The record and other data about an incident typically reveal a sequence of activities— human observations, actions, assessments, and decisions—as well as changes in the state of the process or system. The goal is to examine how people’s mindsets unfolded in parallel with the situation evolving around them, and how people, in turn, helped influence the course of events. Cues and indications from the world influence people’s situation assessments, which in turn inform their actions; these, in turn, change the world and what it reveals about itself, and so forth. This means that if certain actions or assessments are difficult to interpret, the circumstances (and particularly what was observable about them) in which they appeared can hold the key to their sensibility. Indeed, the reconstruction of mindset often begins not with the mind, but with the situation in which the mind found itself.


158

Similarly, if there is a lack of data from system or process sources, certain behaviors that are canonical in particular process states can help reconstruct the state of cues and indications observable at the time. There are various entries to scour the record for events and activities: • Shifts in behavior. There can be points at which people may have realized that the situation was different from what they believed it to be previously. One can see this in their remarks or their actions. These shifts are markers where later one will want to look for the indications unfolding around them that people may have used to come to a different realization. • Actions to influence the process. These may come from people’s own intentions. Depending on the kinds of data that the domain records or provides, evidence for these actions may not be found in the actions, but in process changes that follow from them. As a clue for a later step, such actions also form a nice little window on people’s understanding of the situation at that time. • Changes in the process. Any significant change in the process that people manage must serve as evidence. Not all changes in a process managed by people actually come from people. In fact, increasing automation in a variety of workplaces has led to the potential for autonomous process changes almost everywhere, for example: • Automatic shutdown sequences or other interventions • Alarms that go off because a parameter crossed a threshold • Uncommanded mode changes • Autonomous recovery from undesirable states or configurations However, even if they are autonomous, these process changes do not happen in a vacuum. They always point to human behavior around them, behavior that preceded it, and behavior that followed it. People may have helped to get the process into a configuration in which autonomous changes were triggered. When changes happen, people may or may not notice or respond to them. Such actions, or the lack of them, again give a strong clue about people’s knowledge and current understanding. The way to capture these events and activities during this stage is in context-specific language, which means a minimum of psychological diction; instead, a version of what happened in terms that domain people use to talk about their own work is created. The goal is to miss as few details as possible. Skipping to higher-level descriptions of human performance is seductive, even at this stage, but should be avoided. Seemingly low-level concepts, such as decision making or diagnosis, already are large (meaning they contain a lot of behavior) and are easily mistaken for detailed insight into psychological issues (Woods, 1993; Hollnagel, 1998). Time (and/or space) can be a powerful organizing principle to help lay out the activities and events. Behavior, and the process in which it took place, unfolded over time and, probably, in some space. By organizing data spatially and temporally (e.g., through drawing maps or timelines or both), actions and assessments can become more clearly coupled to the process state and location in which they took place; they can recover their spot in the flow of events of which they were part and that helped bring them forth. Such


159

organization likely yields further clues about why actions and assessments made sense to people back there and then. Step 2: Divide the Sequence of Events into Episodes, If Necessary Accidents do not simply happen; they evolve over a period of time. Sometimes this time may be long and, when it is, it may be fruitful to divide the sequence of events into separate episodes that each deserve their own further human performance analysis. Cues about where to “chunk up” the sequence of events can mostly come from the domain description arrived at earlier, especially at discontinuities in human assessments or actions or process states. There is, of course, inherent difficulty in deciding what counts as the overall beginning of a sequence of events (especially the beginning; the end often speaks for itself). Because, philosophically, there is no such thing as a root cause, there is technically no such thing as the beginning of a mishap. Yet the investigation needs to start somewhere. Making clear where it starts and explaining this choice is a good step toward a structured, well-engineered human performance investigation. Here is one option: take as the beginning of the first episode the first assessment, decision, or action by people or the system close to the mishap—the one that, according to you, set the sequence of events in motion. This assessment or action can be seen as a trigger for the events that unfold from there. Of course, the trigger has a reason, a background, that extends back beyond the mishap sequence in time and in place. The whole point of taking a proximal action or event as the starting point is not to ignore this background, but to identify concrete points to begin the investigation into it. Step 3: Discover How the World Looked or Changed during Each Episode This step is about reconstructing the unfolding world that people inhabited: find out what their process was doing and the data available. This is the first step toward coupling behavior and situation—toward putting the observed behavior back into the situation that produced and accompanied it. Laying out how some of the critical parameters changed over time is nothing new to investigations. Many accident report appendices contain read-outs from data recorders that show the graphs of known and relevant process parameters. However, building these pictures is often the point at which investigations stop today. Tentative references about connections between known parameters and people’s assessments and actions are sometimes made, but never in a systematic or graphic way. The point here is to marry all the events that have been identified with the unfolding process—to begin to see the two in parallel as an inextricable, causal dance-adeux. The point of this step is to build a picture that shows these connections. The record will most likely contain some kind of data about how process parameters were changing over time (e.g., speed until impact, but also traces of changing pressures, ratios, settings, quantities, automation or computer modes, rates, and so forth) and how these were presented to the people in question. Considerable domain knowledge (from the human factors expert or from outside) may be necessary to determine which of the parameters could have counted as a stimulus for the behavior under investigation. The


160

difficulty (reflected in the next step) will be to move from merely showing that certain data were physically available to arguing which of these data were actually observable and made a difference in people’s assessments and actions and why this made sense to them back then. Step 4: Identify People’s Goals, Focus of Attention, and Knowledge Active at the Time What, out of all the data available, did people actually see and how did they interpret it? Given that human behavior is goal directed and governed by knowledge activated in situ, clues are available from looking at people’s goals at the time, and at the knowledge activated to help pursue them. Finding the goals on which people were working does not need to be difficult. It often connects directly to how the process was unfolding around them: • What was canonical or normal at this time in the operation? Tasks (and the goals they represent) relate in systematic ways to stages in a process. • What was happening in the process managed by the people? Systems were set or inputs were made—changes that connect to the tasks people were carrying out. • What were other people in the operating environment doing? People who work together on common goals often divide the necessary tasks among them in predictable or complementary ways. There may be standard role divisions, for example, between pilot flying and pilot not flying, that specify the work for each. It is seldom the case, however, that only one goal governs what people do. Most complex work is characterized by multiple goals, all of which are active or must be pursued at the same time (on-time performance and safety, for example). Depending on the circumstances, some of these goals may be at odds with one another, producing goal conflicts. Any analysis of human performance must take the potential for goal conflicts into account. Goal trade-offs can be generated by the nature of the work. For example, anesthesiologists need to maximize preoperative workup time with a patient to guard patient safety and quell liability concerns, while their schedules interlock with other professions that exercise pressure with respect to timing. Goal conflicts can also precipitate from the organizational level. In this case, not all goals (or their respective priorities) are written down in guidance or procedures or job descriptions. In fact, most are probably not. This makes it difficult to trace or prove their contribution to particular assessments or actions. However, previous occurrences in similar circumstances or in the same organization may yield powerful clues. They can substantially influence people’s criterion setting with respect to a goal conflict. For example, a decision to take off or not to take off in bad weather may be precluded by earlier incidents or, conversely, encouraged by organizational reactions to lack of on-time performance. When it comes to knowledge, not all knowledge that people once possessed is necessarily available when called for. In fact, the problem of knowledge organization (is it structured so that it can be applied effectively in operational circumstances?) and inert knowledge (even if it is there, does it get activated in context?) should attune human factors experts to mismatches between how knowledge was acquired and how it is to be


161

applied in practice. For example, if material is learned in neat chunks and static ways (books and most computer-based training) but needs to be applied in dynamic situations that call for novel and complex combinations, inert knowledge is a risk (Woods et al., 1994). What people know and what they try to accomplish jointly determine where they will look and direct their attention and, consequently, which data will be observable to them (Neisser, 1976). Recognize that this is, once again, the local rationality principle. People are not unlimited cognitive processors (there are no unlimited cognitive processors in the universe). People do not know and see everything all the time, so their rationality is limited, or bounded. What people do, where they focus, and how they interpret cues make sense from their point of view: their knowledge, objectives, and limited resources (e.g., time, processing capacity, workload). Re-establishing people’s local rationality will help one to understand the gap between data availability and what people actually saw or used. In dynamic situations, people direct their attention as a joint result of: • Their current understanding of the situation, which in turn is determined partly by their knowledge and goals. Current understanding helps people form expectations about what should happen next (as a result of their own actions or as a result of changes in the world). • What happens in the world. Particularly salient or intrusive cues will draw attention even if they fall outside people’s current interpretation of what is going on. Keeping up with a dynamic world, in which situations evolve and change, is a demanding part of much operational work and implies two different kinds of “errors.” People may fall behind during rapidly changing conditions and update their interpretation of what is happening constantly, trying to follow every little change in the world. Conversely, people may become locked in one interpretation, even while evidence around them suggests that the situation has changed (see De Keyser and Woods, 1990). Step 5: Step Up to a Conceptual Description The goal here is to build an account of human performance that runs parallel to the one created in the first step. This time, however, the language that describes the same sequence of events is not one of domain terms; it is one of human factors or psychological concepts. One reason for the importance of this step perhaps goes beyond the mandate of an individual investigation. Getting away from the context-specific details—in a language that may not communicate well with other context-specific sequences of events—opens a crucial way to learn from failure: discovering similarities between seemingly disparate events. Instead, if people stress the differences between sequences of events, learning anything of value beyond the one event becomes difficult (Rochlin, 1999). Similarities between accounts of different occurrences can point to common conditions that helped produce the problem under investigation. The Risk of Pseudoscience in Human Factors Testimony The challenge for the human factors expert is to build up an account that moves from the context-specific to the concept-dependent gradually, leaving a clear trace for others to


162

follow, verify, and debate. However, given that the audience of a human factors expert often consists of lay people (e.g., juries and judges), there is a huge risk of human factors pseudoscience slipping into the testimony. Pseudoscience here refers to human factors experts drawing grand conclusions (that sound good) without making a strong theorybased or analytical connection between those conclusions and the actual data about which they speak. In pseudoscience, there is an empty space between the context-specific data and the conceptual description of those data—between the modeled and the model. For example, one cockpit voice recording captured pilots asking questions about the direction of a particular waypoint. The grand conclusion is that “deficient situation awareness is evident” (Aeronautica Civil, 1996, p. 34) and that “the CRM (crew resource management) of the crew was deficient” (Aeronautica Civil, 1996, p. 47). Such “explanations” frequently make it into conclusions and statements of cause and even into human factors testimony. For example, “Aeronautica Civil determines that the probable causes of this accident were:…(3) the lack of situational awareness of the flightcrew…” (1996, p. 57). None of this explains anything. The labels used (e.g., loss of situation awareness or loss of CRM) are theory-begging folk models that do little more than parrot popular contemporary consensus between experts and nonexperts on the nature of an everyday phenomenon, for example, getting lost or confused (Hollnagel, 1998). Juries and judges can understand it, but that does not mean that it technically explains anything. Human factors is more in vogue than it was (this volume is one testimony to that), and, as a result, researchers and others have introduced, loaned, or overgeneralized concepts that try to capture critical features of individual or coordinative behavior and make for accessible lay labels, for example, workload, complacency, stress, and situation awareness. Although easily mistaken for deeper insight into human factors issues by those who do not know any better, these concepts can create more confusion than clarity. The problem is that folk models typically lack the human performance measures or probes that would be necessary to reach down into the context-specific details because they postulate no underlying psychological theory that could deliver any (Hollnagel, 1998). A credible and detailed mapping between the context-dependent (and measurable) particulars of an accident sequence and pertinent conceptual models is often lacking. The jump from accident details to conceptual conclusions is typically a single and large one, which immunizes it against critique or verification. For example, when one pilot asks the other, “Where are we?” (Aeronautica Civil, 1996, p. 33), this may be a clear instance of a loss of situation awareness to a lay observer. However, there is nothing in models of situation awareness (e.g., Endsley, 1999) that dictates that this would be so. The model proposes no performance measurement based on an underlying psychological theory, i.e., that asking a question involving direction indicates a loss of situation awareness. This is similar to the issue faced by psychological field studies, in which the translation from context to concept needs to go through various steps or levels in order for those without access to the original situation to trace and appreciate the conceptual description (Woods, 1993). In other words, by drawing such a grand conclusion in one grand leap, the human factors expert leaves no trace for others to critique or follow. It is unverifiable and nonanalytic and, by definition, pseudoscience.


163

The steps presented earlier are one way to establish a more verifiable mapping between the particulars of context-specific behavior in an accident episode on the one hand and models or descriptions of human performance on the other. Starting with the situation and how it unfolded in parallel with people’s assessments and actions is a good way to leave a trace and a credible human factors analysis (see Woods, 1993; Klein, 1998; and Dekker, 2002, for more achievement in this respect). To be sure, any explanation of past performance that the human factors expert arrives at remains a fictional story, an approximation or tentative match open to revision as new evidence may come in or as other interpretations cover and explain more of the data. In the words of Woods (1993, p. 238): A critical factor is identifying and resolving all anomalies in a potential interpretation. We have more confidence in, or are more willing to pretend that, the story may in fact have some relation to reality if all currently known data about the sequence of events and background are coherently accounted for by the reconstruction. Explaining vs. Excusing Human Behavior To some attorneys or opposing experts, the human factors work described here may seem like a ploy to excuse defective operators, similar to an apologist scheme to let erratic people off the hook and allow them to plod on as uncorrected unreliable elements in an otherwise safe system. Human factors experts, however, should aim to help explain the riddle of puzzling human performance, not to explain it away. There is a difference between explaining and excusing human performance. Human factors expertise is about explaining behavior, not excusing it. The latter is not within its purview, but rather is something for legal proceedings to decide that hinges on the norms and laws that govern practice and is shaped by reactions to failure within the organization or profession. The commitment of a human factors expert is neither to excuse nor to judge people for what they did—that is the job of the legal people who hire the expert—but rather to explain why people did what they did. The bottom line is this: people’s actions and assessments must make sense when viewed from their position inside the situation. If, despite the human factors expert’s best efforts, people’s actions and assessments still make no sense, then human factors is probably not the field that needs to be engaged for guidance on how to explain their behavior. Attorneys may need to go to psychiatry or clinical psychology if they really cannot make sense out of certain decisions or actions. Those fields could help them explain behavior on the basis of suicidal tendencies or sabotage, for example. As long as human performance can be made to make sense using human factors concepts (and most performances in complex dynamic worlds can be), attorneys should be safe with a human factors expert. Human factors experts should push on performance until it makes sense, because it probably will.


164

8.4 Example: Two Perspectives on an Accident Sequence Taking the perspective of a person caught up inside a situation is difficult technically and because of the hindsight bias. It is easier to retain the position of retrospective outsider and judge people for what they did or did not do; however, that explains nothing. The example here shows the tension between the two perspectives, even within one account of one accident by one set of authors (Rodgers et al., 2000). The two perspectives struggle into view alternately; the accident account oscillates between the normative/ judgmental and the insightful. Here the two perspectives will be reviewed for their contribution to understanding the grounds for people’s performance in this particular situation. As expected, the retrospective outsider position contributes little or nothing to understanding; it only judges from a rationalist point of view. In contrast, the insider perspective (to the extent that it is reconstructed) can carry the interpretative load. The accident occurred to a DC-9 airliner in 1994 (NTSB, 1995). It crashed next to the Charlotte, North Carolina, airport during a severe thunderstorm. While trying to break off its approach and go around, the crew got caught in a severe downdraft and microburst (a huge parcel of air coming down and exploding in all directions when it hits the ground that is akin to a water balloon dropping on the floor and bursting open, except for no balloon and more air than water). The microburst pushed the aircraft toward the ground and slowed it down below safe flying speed. This is a typical accident for a human factors expert in the sense that, in hindsight, there appeared to be two possible interpretations of the situation. In one, there was going to be a microburst; in the other there was not. The crew chose the wrong one and people wondered how they could have missed all the critical cues that were, after all, available to them. Rodgers et al. write how they aimed at “understanding the crew’s attempt to continue the flight into the adverse weather because no rational pilot would deliberately endanger the safety of the flight by attempting to traverse such weather” (2000, p. 104). This is the local rationality principle: the starting point for trying to make sense of people’s behavior. Crashing is generally not deliberate; therefore, the behavior must have made sense to people at the time. The point is not to find out what people missed, but to understand why it made sense for them to continue based on their incomplete, continually unfolding interpretation of the situation at the time. Indeed, continuing did make sense; Charlotte Air Traffic Control (ATC) had told the crew to expect a visual approach and, later, that there was only some rain south of the field. It indicated that the weather was basically good. Not much later, however, ATC changed the crew’s approach into an instrument one (used in clouds or bad visibility). Rodgers and colleagues’ perspective is pulled outside the sequence of events, looking back onto the total situation and its outcome when they say, “There was no indication, either from their conversations on the CVR or with ATC, that they understood its implications” (2000, p. 105). These “implications,” however, would only become clear with the full benefit of hindsight, with the rubble of the aircraft scattered next to the airport. In reality, nobody (including ATC) knew that a microburst was going to hit or that it would be so severe. Crews make instrument approaches all the time; it is no indication of abnormality. Rain and thunderstorms occur in the South (and elsewhere) all the time, especially in the summer; they are not abnormal. Thus, the judgment that “implications” were not understood is made from the outside and from hindsight. The crew’s understanding (or


165

lack thereof) is contrasted against an after-the-fact world that is rich, filled with data, completely certain, and with an outcome about which we now know. However, this explains nothing about why the crew did what they did on the inside of their uncertain, incomplete, and unfolding world at the time, for which they did not yet have an outcome. Rodgers et al. then continue from the normative, judgmental stance and indicate that “although they addressed the possibility of a go-around, the CVR conversation described their attempt to avoid the storm cell…rather than a formal review of the missed approach procedure…as was required” (2000, p. 105). Actual behavior (attempts at storm avoidance) is contrasted against “required” written guidance (formal missed-approach procedure). No explanation is given about why the crew may have wanted to adapt their practice in the face of the conditions ahead or why the formal plan may in fact have been a bad idea at the time. (As Chapman [1987, p. 91] put it, “No plan survives contact with the enemy”) All that is said is that there was a mismatch between procedures and practice. This explains nothing. Yet, then there is another shift, away from the normative outsider perspective, back into the cockpit at the time: …the crew repeatedly discussed the weather conditions during the approach to Charlotte. They observed and commented on the heavy rain over the field. They identified the storm cell on their airborne radar and attempted to locate it. They even discussed the potential presence of windshear and were prepared to execute a go-around if the weather conditions so warranted. Finally they requested the tower controller to provide them with the airport surface winds and the ride conditions experienced by the pilots of the aircraft just ahead of them…it was clear by both their statements and by their initiation of apparently routine goaround procedures that, until just prior to impact, they were unaware of the magnitude of the microburst they were encountering (2000, p. 106). Concrete crew actions are employed as probes into their possible awareness inside their situation. This is critical for reconstructing sensemaking; any claims about what other people were aware of can, of necessity, only be made from their local perspective at the time. From inside the situation, there was not going to be a microburst and certainly not one of this magnitude. Rodgers et al. do not stay inside the situation for long, though. Pulling themselves out once again and looking back onto the entire sequence of events, where it is clear that the crew should have zigged instead of zagged, they judge: The fact that the crew had attempted to traverse a severe microburst in itself demonstrates that their SA regarding the weather conditions along the final approach path was deficient and this deficiency led to their decision to continue the approach…the evidence also indicated that the crew had obtained, but did not perceive or comprehend, considerable information that would have supported an alternative assessment regarding the severity of the weather (2000, p. 106).


166

The nonperception and noncomprehension of information that can only in hindsight be construed as “considerable” is simply an outsider’s judgment of crew performance, not an explanation from within the unfolding, uncertain situation at the time. It judges people for not seeing or understanding what can be seen or understood only with full knowledge and hindsight. “Considerable information,” in addition, is that typical shopping bag full of epiphanies (“Look at all the information available to them! And they still didn’t get the message!”). It does injustice to how cues and indications about the nature of the situation emerged in reality: over time and in the midst of all kinds of other tasks and indications. The assertion is also pseudoscientific because it claims that “deficient SA” caused something (to continue an approach and subsequently to crash); however, nothing in theories of situation awareness endows it with such causal power. “Deficient SA” is merely an after-the-fact judgment about a past situation (saying we now know more than people did at the time). It cannot logically have caused anything in that situation. As Billings (1996, p. 3) put it: The most serious shortcoming of the situation awareness construct as we have thought about it to date, however, is that it’s too neat, too holistic and too seductive. We heard here that deficient SA was a causal factor in many airline accidents associated with human error. We must avoid this trap: deficient situation awareness doesn’t “cause” anything. Faulty spatial perception, diverted attention, inability to acquire data in the time available, deficient decision-making, perhaps, but not a deficient abstraction! Even if this were resolved, Rodgers and colleagues (2000) offer no criteria (nor do other authors) by which SA can be measured to have been “deficient” (other than the rubble of an aircraft next to the airport). Its deficiency is an appealing and easy judgment in hindsight, not the outcome of anything resembling scientific analysis. In addition, no underlying theory or model is presented that links deficient situation awareness to the decisions and the outcome that followed. It is all folk modeling—an appeal to smoothsounding labels that lack any deconstruction or traceable connection to the data about which they pretend to speak. Then, from the lofty perspective looking down on the errors of others and judging them for their deficiencies in modern human factors labels, Rodgers and colleagues once again descend into the situation and discover that: It is likely that the ride report from the aircraft crew just ahead was most influential in the crew’s SA regarding the weather. As a crew from the same airline as the accident crew and flying turbojet aircraft, the crew of the plane ahead was well versed in the kind of weather that would exceed the capabilities of the accident airplane and was knowledgeable on the company’s guidance regarding weather conditions its pilots could not safely traverse. Further, because they were just ahead of the accident crew and closest to the airspace conditions the accident crew was about to enter, their report contained the most timely information on weather (2000, p. 107).


167

From the point of view of the people inside the unfolding situation, who did not have knowledge of the outcome, it made sense to continue with the information at hand. Lots of information about the weather was in fact not at hand (e.g., ATC did not communicate everything that it could have; an onboard windshear warning system was momentarily deactivated because of its logic; the crew had to switch radio frequencies, just missing important announcements about the weather). Yet, in the tugging of perspectives, the retrospective outsider still wins in the end, apparently making for a quicker, cleaner, more holistic conclusion. As the hindsight bias would predict, it is failure explaining failure. Unfortunately, it explains nothing and is short on scientific credibility: “the incorrect SA of the pilots led them, in addition to continuing an approach beyond the point that it should have been discontinued, to fail to anticipate a microburst of the severity that they encountered.” This is counterfactual, judgmental, and pseudoscientific. It is counterfactual because it says the pilots continued with an approach that should have been discontinued. Saying that which should have been done but what was not does not explain why people did what they did. Furthermore, claiming that the pilots failed to anticipate the weather for what it (in hindsight) turned out to be is a judgment from outside the situation. It identifies an alternative pathway that the people inside did not take and calls it their failure. It is not an explanation from people’s point of view within. Finally, “incorrect SA” is, of course, a judgment and a pseudoscientific one at that. If human factors experts want to reconstruct sensemaking, they must avoid pseudoscience, counter-factuals, and judgments. One way to do this for the preceding case is to reconstruct a detailed timeline to see when, and in which context, the various competing cues about the equivocal nature of the situation emerged, and what these would have meant in domain semantics. This rebuilds, per second or minute, the continually evolving situation in which the crew found itself and can be the scaffolding for beginning to understand how the crew’s assessment evolved in parallel. When it comes to mapping these context-specific data into a larger human factors concept, various options, other than reverting to normative outsider judgments such as “incorrect SA,” are open to the expert. One is to use the concept of plan continuation (rather well specified and deconstructed; see Orasanu et al., in press), which describes how people continue with their original plan of action if the cues and indications in favor of the plan are consistently more compelling, numerous, and stronger than cues that (in hindsight) would have suggested the plan should be discontinued. Another way to see the Charlotte accident conceptually is as an adaptive control problem (which overlaps in some sense with the plan continuation concept), in which the situation is initially too uncertain and incomplete for effective feedforward (ambiguous cues about the weather situation, smooth rides by company aircraft in front), but then quickly becomes too dynamic for effective feedback (the wind speeds changing so rapidly that they outrun the aircraft’s performance capabilities). Here, too, using control theory, there is plenty of opportunity for detailed decomposition making verifiable links between the specifics of the data trace and the model that explains what went on inside.


168

8.5 Checklists for Reconstructing Behavioral Sequences The following three checklists remind human factors experts what to consider and do (and not do) when reconstructing behavioral sequences. The first is about rules to follow in the analysis, the second is about typical errors made in human factors analysis, and the third is about the steps necessary to reconstruct a behavioral sequence. Checklist 8.1: Rules for Human Factors Analysis of Behavior Sequences 1. Do not use the outcome of a sequence of events to assess the quality of the decisions and actions that led up to it. 2. Do not mix elements from your own reality now into the reality that surrounded people at the time. Put performance back into the circumstances that brought it forth. 3. Do not create shopping bags full of epiphanies (“Look at all this! It should have been so clear!”) because evidence about the unfolding situation did not reach people that way at the time. 4. To understand and evaluate human performance, it is necessary to understand how the situation unfolded around people at the time and to take on the view from inside that situation. From there, try to understand how actions and assessments could have made sense. 5. Recognize that consistencies, certainties, and clarities are products of hindsight and not data available to people inside the situation. That situation was most likely marked by ambiguity, uncertainty, and various pressures. 6. Remember that the point of a human error investigation is to understand why people did what they did, not to judge them for what they did not do. 7. At all times, remember the local rationality principle. People are not unlimited cognitive processors (in fact, there is not a single unlimited cognitive processor, machine or human, in the entire world). They cannot know everything at all times. What people did made sense from their point of view, their knowledge at the time, their objectives, and their limited resources. Checklist 8.2: Typical Errors Made in Human Factors Analysis of Behavioral Sequences • The counterfactual reasoning error. You make this error when you say what people could or should have done to avoid the mishap. (“If only they…”). Saying what people did not do but could have done does not explain why they did what they did. • The data availability-observability error. You make this error when you highlight the data available in the world surrounding people and wonder how they could have possibly missed them. Pointing out the data that would or could have revealed the true nature of the situation does not explain people’s interpretation of the situation at the time. For that you need to understand which data were observed or used and how and why. • The micromatching error. You make this error when you try to match fragments of people’s performance with all kinds of rules and procedures that you excavate from


169

written documentation afterward. Of course, you will find gaps where people did not follow procedures. This mismatch, however, does not at all explain why they did what they did. For that matter, it is probably not even unique to the sequence of events that you are investigating. • The cherry-picking error. You make this error when you identify an overarching condition in hindsight (“they were in a hurry”), based on the outcome, and trace back through the sequence of events to prove yourself right. This violates rule 2: leave performance in the context that brought it forth. Do not lift disconnected fragments out to prove a point you can really only make in hindsight. Checklist 8.3: Steps to Take in the Reconstruction of a Behavioral Sequence 1. Lay out the sequence of events based on the data gathered. You can use language of the domain in which the mishap occurred (a context-specific language) to structure the events and use time and space as principles along which to organize them. 2. Divide the sequence of events into episodes that you can study separately for now. Each of these episodes may fit a different human factors explanation, but you may also find that you must readjust the boundaries of your episodes later. 3. Reconstruct critical features of the situation around each of these events. What did the world look like; what was the process doing at the time? What data were available to people? 4. Identify what people were doing or trying to accomplish during each episode. Reconstruct which data were actually observable. See what goals people were pursuing, what knowledge they used, and where, as a consequence, their attention was focused. Be relentless; press on their behavior until it makes sense to you. 5. Link the details of your sequence of events to human factors concepts. In other words, build an account of the sequence of events that runs parallel to the account in step 1 but is instead cast in concept-dependent, or human factors, terms. This will help you synthesize across mishaps and understand broader patterns of failure.

Defining Terms Cherry picking—Selecting disparate performance fragments from an accident record to prove a point that can only be made in hindsight. Counterfactual reasoning—Saying what did not happen, but if it happened, it would have avoided the accident. Counterfactual reasoning does not explain why people did what they did. Ecological approach—The approach in human factors that tries to understand performance difficulties with reference to constraints imposed by the world on people’s goal-oriented behavior. Hindsight bias—Bias in looking at past performance of which one already knows the outcome. If the outcome was bad, then people’s performance must have been bad, too.


170

Local rationality principle—Principle based on cognitive science that says that people are limited cognitive processors; what they were doing had to make sense from their local perspective: their knowledge, objectives, and focus of attention at the time. Micromatching—Picking fragments from the accident record and holding them against an after-the-fact world now known to be true (e.g., procedures or available data). This does not explain why people did what they did at the time. Pseudoscience—Putting large labels on a behavioral sequence (e.g., “loss of situation awareness”) without leaving an analytic trace for others to follow or verify. Retrospective outsider—One looking back onto a sequence of events and overlooking it in its entirety (including outcome). From this perspective, it is possible only to judge people for not doing what they should have done (or vice versa). One cannot understand why people did what they did; for that, it is necessary to attain the incomplete, unfolding perspective of the people inside the situation at the time. Situated performance—Behavior that took place in a complex, dynamic situation and that can only be understood in parallel and connection with the unfolding situation. Fragments of performance considered outside that situation are robbed of their meaning. References Aeronautica Civil (1996). Aircraft accident report: controlled flight into terrain, American Airlines flight 965, Boeing 757–223, N651AA near Cali, Colombia, December 20, 1995. Bogota, Colombia: Aeronautica Civil. Amalberti, R. (2001). The paradoxes of almost totally safe transportation systems. Safety Sci. , 37(2001), 109–126. Billings, C.E. (1996). Situation awareness measurement and analysis: a commentary. In D.J.Garland and M.R.Endsley (Eds.), Experimental Analysis and Measurement of Situation Awareness , Daytona Beach, FL: Embry-Riddle Aeronautical University Press, 1–5. Boeing Commercial Airplane Group (1996). Boeing submission to the American Airlines Flight 965 Accident Investigation Board. Seattle, WA: Boeing. Chapman, G. (1987). The new generation of high-technology weapons. In D.Bellin and G.Chapman (Eds.), Computers in Battle: Will They Work? New York: Harcourt Brace Jovanovich, 61–100. Clark, A. (1997). Being There: Putting Brain, Body and World Together Again . Cambridge, MA: MIT Press. De Keyser, V. and Woods, D.D. (1990). Fixation errors: failures to revise situation assessment in dynamic and risky systems. In A.G.Colombo and A.Saiz de Bustamante (Eds.), System Reliability Assessment , The Netherlands: Kluwer Academic, 231–251. Dekker, S.W.A. (2002). The Field Guide to Human Error Investigations . Bedford U.K.: Cranfield University Press (also Aldershot, U.K.: Ashgate Publishing Co.). Degani, A. and Wiener, E.L. (1991). Philosophy, policies and procedures: the three Ps of flightdeck operations. Paper presented at the 6th International Symposium on Aviation Psychology, Columbus, OH, April. Endsley, M.R. (1999). Situation awareness in aviation systems. In D.J.Garland, J.A.Wise, and V.D.Hopkin (Eds.), Handbook of Aviation Human Factors , Mahwah, NJ: Erlbaum, 257–276. Fischoff, B. (1975). Hindsight is not foresight: the effect of outcome knowledge on judgment under uncertainty. J. Exp. Psychol: Hum. Perception Performance , 1(3), 288–299. Gibson, J.J. (1979). The Ecological Approach to Visual Perception . Boston, MA: HoughtonMifflin.


171

Hollnagel, E. (1998). Measurements and models, models and measurements: You can’t have one without the other. In Proceedings of NATO AGARD Conference , Edinburgh, U.K. Hutchins, E. (1995). Cognition in the Wild . Cambridge, MA: MIT Press. Klein, G. (1998). Sources of Power: How People Make Decisions. Cambridge, MA: MIT Press. Lanir, Z. (1986). Fundamental Surprise . Eugene, OR: Decision Research. National Transportation Safety Board (1995). Flight into terrain during missed approach. USAir Flight 1016, DC-9–31, Charlotte/Douglas International Airport, Charlotte, NC, 7/2/94 (NTSB Rep. No. AAR/95/03). Washington, D.C.: NTSB. Neisser, U. (1976). Cognition and Reality . San Francisco, CA: Freeman. Orasanu, J. and Connolly, T. (1993). The reinvention of decision making. In G.A.Klein, J.Orasanu, R. Calderwood, and C.E.Zsambok (Eds.), Decision Making in Action: Models and Methods , Norwood, NJ: Ablex, 3–20. Orasanu, J., Martin, D., and Davison, J. (in press). Sources of decision error in aviation. In G.Klein and E.Salas (Eds.), Applications of Naturalistic Decision Making , Mahwah, NJ: Lawrence Erlbaum Associates. Perrow, C. (1984). Normal Accidents . New York: Basic Books. Reason, J. (1990). Human Error . Cambridge, U.K.: Cambridge University Press. Reason, J. (1997). Managing the Risks of Organizational Accidents . Aldershot, U.K.: Ashgate. Rochlin, G.I. (1999). Safe operation as a social construct. Ergonomics , 42(11), 1549–1560. Rodgers, M.D., Mogford, R.H., and Strauch, B. (2000). Post hoc assessment of situation awareness in air traffic control incidents and major aircraft accidents. In M.R.Endsley and D.J.Garland (Eds.), Situation Awareness Analysis and Measurement , Mahwah, NJ: Lawrence Erlbaum Publishers, 73–112. Starbuck, W.H. and Milliken, R.J. (1988). Challenger, fine-tuning the odds until something breaks. J. Manage. Stud. , 25, 319–340. Suchman, L.A. (1987). Plans and Situated Actions: The Problem of Human-Machine Communication . Cambridge, U.K.: Cambridge University Press. Tuchman, B.W. (1981). Practicing History: Selected Essays . New York: Norton. Varela, F.J., Thompson, E., and Rosch, E. (1995). The Embodied Mind: Cognitive Science and Human Experience . Cambridge, MA: MIT Press. Vaughan, D. (1996). The Challenger Launch Decision . Chicago, IL: University of Chicago Press. Weick, K. (1995). Sensemaking in Organizations . London: Sage. Winograd, T. and Flores, E (1987). Understanding Computers and Cognition . Reading, MA: Addison-Wesley. Woods, D.D. (1993). Process-tracing methods for the study of cognition outside of the experimental laboratory. In G.A.Klein, J.Orasanu, R.Calderwood, and C.E.Zsambok (Eds.), Decision Making in Action: Models and Methods , Norwood, NJ: Ablex, 228–251. Woods, D.D., Johannesen, L.J., Cook, R.I., and Sarter, N.B. (1994). Behind Human Error: Cognitive Systems, Computers and Hindsight . Dayton, OH: CSERIAC.

Further Information A very concrete guide on how to conduct human error investigations and how to reconstruct situated performance is Sidney Dekker’s The Field Guide to Human Error Investigations (Cranfield University Press, Bedford, U.K., and also Ashgate Publishing Co., Aldershot, U.K., in 2002). Gary Klein’s Sources of Power: How People Make Decisions is an excellent treatise on how people make decisions in actual complex, dynamic situations (1998, MIT Press, Cambridge, MA). David Woods’ chapter on process-tracing methods for the study of cognition


172

outside the experimental laboratory was one of the first to pull together the issues associated with reconstructing situated performance (in G.A.Klein, J.Orasanu, R.Calderwood, and C.E.Zsambok [Eds.], Decision Making in Action: Models and Methods , Norwood, NJ: Ablex, 1993, 228–251).

9 Causation Issues in Workers’ Compensation Roger C.Jensen Montana Tech Francisco J.Bricio The Bricio Law Firm 0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

9.1 Introduction Ergonomics experts occasionally serve as experts in workers’ compensation (WC) cases. These cases typically involve a causation issue for disorders such as cumulative trauma disorders of the upper extremities, low back troubles, vibration-induced white finger, or heat disorders. This article provides ergonomists with an overview of various WC systems, explains legal issues most relevant to ergonomics, and provides an example. 9.2 Objectives and Scope The objective of this chapter is to explain the WC legal issues that an ergonomics expert may help resolve. Such cases involve cumulative trauma disorders, low back injuries, vibration disorders, and heat disorders. An ergonomist asked to provide expert testimony in a WC case should understand which WC system applies because there are many WC systems and each has its unique features. Of particular concern are different tests for causation—the issue most likely the focus of the ergonomist’s testimony. Thus, it is necessary to understand the specific legal requirement for an employee’s injury or disease to qualify as having been caused by employment. This chapter begins with a brief history of WC laws. A section is included to educate the ergonomist on the unique distinction made in WC systems between injury and disease. A section discusses the common causation tests used in WC systems. Following this are sections on ergonomic principles and the potential contribution of an ergonomics expert in WC cases. A case involving a heat disorder is used to illustrate the principles and the role of the ergonomist in this kind of case.

Causation issues in workers' compensation

175

9.3 Discussion of Principal Issues Historical Perspective A concise history of workers’ compensation laws is provided in an annual publication by the U.S. Chamber of Commerce (2002). Some key aspects are summarized in this chapter. Before WC laws, a worker injured on the job received nothing unless the employer felt generous. Families of dead and disabled workers had to seek charity. The charities were using much of their resources to take care of families who had no income due to the temporary or permanent loss of their breadwinner. The injured workers, or the families of deceased workers, could sue the employer. However, they had to prove the employer was negligent. The employer, of course, would contest the negligence allegations as his core defense. In addition to denying negligence, the employer could assert other defenses. The three defenses most used were known as “three witches” or “unholy trinity” because they acted as three nasty hurdles to a successful suit. The three witches were: • Contributory negligence (the employee contributed to the accident) • Fellow servant (another employee contributed to the accident) • Assumption of risk (the employee knew of the hazards and still chose to work) The employer could win by proving any one of the three defenses. The law was so stacked in favor of business that few workers could win. Stories about injustices from this law began to affect public sentiment in the early 1900s. Employers began to recognize that a change was going to take place. The charitable institutions also favored a change, reasoning that business caused the problem so business should pay for it. In 1908 the U.S. government enacted the Federal Employers’ Liability Act. It provided a system for compensating railroad employees who were injured on the job. Subsequent legislation brought merchant seamen under this same compensation system. Under these systems, injured workers must sue their employer for compensation and prove that their employer’s negligence was, at least, partially responsible for the injury. Between 1910 and 1949, all U.S states enacted a workman’s compensation law. Some states later renamed the laws “Workers’ Compensation” to be gender neutral. Canadian provinces and Australian states enacted similar compensation laws. A WC law for civilian employees of the U.S government, the Federal Employees’ Compensation Act, was enacted by the federal government. Another federal act established a compensation system for longshore and harbor workers. In these “no fault” systems, the majority of claims are handled as ordinary insurance claims. Disputed claims are handled through adversarial litigation. Today, jurisdictions differ in the specifics of their WC laws, but many commonalities exist. Lencsis (1998, pp. 14–17) provides a concise comparison of the WC systems in U.S. states, Canadian provinces, Australian states, Japan, the United Kingdom, France, and New Zealand.


176

Many jurisdictions made WC the exclusive remedy for an injured employee against his employer. Recently, some U.S. states have opened a narrow pathway around the exclusive remedy provision that allows an employee to collect WC benefits and sue his employer. An ergonomist may be asked to testify in such a case, so an understanding of the burden of proof is essential. In his suit, the employee must prove that his injury was caused by the employer knowingly engaging in some form of very bad conduct. Jurisdictions differ somewhat on the specific standard for defining very bad conduct, using terms like egregious conduct, grossly negligent conduct, or wanton disregard for employee safety. These cases are not WC cases, and this chapter does not address them. Injury or Disease WC laws were originally established to compensate workers for work-related injuries. Gradually, certain work-related diseases were brought within the scope of coverage. Today, many jurisdictions have different statutory provisions for injuries and diseases regarding the test of causation, the statutory time limit for

TABLE 9.1 Common Causes of Occupational Disease Covered by WC Exposure peculiar to work Extended exposure to airborne particulates or hazardous chemicals

Examples Byssinosis resulting from years of breathing cotton dust in a cotton mill Leukemia resulting from years of exposure to benzene Hepatitis contracted by a nurse from infected blood

Exposure to an infectious agent by a healthcare worker Exposure to a hazard having a latency Radiation technician develops cancer 6 years after period between exposure and onset of the receiving a high exposure disorder Assembly line worker diagnosed with carpal tunnel syndrome after 10 months of performing highly repetitive work

filing a claim, and the amount of benefits. Thus, determining the proper classification of an employee’s disorder is important and not always easy. The plaintiff’s attorney will certainly want to avoid a situation in which, for example, the claim is filed as a disease and his or her expert calls it an injury. The following discussion aims to help the ergonomist understand the basis for the WC distinction between an injury and a disease. The first thing to recognize is that the distinction between an injury and a disease is based on the statute. For example, the U.S. Supreme Court ruled that the hearing loss of a retired harbor worker was an injury because hearing loss is a “scheduled injury” in the Longshore and Harbor Workers’ Act (Bath Iron Works v. Director, Office of Workers’ Compensation Programs, 1993). Thus, the specific WC statute is the principal source of law for making the classification. If the statute is not crystal clear, the classification may be made by consulting case law based on the statute or similar statutes. Research into statutes and case law is a job for the attorney, not the ergonomist. If the statute and case law are unclear, expert opinion and logic may be helpful.


177

Covered injuries are traditionally those resulting from accidents. However, the definition of accident is imprecise. It clearly includes an event not expected by the employee, such as falling, being struck by something, overexertion, or contacting a very hot object. It also appears to include a single exposure to a hazard lasting anywhere from seconds to a few hours. The duration of the event does not appear to be the controlling factor. For example, the amputation of a press operator’s finger takes less than 1 second, drowning takes several minutes, and death from carbon monoxide poisoning may occur in an hour or two. The harmful effects of such “accidents” and exposures are generally classified as injuries, but the laws of each jurisdiction should be consulted. Most covered diseases are adverse medical conditions resulting from one of the exposures listed in Table 9.1. In addition to those listed, some WC laws cover cases of mental stress caused by employment. Again, the duration of exposure does not determine if the harm is a disease or an injury. The nurse who contacted hepatitis may have been in contact with the infected blood for only a few seconds, but the harm is still classified as a disease. The radiation technician who developed cancer 25 years after being exposed to a very high level of radiation for a short period would have her cancer classified as a disease under most WC systems. Ergonomists serving as experts in WC cases are advised to have the attorney explain the classification system of the applicable WC law. It is very possible that a particular disorder may be referred to as a disease or illness in the ergonomics literature, but be classified as an injury in the WC system. This may be illustrated by some familiar types of disorders. Historically, the occupational safety and health literature has referred to gradual, noise-induced hearing loss as an occupational disease. However, in the WC systems, hearing loss is classified as an injury. The rationale is that each single exposure to loud noise may result in irreversible damage to hearing. Future single exposures may add to the damage. The cumulative effect over years of exposure is the end result of many injuries. Eventually, damage becomes detectible by audiometric testing. In essence, occupational hearing loss is the result of many single exposures, each causing irreversible injury to the hearing mechanisms. Categorizing vibration-induced white finger is challenging. On the one hand, it is similar to noise-induced hearing loss in that the disorder gradually builds up as a result of many single exposures to the stressors. On the other hand, the disorder differs from noiseinduced hearing loss in that the damage caused by single exposures, at least during the early stages of the disease progression, is reversible. Another difference is that vibrationinduced white finger has a latency period between exposure and onset. This latter characteristic favors classification as a disease. Occupational carpal tunnel syndrome also seems to fit better into the disease category. It typically develops from extended repetition of stressful exertions over time. There is a latency period during which resting the affected tendons will allow their return to normal health. There is also a point at which the disorder is no longer reversible by rest alone. Therefore, these two disorders would seem to belong in the WC classification of disease, but the laws of each jurisdiction should be consulted. Low-back sprains/strains are classified as an injury in WC systems. At first blush, this might seem in conflict with biomechanical theories of cumulative low-back deterioration resulting from repeated biomechanical stressors. The rationale behind the WC


178

classification is that, even if the individual’s low back may have deteriorated over the years to a point at which it could handle little biomechanical stress, the incident that triggered the pain was a single event, which is typical of an accident, as defined in WC laws. Under these laws, it does not matter if the employee’s back was sturdy or fragile immediately before the incident. Also deserving some discussion are disorders resulting from exposure to extremes of heat or cold. Cases of hypothermia and frostbite result from exposure to extremes of cold. The time from onset of exposure to the disorder may range from less than 1 minute (in the case of frostbite from touching a very cold piece of metal) to several hours of being in a cold environment with inadequate clothing. Cases of heat exhaustion and heat stroke result from exposure to excessive heat from environmental and metabolic sources. The time from initial exposure to onset of the disorder depends on the rate of heat transfer and the individual’s acclimatization and health. A relatively short onset time would be 1 hour, and a long onset would be an entire work shift of 8 to 10 hours. In the heat stroke case discussed in this chapter, the applicable WC law was not crystal clear on whether to classify heat stroke as an injury or a disease. If the jurisdiction defines an injury as an unexpected event or a single exposure to a hazard, then these thermal stress cases fit best within the injury classification. An ergonomist consulting on a WC case should be sure to ask the attorney about the possible significance of the injury vs. disease classification. Causation in WC Systems All jurisdictions require some causal relationship between an employee’s disorder and his work. The general requirements for U.S. states are summarized next. Nearly all work-related injuries are covered. Covered injuries are traditionally those resulting from “accidents” or single exposures. Injuries resulting from assaults have generated some jurisdictional differences. Generally, an injury resulting from a workrelated assault is covered, but an injury resulting from an assault of a personal nature is not. Some work-related diseases are covered. Those most widely authorized by WC laws are noted in Table 9.1. WC laws generally exclude ordinary diseases of life (e.g., common cold, flu) and diseases that are not peculiar to or characteristic of the employee’s occupation. Some infectious diseases that anyone can get are compensible if the working conditions exposed the worker to significantly greater risk than that of other work places or public places, e.g., a nurse who contracts tuberculosis from a patient. The basic WC test of causation for an injury or disease is: did it arise out of and occur in the course of employment? Application of these tests to specific cases has generated an extensive body of case law. An ergonomist is most likely to be involved in the “arising out of” issue. This test is clearly met if the risks for the injury were distinctly associated with the employment. An example is the power press operator who amputated a finger while operating a press as part of her employment. The test is not met if the risks were entirely personal to the individual, for example, if a bank teller suffers a heart attack while working on a routine day. In many situations the risks are not exclusively tied to the work or to the individual.


179

These are sometimes referred to as neutral risks. Jurisdictions tend to follow one of three doctrines for neutral risk cases (Lencsis, 1998, pp. 35–37). • Jurisdictions that follow the actual risk doctrine find for compensation if the risk was actually part of the employment. This is the most liberal doctrine. • Jurisdictions that follow the positional risk doctrine find the injury or disease arose out of employment if the incident occurred while the employee was in a position appropriate for his job. There is no need to prove the employee was at increased risk because of the job. Another way of looking at this test is to ask whether it would not have occurred but for the conditions and obligations of employment. For example, suppose a janitor cleaning a building gets crushed when the building collapses. His case would meet this test because, had it not been for his employment, he would have been somewhere else when the building collapsed. His attorney would not need to prove that janitors are at increased risk of death by building collapse compared to people in other occupations or the general public. • Jurisdictions that follow the increased risk doctrine find the injury or disease arose out of employment if the claimant proves his risk of the injury or disease was increased by (or peculiar to) the employment. The plaintiff needs to prove that employment put him at an increased risk of the injury or disease compared to the general public. This is when testimony of an appropriate expert may be helpful. Examples include an employee who develops carpal tunnel syndrome, sciatica, or heat stroke. The expert may help identify the risk factors and determine if workplace risk factors made this individual’s risk greater than that of the general public. With this brief summary of WC systems, we turn to the principles of ergonomics that apply to certain litigated WC cases. Ergonomic Principles Ergonomists work to match task demands to mental as well as physical human capabilities (Kroemer and Grandjean, 1997). Throughout the mental-physical continuum, a distinction is made between stress and strain. Stress is some form of load on the person and may be a physical, mental, or mixed load. Strain refers to the person’s response to the stress. Ergonomists are asked to assist with WC cases when a question arises about the occupational stress level’s being sufficiently intense to overstrain an individual and/or cause a disorder. The ergonomics community has devoted considerable effort toward techniques to assess workplace stress levels and define the limits of safe levels. Thus, ergonomists are well suited to serve as experts on these issues. Thermal stress provides an excellent example. Humans function well and maintain a constant body temperature in a broad range of climatic conditions (Bernard, 1996; Bishop, 1997). Figure 9.1 illustrates the concept. At the lower end of this region are conditions referred to a cold stress, in which the human body responds by shifting blood flow to protect vital organs and by shivering to increase metabolic heat production. When the body’s ability to adapt to cold stress is exceeded, body temperature decreases. At the upper end of the region are conditions referred to as heat stress, in which the human body responds by shifting blood flow toward the skin and sweating to increase


180

evaporative cooling. In this region, when the body’s ability to adapt is exceeded, body temperature increases. In a WC case, an ergonomist might be asked to determine if the heat exposure of an individual was sufficient to cause a rise in body temperature and/or a heat stroke. The Ergonomics Expert in WC Litigation In WC litigation, the injured employee is characterized as the plaintiff. In the case of a deceased employee, the plaintiff is the spouse or other lawful heirs. The employer is the defendant. Often the case also names the employer’s WC insurance company as a defendant. The plaintiff bears the burden of introducing evidence to prove entitlement to compensation. The defendant will try to discredit the plaintiff’s evidence and may introduce evidence to refute his case. The plaintiff needs evidence to establish that an employer-employee relationship existed, the employment was within the scope of the jurisdiction’s WC law, and the injury or disease developed in the course

FIGURE 9.1 Body core temperature related to environmental heat level for a person doing light work. of employment. These elements of a WC claim are outside the scope of this chapter. For a more complete explanation suitable for nonattorneys, see Lencsis (1998). For a comprehensive treatise on the subject suitable for attorneys, see Larson and Larson (1996). The focus of this chapter is the element involving proof that the disorder arose out of employment. In jurisdictions following the increased risk doctrine, a part of this burden involves proving that the plaintiff’s risk for the injury was increased by (or peculiar to) the employment. This chapter explores this legal concept as it applies to WC claims for heat-related disorders and cumulative trauma disorders because these are types of cases, for which ergonomists are most likely to contribute. The major heat-related disorders are heat cramps, heat exhaustion, and heat stroke. Heat exhaustion is a moderately serious disorder sometimes resulting in a WC claim. Heat stroke is a very serious disorder that requires immediate medical treatment; and if it occurs at work, it will involve a WC claim.


181

Some disorders associated with cumulative trauma are carpal tunnel syndrome, hearing loss, and white-finger disease. All of these disorders may be developed on or off the job. Thus, an employee who develops one of these disorders does not automatically qualify for WC. To do so requires evidence to establish that the disorder arose out of the job. What sort of evidence can a plaintiff introduce to support his position that the disorder arose from a job? The following types of testimonial evidence should be helpful whether the disorder is classified as an injury or a disease. • What are the known personal and work-related risk factors for the disorder? • Did the work performed by the plaintiff involve exposure to the work-related risk factors? • Was the plaintiff’s exposure to the work-related risk factors of sufficient intensity and duration to cause or contribute to the disorder? • Was the plaintiff’s occupational exposure to the work-related risk factors considerably greater than that of the general public? • Do employees who perform jobs similar to the plaintiff’s tend to have higher prevalence or incidence rates of the disorder than employees in other jobs or the general public? • Were the plaintiff’s symptoms consistent with those known to be associated with the hazardous exposure according to authoritative scientific literature? A plaintiff may need more than one expert to address all these questions. The trier of fact (e.g., judge or hearing examiner) needs to determine the qualifications of the proposed expert and decide on the scope of his or her testimony based on the individual’s credentials. There may be legal precedence in the jurisdiction specifying the credentials of the expert for specific issues. For example, the North Carolina Supreme Court declared that the expertise of a medical expert is appropriate for testimony on risk factors for heat disorders (Dillingham v. Yeargin Construction Co., 1987). In some jurisdictions, a qualified ergonomist may be permitted to testify about work-related risk factors. If not, he or she can provide copies of important literature to the attorney, who can share it with a medical expert for testimony. Testimony regarding studies involving prevalence and incidence rates should be offered by a qualified individual who knows and understands the studies. These concepts are illustrated with the following example. 9.4 Example A Mexican farm laborer was working in North Carolina. He had a proper work visa and his employer had WC insurance to cover him and his fellow employees. During the middle of the summer, he was working long hours harvesting crops. On the 19th straight work day, he started work at 6:30 a.m., had a half-hour lunch break, and continued working, but in the late afternoon he was unable to continue working. His foreman had two fellow laborers take him to an area to rest. They took him to an outdoor location beside the trailers where they resided while working at the farm. Sadly, the employer had a walk-in cooler where the worker could have been placed. About an hour later, he was found completely unconscious and unresponsive by a fellow laborer. Someone called for an ambulance, which arrived 40 minutes later after getting lost on the way to the farm. It


182

took 30 minutes to get the laborer to the hospital. His diagnosis was coma resulting from exertional heat stroke, from which he never recovered. The employer reported the incident to the North Carolina Industrial Commission. The Commission appointed a guardian, who then retained an attorney to represent the worker. He filed a WC claim with the employer’s WC insurance company. The company denied the claim. When the hospital learned of the denial, they had the worker flown back to Mexico rather than continue providing charitable care. The attorney challenged the denial of the claim. The case was handled by the North Carolina Industrial Commission, Court Division. The WC insurance company took the position that heat stroke was not peculiar to a farm laborer’s job (an earlier version of the increased risk doctrine). Thus, the plaintiff’s attorney had the burden of proving that the employee’s job involved exposure to risk factors for heat disorders peculiar to his employment. After conducting a literature search, he found an article comparing risk of heat disorders for different occupational groups and different industries. The attorney contacted the author and explained the case. After some discussion, the author agreed to serve as an expert for the case. The author’s testimony was taken in a deposition, which began with testimony about his qualifications to serve as an expert for the case. After this testimony, the plaintiff’s attorney moved to tender the author as an expert in the area of occupational safety and, in particular, on the area of heat stress in work environments. Defense counsel objected but offered no basis for the objection. The plaintiff’s attorney continued with the deposition. In response to a series of questions by plaintiff and defense attorneys, the expert provided the following testimony. On the day of the incident, the weather was hot. The conditions were mostly sunny with air temperatures in the range of 32.8 to 33.9°C (91 to 93°F) all afternoon. The humidity level was moderate, about 35% relative humidity. According to the expert and both physicians, this level of heat is enough to cause a heat disorder in a worker doing physical labor for several hours in the sun. The expert mentioned a study of Nebraska WC claims finding that when the maximum daily temperature was in this range, the rate of WC claims for heat disorders was noticeably increased (Jensen, 1987). The relationship is shown in Figure 9.2. Two physicians who treated the plaintiff were deposed, and both stated the risk factors for heat-related injuries. The safety and health expert’s testimony regarding risk factors for heat disorders was only to supplement the testimony of the physicians. His testimony was that risk factors for heat stroke include personal factors and occupational factors. The expert expressed no opinions on personal risk factors. On the subject of occupational risk factors, he explained that sources of heat load are primarily metabolic heat from performing manual labor and radiant heat from the sun. The air temperature was similar to that of the skin; therefore, little heat was transferred between the plaintiff’s body and the air through convection. The principal mechanism for transferring heat from the plaintiff’s body was through evaporation of sweat from his skin. The plaintiff performed physical labor outdoors all day, beginning at 6:30 a.m. His extended exposure to a significant heat load substantially increased the risk of a heat disorder.


183

FIGURE 9.2 Relationship between daily incidence of heat disorder WC claims and population-weighted daily maximum temperature during a hot summer in Nebraska. Adapted from Jensen, 1987. Another key topic of testimony concerned comparison of different occupational groups for heat disorders. This testimony was based on a paper by the expert (Jensen, 1983). Occupational groups were compared on their annual incidence rate of workers’ compensation claims for heat-related disorders. A comparison based on occupational classification indicated that the rate for farm laborers was 13 cases per 100,000 employees, whereas all occupations combined had a rate of 2 cases per 100,000 employees. This testimony helped support the plaintiff’s burden of proving that the job of farm laborer involves greater risk of a heat disorder than that of the broad spectrum of all employees in diverse occupations. The decision in the case was for the plaintiff. As a result, testimony on the issue of damages was needed. Medical experts testified on the likelihood of the plaintiff coming out of the coma, how long he would likely live, and what types of care would be needed. This case also illustrates the no-fault character of WC laws. Here, the plaintiff suffered a serious heat disorder and the employer failed to provide prompt and proper first aid. The employer had a walk-in cooler where the plaintiff could have been placed while waiting for the ambulance. Instead of applying this basic form of first aid, the plaintiff was left lying outdoors in hot weather for over an hour. Thus, the employer was at fault for failing to provide proper care; however, that is not relevant to the issue of qualifying for WC. Remember that WC is a no-fault system. However, the lack of proper first aid did become an issue in the plaintiff’s claim for an extra 10% monetary award based on North Carolina’s WC law allowing extra compensation if the injury was caused by the employer’s violation of a statutory safety duty.


184

9.5 Checklist Table 9.2 is a checklist for use by an ergonomics expert who is asked to assist with a WC case. The checklist emphasizes the causation issues. Defining Terms Arising out of—A phrase expressing a causation-related requirement for an injury or disease to qualify for workers’ compensation. Basically, the cause of the injury or disease must be related to employment. Jurisdictions differ on rules for applying this requirement. Heat stress—A condition in which the thermal demands placed on a human (or animal) due to metabolic heat plus environmental heat exchange invoke a physiological cooling response, such as sweating. In the course of employment—A phrase expressing a causation-related requirement for a workers’ compensation claim to qualify for compensation. Part of the employee’s burden of proof is to

TABLE 9.2 Checklist of Primary Factors for the Causation Issue in a WC Case Primary factor Was employment within workers’ compensation coverage? Arose out of employment

Considerations

Check law of the jurisdiction The responsible employer is identified The risk for the injury/disease was distinctly associated with the employment. If not, which doctrine applies in the jurisdiction? Not if the risk factors for the injury/disease were entirely personal, unrelated to the employment Generally, not commuting between home and workplace unless performing a task of employment Not an intentional injury of a personal nature Occurred in the course of Occurred at the employee’s place of work, including the employment parking lot Occurred during a time when employee had a work-related reason to be there Occurred while the employee was engaged in an employment-related activity, and not during unreasonable horseplay or while on a curiosity venture Workplace stressors were related to the type of strain known Cumulative trauma disorders in jurisdictions following the increased to cause or contribute to the disorder risk doctrine Employee’s exposure to work-related risk factors was great enough in duration and intensity to cause the employee’s symptoms Time of exposure to onset of symptoms is consistent with known scientific knowledge Symptoms were consistent with the workplace stressors or exposures. Can inconsistent symptoms be explained?


185

Job involved increased risk of the disorder compared to risks faced by the general public Is there epidemiological documentation that the employee’s job involved greater risk of the CTD than most other jobs did? Heat disorders in jurisdictions Heat stress experienced by employee was enough to exceed following the increased risk doctrine ability to regulate body temperature Employee’s occupational exposure to heat stress was greater than that of the general public The job involved greater risk of a heat disorder than that of the general public and/or employees in most other jobs Is there epidemiological documentation that the employee’s job involved greater risk of a heat disorder than most other jobs?

offer evidence indicating the injury or disease occurred while he or she was engaged in employment activities rather than in personal activities. No-fault system—A phrase used to indicate that WC systems were intended to compensate injured workers without regard for fault. In a no-fault system, the employee does not need to prove the employer was negligent or otherwise at fault, and the employer cannot deny the claim because the employee was at fault. Today, a few exceptions are recognized, depending on the jurisdiction. Stress—A form of load on an object, person, or body part. It may be a physical load, mental load, or a combination load. Strain—A response to stress. Engineers and physicists define strain as a deformation in response to a physical stress. The occupational safety and health community uses the term to describe an injury or impairment of a tendon or muscle caused by overstretching, overexerting, overusing, or wrenching. The various human responses to thermal stress are referred to as strain responses. References Bath Iron Works v. Director, Office of Workers’ Compensation Programs, 506 U.S. 153 (1993). Bernard, T.E., 1996, Occupational heat stress. In Bhattacharya, A. and McGlothlin, J.D. (Eds.) Occupational Ergonomics; Theory and Applications . New York: Marcel Dekker, 195–218. Bishop, P.A., 1997, Applied physiology of thermoregulation and exposure control. In DiNardi, S.R. (Ed.) The Occupational Environment—Its Evaluation and Control Fairfax, VA: American Industrial Hygiene Association, 629–658. Dillingham v. Yeargin Construction Co., 320 N.C. 499 (1987). Jensen, R.C., 1987, Incidence of workers’ compensation claims for heat illness as a measure of the effects of a heat wave. In Asfour, S.S. (Ed.) Trends in Ergonomics/Human Factors IV . Amsterdam: North-Holland, 341–348. Jensen, R.C., 1983, Workers’ compensation claims relating to heat and cold exposure. Prof. Safety , 28(9), 19–24. Kroemer, K.H.E. and Grandjean, E., 1997, Fitting the Task to the Human: a Textbook of Occupational Ergonomics . London: Taylor & Francis, 230–231.


186

Larson, A. and Larson, L.K., 1996, Larson’s Workers’ Compensation Law . New York: Matthew Bender. Lencsis, F.M., 1998, Workers’ Compensation: A Reference and Guide . Westport, CN: Quorum Books. U.S. Chamber of Commerce, 2002,2002 Analysis of Workers’ Compensation Laws . Washington, D.C.: U.S. Chamber of Commerce.

10 Legal Issues in Work-Related Musculoskeletal Disorders: a European Perspective from the U.K. Vincent Kelly University of Surrey Jason Devereux University of Surrey 0–415–2887–3/05/$0.00+$1.50 © 2005 by CRC Press

10.1 Introduction to Human Factors Principles and Relevance to Standard of Care Ergonomics is concerned with understanding the interactions between the work system and individual capacities in order to minimize worker ill health, improve comfort and safety, and increase production. The work system is formed by the technology, organization, task, environment, and the person (Smith and Sainfort, 1989; Carayon et al., 1999). An imbalance between the work system demands and individual capacity can result in musculoskeletal disorders (MSDs). Such an outcome as a result of this imbalance has been referred to as an ergonomic injury (Pheasant, 1996). There is substantial evidence that some musculoskeletal disorders are a significant problem within the European Union in terms of a major ill-health and financial burden (Buckle and Devereux, 2002). In the U.K., a four-tier hierarchy sets the standard of care: statutory requirements, approved code of practice, guidance, and general and approved practice. For example, in relation to MSDs in the U.K., there are statutory regulations and guidance on manual handling and VDU work. Work-related upper-limb disorders are currently covered by guidance issued by the Health and Safety Executive (HSE). The ergonomic approach is central to the European Directive on manual handling (90/269/EEC) and to the U.K. Manual Handling Operations Regulations 1992 (HSE, 1998). The Health and Safety Display Screen Equipment Regulations 1992 also take into account ergonomic principles in the design, selection, and installation of display screen equipment; the design of the workplace; and the organization of the task and software particular to human data processing (HSE, 2001). The HSE guidance on upper-limb disorders in the workplace, HSG60 (HSE, 2002), states that an ergonomic approach is the most effective way of dealing with work-related upper-limb disorder problems. General and approved practice in industry is the primary guide in determining the standard of care, but it is not inflexible and it may be departed from when there is a


188

failure to take account of some danger (de Navarro et al., 2001). The prevention of ergonomic injuries is based on the application of ergonomic knowledge to work practice. An approved work practice is one that takes account of current ergonomic knowledge. According to the U.K. Health and Safety Commission (HSC, 2000), authoritative sources of good practice are prescriptive legislation, approved codes of practice, and guidance produced by government and HSE inspectors. Other sources include standards produced by standard-making organizations and guidance agreed upon by a body representing an industrial or occupational sector, provided the guidance has gained general acceptance. Consequently, ergonomic and human factors principles underpin the standard of care required in musculoskeletal disorders litigation. Ergonomics is the science that bridges the gap between the generality of the law and the specifics of each individual case (Kelly, 1998). 10.2 Objective and Scope A report by the European Agency for Safety and Health at Work (2000) highlighted that six of the fifteen member states of the European Union use legal proceedings as a means of compensating employees for MSDs, in particular upper-limb disorders. The U.K. is the member state in which litigation is most frequently used. In the U.K., claims for damages are based on the tort of negligence or on breach of statutory duty (Barrett and Howells, 1997). The definition of negligence is set out in the judgment of Alderson B in Blyth v Bermingham Water Works Company (1856): “Negligence is the omission to do something which a reasonable man, guided upon those considerations which ordinarily regulate the conduct of human affairs, would do, or doing something which a prudent and reasonable man would not do.” To succeed in a legal action in negligence, the claimant must show that: • He has been injured. • The injury was a direct consequence of risks to which he had been exposed in the course of his work. • The employer was in breach of his duty of care, that is, the claimant must prove (i) the risk to which he was exposed was reasonably foreseeable (ii) it would have been reasonably practicable to circumvent the risk (Pheasant, 1991). In order to succeed on a breach of statutory duty, the claimant must prove: • He belongs to the class of persons the statute is designed to protect. • The defendant was the person on whom the duty was imposed. • The defendant was in breach of the duty. • The breach caused the damage (Barrett and Howells, 1997). The remainder of this chapter provides a description of the legal issues concerning the tort of negligence and breach of statutory duty in relation to claims for musculoskeletal disorder compensation. These legal issues are exemplified through case law concerning work-related musculoskeletal disorders affecting the back and upper limbs. Finally,

Legal issues in work-related musculoskeletal disorders

189

criteria that need to be addressed in a legal action are provided as a checklist for use by ergonomists and other professionals involved in litigation. 10.3 The Legal System in the U.K. The English legal system is based on Common Law, which developed from the decisions of judges whose rulings over the centuries have created precedents for other courts to follow; these decisions were based on “custom and practice of the realm” (Carter and Howard, 1995). Statute Law is passed by Parliament and in recent times, in relation to safety, health, and welfare at work, originates increasingly from the European Union, has an accident prevention focus, and is rooted in scientific knowledge. An injured employee is entitled to sue the employer for damages for injury resulting from a breach of a duty at common law and a statutory duty. This has led to the emergence of a “double-barreled” action against the employer. In such cases, an injured employee sues separately, though simultaneously, for breach of both duties on the part of the employer (Stranks, 1999). When considering an action brought in negligence, a number of issues must be addressed: causation; foreseeability; knowledge; reasonable practicability; contributory negligence; vicarious liability; and the duty of the employer. Causation A claimant must show the relationship between the injury for which compensation is being claimed and the breach of duty that is said to give rise to this injury, thus establishing a causal link between a breach of duty and the injury. It is particularly important in cases of alleged work-related upper-limb disorders to relate the injury caused to the breach alleged, as opposed to the work in general. If the absence of a warning is to be relied upon as a breach, then the claimant must be able to say that the warning would have made a difference (Scott and Langstaff, 2001). Reasonable Foreseeability What can be foreseen depends on knowledge: what the defendant actually knows or what a reasonable man in his position should know (Munkman, 1990). In relation to the reasonable man, Lord McMillan in Glasgow Corporation v Muir (1943) said: “The standard of foresight of the reasonable man is in one sense an impersonal test. It eliminates the personal equation and is independent of idiosyncrasies of the particular person whose conduct is in question. The reasonable man is presumed to be free of over apprehension and over confidence.” Of course, a strict correlation exists between the extent of the duty on the one hand and negligence (the breach of duty) on the other. Both depend on the degree of risk foreseeability (Munkman, 1990). What is foreseeable depends largely on technical knowledge available at the time when the claimant was injured. Advances in technical and scientific knowledge make occurrences in the present foreseeable which would not


190

have been foreseeable in the past. Most accident investigations disclose the cause of accidents and enable others to learn from the experience (Barrett and Howells, 1997). Knowledge The issue of knowledge in a legal action is complex and important for determining what is reasonably foreseeable. In Stokes v Guest, Keen and Nettlefold (Bolts and Nuts Ltd.) [1968], Swanwick J considered the case law and said: From these authorities I deduced the principles, that the overall test is still the conduct of the reasonable and prudent employer, taking positive thought for the safety of his workers in the light of what he knows or ought to know; where there is a recognised and general practice which has been followed for a substantial period in similar circumstances without mishap, he is entitled to follow it, unless in the light of common sense or newer knowledge it is clearly bad; but, where there is developing knowledge, he must keep reasonably abreast of it and not be too slow to apply it; and where he has in fact greater than average knowledge of the risks, he may be thereby obliged to take more than the average or standard precautions. He must weigh up the risk in terms of the likelihood of injury occurring and the potential consequences if it does; and he must balance against this the probable effectiveness of the precautions that can be taken to meet it and the expense and inconvenience they involve. If he is found to have fallen below the standard to be properly expected of a reasonable and prudent employer in these respects, he is negligent. This passage of Swanwick J was cited with approval in two Court of Appeal cases, Joseph v Ministry of Defence (1980) and White v Hallbrook Precision Castings Limited [1985]. It was further endorsed by the Court of Appeal in Hayes v Pilkington Glass Limited [1998]. In general, an employer is expected to keep reasonably abreast of current knowledge concerning dangers arising in trade processes and should be acquainted with pamphlets issued by the HSE and other safety organizations drawing attention to risks that have come to light and the means of avoiding them (Wright v Dunlop Rubber Co. Ltd and ICI Ltd., 1972). The issue of warning pamphlets by the HSE may be decisive: Cartwright v G.K.N. Sankey Ltd (1973) shows that the courts expect these pamphlets to be read carefully by a person with authority and not skimmed over quickly. Depending on the size of the employer, these may well fix the date from which liability arises (Thompson v Smiths Ship Repairers Northshields Limited, 1984). In McSherry v British Telecommunications PLC [1992], it was held that an organization like British Telecom should have been aware by 1985 of the risks of upperlimb injuries from repetitive keyboard use. However, the mere mention of the possibility of a risk in an HSE publication will not necessarily be sufficient to find liability (Walker v Wabco Automotive U.K. Limited, 1999).


191

The influence of advancing knowledge and what is considered to be reasonable care is shown in the judgment in Bankstown Foundry Pty Ltd. v Braistini (1986) per Brennan and Dean J.J: Contemporary decisions about what constitutes reasonable care on the part of an employer towards an employee in the running of a modern factory are in sharp conflict with what would have been considered reasonable in the 19th century workshop and, for that matter, reflect more demanding standards than those of 20 or 30 years ago. While it is true that that has come, in part, being the consequence of the elucidation and development of legal principle, it has, to a greater extent, reflected the impact, upon decisions of fact, of increased appreciation of the likely causes of injury to the body, of the more general availability of the means and methods of avoiding such injury and of the contemporary tendency to reject the discounting of any real risk of injury to an employee in the assessment of what is reasonable in the pursuit by an employer of pecuniary profit. Changing standards in the community also have an impact on what is considered reasonable care. The judgment by Mason, Wilson, and Dawnson J.J in Bankstown Foundry Pty Ltd. v Braistini (1986) said: What reasonable care requires will vary with the advent of new methods and machines and with changing ideas of justice and increasing concern with safety in the community. This must be so, because in every case the tribunal of fact, be it a judge sitting alone or a jury must determine whether or not in the circumstances of the particular case the employer failed to take those precautions, which an employer acting reasonably could be expected to take. What is considered to be reasonable in the circumstances of a case must be influenced by current community standards. In so far as legislative requirements touching industrial safety have become more demanding upon employers, this must have its impact on community expectations of the reasonably prudent employer. Therefore, as new standards emerge and more knowledge becomes available, the employer is expected to keep up to date and take reasonable action to ensure compliance. Reasonable Practicability A statutory requirement is qualified by the phrase “so far as is practicable” when it implies that if, in the light of current knowledge and invention, it is feasible to comply with this requirement, then, irrespective of the cost or sacrifice involved, such a requirement must be complied with (Schwalb v. H.Fass & Son Ltd, 1946). “Practicable” is equivalent to physically possible and implies a higher duty of care than a duty qualified by the phrase “so far as is reasonably practicable” (Stranks, 1999). “Reasonably


192

practicable” was defined by Lord Justice Asquith in Edwards v. National Coal Board [1949] when he said: Reasonably practicable is a narrower term than “physically possible” and seems to me to imply that a computation must be made in which the quantum of risk is placed on one side of the scale and the sacrifice involved in measures necessary for averting that risk (whether in money, time or trouble) is placed on the other, and that if it be shown that there is a gross disproportion between them, the risk been insignificant in relation to the sacrifice—the employer discharges the onus which is upon them. The Approved Code of Practice (HSC, 2000) in the Management of Health and Safety at Work Regulations 1999 deals with risk assessment. In this approved code: • A hazard is something with the potential to cause harm (this can include articles, substances, plant or machines, methods of work, the working environment, and other aspects of work organization). • A risk is the likelihood of potential harm from the hazard being realized. The extent of the risk will depend on: • The likelihood of that harm occurring • The potential severity of that harm, i.e., of any resultant injury or adverse health effect • The population which might be affected by the hazard, i.e., the number of people who might be exposed Risk reflects the likelihood that harm will occur and its severity (HSC, 1998). Lord Reed said, in Morris v West Hartlepool Steam Navigation Co. Ltd. [1956]: It is the duty of an employer, considering whether some precaution should be taken against the foreseeable risk, to weigh, on one hand, the magnitude of the risk, the likelihood of an accident happening and the possible seriousness of the consequences if an accident does happen, and, on the other hand, the difficulty and the expense and any other disadvantage of taking the precaution. Consequently, in manual handling situations an understanding of what is reasonably practicable can only be arrived at on the basis of ergonomic knowledge concerning likelihood of injury, the extent of injury, and the measures necessary to avert the risk. In Bolton v Stone [1951], the decision of the House of Lords was that because of the slightness of the risk, there was no negligence in failing to take measures that would involve great sacrifice at a cost; however, Lord Reed pointed out, referring to Bolton v Stone [1951] in The Wagon Mould (2) [1967], that it is negligent to allow even a small risk to arise if it can easily be avoided. The first authoritative decision from the Court of Appeal on the Manual Handling Operations Regulations 1992 is Hawkes v London Borough of Southwark (1998). In relation to the concept of reasonable practicability, the Court of Appeal made it clear that the traditional interpretation should be given to this concept: “I believe it is proper to


193

conclude that Parliament had in mind, when they enacted the regulations, the construction of words ‘reasonably practicable’ which has been accepted by the court since 1938” (per Aldous L.J). In this case, the balancing approach was considered in the context of whether or not additional help should have been provided to the claimant, who was injured lifting a heavy door in a block of flats, as follows: In my view the risk was slight as was the sacrifice. There was no evidence to suggest that the defendants, with their workforce, could not have made available a second man to help carry the door upstairs. How many extra man hours would have been needed over a year was not clear and I do not believe it appropriate to assume, in the defendants’ favour, that they would be other than minimal in the context of the total number of hours worked by the relevant personnel employed by the respondent. If so, the risk was not in my view insignificant in relation to the sacrifice (per Aldous L.J). Contributory Negligence The Law Reform (Contributory Negligence) Act 1945 s(l) provides that: Where any person suffers damage as a result partly of his own fault and partly of the fault of any other person or persons, a claim in respect of that damage shall not be defeated by reason of the fault of the person suffering the damage, but the damages recoverable in respect thereof shall be reduced to such an extent as the court thinks equitable having regard to the claimant’s share in the responsibility for the damage. That is to say, the damages awarded would be affected by the claimant’s contribution to the injury through his own negligence. In considering contributory negligence in Jones v Livox Quarries Limited [1952], Denning LJ said: “…although contributory negligence does not depend on a duty of care it does depend on foreseeability. Just as actionable requires the foreseeability of harm to others, so contributory negligence requires the foreseeability of harm to oneself.” In practice, this means that a person whose injury is caused in part by his own negligence will receive damages reduced proportionately at the discretion of the court. Vicarious Liability Under the doctrine of vicarious liability, an employer is liable for the negligence of an employee when it is committed in the course of that employee’s employment (Hendy and Ford, 2001). Vicarious liability rests on the employer simply as a result of the fact that he is the employer and is deemed to have ultimate control over the employee in what is known as a “master and servant” relationship. This liability must be insured against under the Employer’s Liability Compulsory Insurance Act, 1969 (Stranks, 1999).


194

Duty of the Employer The concept of personal nondelegable duty in the context of an employer’s duty of care and negligence to his servants received recognition from the House of Lords in Wilsons and Clyde Coal Co. Ltd. v English [1938]—a Scottish appeal that marked the development of the modern law of employers’ liability and negligence (White, 1993). Exposition of the employer’s personal duty to his servants is found in particular in the judgment of Lord Wright, who said, “The obligation is threefold—the provision of a competent staff of men, adequate material and a proper system and effective supervision.” Lord Wright further broke down the element of “material” into plant and appliances. It has become customary to separate the place of work from the other element of “plant and appliances” with the result that the obligation is usually spoken of today as being fourfold, as set out by Streatfield J in Hudson v Ridge Manufacturing Company Ltd [1957]—namely, to exercise reasonable care in: • Provision of a safe system of work • Provision of a safe place of work • Provision of proper equipment • Provision of competent staff These are not four separate duties but, as Pierce L.J said in Wilson v Tyneside Window Cleaning Co. [1958], are “ultimately only manifestations of the same duty of the master to take reasonable care to so carry out his operations so as not to subject those employed by him to unnecessary risk.” Provision of a Safe System of Work A very important branch of the employer’s duty is that he must take reasonable care to establish and enforce a proper system or method of work. The importance of “safe system” is that it stresses the obligation to plan the work in advance with due regard to safety (Munkman, 1990). The organization or “system” is a broad term including such matters as set out by Hendy and Ford (2001): • Order of the work • Coordination of different departments, workers, and activities • Numbers and roles of workers • Layout of plant and appliances for special tasks • Methods for using particular machines or carrying out particular processes • Instruction and supervision of workers, especially of trainees and inexperienced workers • Precautions to be taken against risk In Colfar v Coggins and Griffiths (Liverpool) Ltd [1945], Lord Greene said that “the safety of a system must be considered in relation to the particular circumstances of each particular job.” He expressed the opinion that the system may include the physical layout of the job, the sequence in which the work is to be carried out, the provision in proper cases of warnings and notices, and the issue of special instructions.


195

In General Cleaning Contractors v Christmas [1953], Lord Oaksey said, “It is the duty of the employer to give such general safety instructions as a reasonably careful employer who has considered the problem presented by the work would give to the workman.” In planning the system of work, the employer must take into account the fact that workmen become careless about the risks involved in their daily work (Munkman, 1990). In General Cleaning Contractors Ltd v Christmas [1953], Lord Reid and Lord Oaksey considered this issue: Where a practice of ignoring an obvious danger has grown up I do not think it is reasonable to expect an individual workman to take the initiative in devising and using precautions. It is the duty of the employer to consider the situation, to devise a suitable system, to instruct his men what they must do and to supply any implements that may be required (Lord Reid). It is well known to employers that their work people are very frequently, if not habitually, careless about the risks which their work may involve. It is for that very reason that the common law demands that the employer should take reasonable care to lay down a reasonably safe system of work. Employers are not exempted from this duty by the fact that their men are experienced and might be, if they were in the position of an employer, able to lay down a reasonably safe system of work themselves. Workmen are not in the position of employers. Their duties are not performed in a calm atmosphere of a boardroom with the advice of experts. They have to make their decisions on narrow sills and other places of danger and circumstances where the dangers are obscured by repetition (Lord Oaksey). On the question of supervision in Clifford v Charles H.Challen and Son Ltd. (1951), Denning L.J said: The standard which the law requires is that they should take reasonable care for the safety of their workmen. To discharge that duty properly an employer must make allowance for the imperfections of human nature. When he asks his men to work with dangerous substances he must provide proper appliances to safeguard them. He must then set in force a proper system by which they use the appliances and take the necessary precautions; he must do his best to see that they adhere to it. He must remember that men doing a routine task and often heedless of their own safety may become careless about taking precautions. He must, therefore, by his foreman do his best to keep them up to the mark and not tolerate any slackness. He cannot throw all the blame on them if he has not shown good example himself.


196

Consequently, in sofaras a safe system of work is concerned, it is the duty of the employer to consider the situation with regard to the state of knowledge at that time and take reasonable precautions deemed necessary for the safety of his employees. Provision of a Safe Place of Work Since Wilsons and Clyde Coal Co. Ltd. v English [1938], it is the duty of the employer to take reasonable care, by himself or by his servants or agents, to provide a safe place of work (Munkman, 1990). In Cole v De Trafford (2) [1918], Scrutton L.J had already said that “the master is bound to use reasonable care to provide safe premises and appliances for his servants to work in and with, and to use reasonable care to keep them safe.” Workplaces must be adequately lit: Garcia v Hartland and Wolfe Limited [1943], but it must not be so excessive as to be dangerous: Russel v Criterian Film Productions Limited [1936]. The employer must take reasonable care to make and keep floors safe to prevent slipping and tripping: Davidson v Handley Page Limited [1945]. In the case of known dangers, it is a question of fact whether a defendant ought to have taken a special precaution or whether those in fact taken were sufficient (Hendy and Ford, 2001). The duty of an employer is to take reasonable care for the safety of his workmen throughout the course of their employment. This duty does not come to an end because the workmen are sent to work on premises that do not belong to the employer. In McQuilter v Gouldandris Bros. Ltd. [1951], Lord Guthre said: The fact that the work had to be carried out in the premises of a third party did not absolve the employer from his duty of exercising reasonable care for the safety of his workmen. The duty must still be fulfilled, although scope is circumscribed by the fact that the work was been done on premises not in the possession and control of the employer. But he was still under the duty of exercising reasonable care to safeguard him against dangers which he should anticipate and which he had the power to avert. This duty to provide a safe place of work is of particular importance when a manual handling job must be carried out in crouched or confined spaces or when the footing is insecure or the foot placement is limited. Provision of Equipment In Wilsons and Clyde Coal Co. Ltd. v English [1938], the duty as laid down by Lord Wright was to provide suitable plant; however, it was also indicated in his judgment that the plant must be kept in good order: “The obligation to provide and maintain proper plant and appliances is a continuing obligation.” An employer is bound to keep reasonably abreast of safety developments within his type of business, and this may require the adoption and provision of new or improved equipment in the interest of safety. In Toronto Power Co. Ltd. v Paskwan [1915], the defendants were found to have been negligent in circumstances in which the deceased had been killed by a falling block from a traveling crane; evidence indicated that a safety


197

device was in existence at the time of the accident and would have prevented the overwinding of the chain that caused the accident. The duty concerning new and improved equipment is set out clearly in the Canadian case Reed v Ellis (1916): “A master is not bound to provide all the latest devises for the care or benefit of those he employs; he is bound to take reasonable means to protect them from injury in his service.” In Munrow v Plymouth Health Authority (1991), a case involving the lifting of a very heavy female patient, the judge said “one does not look for perfection, one looks for reasonable steps that are practicable. A hoist is a reasonable step that is practicable.” Provision of Staff The employer is now liable for the negligence of a worker’s fellow employees (or even other workers under their control). However, in some cases a claimant will need to rely on breach of the primary duty on the part of the employer to select competent staff (Hendy and Ford, 2001). Vicarious liability requires original negligence on the part of a fellow worker and, if the employer requires an inexperienced workman to do a job outside his competence, the workman may not be negligent when he fails to do the job properly with resulting injury to the claimant. In such cases, therefore, the injured employee, if he is to succeed in his claim, must rely upon a breach of the employer’s duty to provide competent staff by proving that the employer initially employed an incompetent person or failed to ensure the employee’s competence by instruction, training, warning, and supervision. The standard of care required from workmen is that appropriate to their status and duties, and it is not correct to say that an everyday act of carelessness or inadvertence in a factory cannot amount to negligence (Munkman, 1990), or to apply an especially lenient standard of conduct in a case in which workmen are collaborating and working in a team: Staveley Iron and Chemical Co. Ltd. v Jones [1956]. In McCann v J.R.McKeller (Alloys) Ltd. (1969), the House of Lords sustained the verdict of a Scottish jury that it was negligent for a workman to drop his end of a heavy steel ingot, although he excused himself by saying that his hand was suddenly “jagged” or “pricked” by a sharp edge. A proper system of work requires that, when the job cannot be safely carried out by the claimant on his own, he should be provided with sufficient helpers to enable the work to be safely carried out. An employer must exercise reasonable care to ensure that the exertions that he requires from his workers will not be such, by reason of the physical and mental strain that they produce, as to result in injury to workers (White, 1993). The duty is owed to the individual worker. In Paris v Stepney Borough Council [1951], Lord McDermot said, “It is no less clear that the duty is owed to the workman as an individual and that it must be considered in relation to the facts of each particular case.” He further added: It seems to me to follow that in the known circumstances that a particular workman is likely to suffer a graver injury than his fellows then that happening of a given event is one which must be taken into consideration in assessing the nature of the employer’s obligation to that workman.


198

On the issue of suitability for a specific job and the workload involved in Wilsons and Clyde Coal Co. Ltd v English [1938], Lord Thankerton said that “if the employer knows or ought to know that the workman has a vulnerable back they are in breach of duty in requiring him to lift and move weights which are likely to cause him injury even if a normal man can carry them without risk.” The individual worker’s stamina and fatigue were considered by Stuart-Smith LJ in Johnstone v Bloomsbury Health Authority [1992] when he said: Take the case of a man whose contract requires him to work a certain number of basic hours and overtime in addition if required. If he is required to work such long hours that he is exhausted and his attention or concentration fails so that he suffers an accident it is no defence for the employer to say that the workman expressly agreed to work such hours. It must be remembered that the duty of care is owed to the individual employee and different employees may have different stamina. If the authority in this case knew or ought to have known that by requiring him to work the hours which they did, that they exposed him to risk of injury to his health, then they should not have required him to work in excess of those hours that he could safely have done. These cases highlight the duty of the employer to take steps to become aware of each employee’s condition and suitability for the job, in effect indicating the requirement for proper selection procedures and ongoing health surveillance. It is well established at Common Law that a defendant must take the claimant as he finds him; this is commonly called “the eggshell skull principle.” Lord Parker, in Smith v Leech Brain and Co. Limited [1962], said, “It has always been the law in this country that a tortfeasor takes his victim as he finds him.” In relation to musculoskeletal disorders, Judge Byrt, QC, said in McSherry v British Telecommunications PLC [1992]: It is sufficient in law if the tortfeasor should reasonably have foreseen that his breach of duty was likely to cause injury within a broad category of the injury that was in fact caused. I am satisfied that the defendants should have been aware that bad posture could cause musculoskeletal problems. The fact that the injuries sustained were more extensive than those they might have envisaged is of no consequence in law. Statutory Duty All statutory duties are of relevance in criminal proceedings, but only those that carry civil liability can be relied on in civil actions for a breach of statutory duty. Regulations introduced under the Health and Safety at Work etc. Act 1974 give rise to civil liability, unless otherwise stated, under section 47(2) of that Act, which reads: “Breach of a duty imposed by health and safety regulations shall, so far as it causes damage, be actionable except in so far as the regulations provide otherwise.” Health and safety regulations are made under section 15 of the act.


199

Statutory duties may influence the common law standard of care (de Navarro et al., 2001). Statutory duties maybe evidence of good practice that a reasonable employer should adopt (Franklin v Gramophone Company Limited [1948]; National Coal Board v England [1954]; Hewett v Alf Brown’s Transport Limited [1991]). In Bux v Slough Metals Limited [1974], Stephenson L.J, in considering the relevance of statutory regulations to a liability in negligence, said that “the employer must try and make the law of the land the rule of the factory.” Breach of Statutory Duty Whatever an act requires to be done must be done, although it is always a matter of construction precisely what is the nature of the obligation. It is in this sense that, subject to causation, every statutory duty is absolute (Ford and de Navarro, 2001). As Lord Akin said in Smith v Cammell Laird and Company Limited [1940]: “It is precisely the absolute obligation imposed by statute to perform or forbear from performing a specified activity that a breach of statutory duty differs from the obligation imposed by common law, which is to take reasonable care to avoid injuring another.” Ford and de Navarro (2001) consider that in this sense, statutory duty is absolute even though it is qualified by the words “as far as is reasonably practicable” because it is a question of fact whether a thing is reasonably practicable or not and unless, on the facts, it is not reasonably practicable, the requirement must be carried out: Marshal v Gotham Company Limited [1954]. Management of Health and Safety at Work Regulations 1999 The Management of Health and Safety at Work Regulations 1999 contains the principle methods of implementing the EEC Framework Directive (89/391/EEC). These regulations carry civil liability in actions brought by employees under regulation 6 of the Management of Health and Safety at Work and Fire Precautions (Workplace) (Amendment) Regulations 2003. The directive may be useful as evidence of approved employer practice for the purpose of establishing negligence (Smith and Ford, 2001). The regulations are accompanied by an approved code of practice (ACOP) and guidance (L21). In relation to the code, HSC (2000) says, “If you follow the advice you will be doing enough to comply with the law in respect of those specific matters on which the code gives advice” and, in relation to guidance, “Health and safety inspectors seek to secure compliance with the law and may refer to this guidance as illustrating good practice.” Regulation 3 deals with risk assessment and reads: “(1) Every employer shall make a suitable and sufficient assessment of: (a) The risks to the health and safety of his employees to which they are exposed while they are at work; and (b) The risks to the health and safety of persons not in his employment arising out of or in connection with the conduct of his undertaking.”


200

The ACOP gives a general indication as to “suitable and sufficient risk assessment”; this requires a systematic general examination of the effect of the work activities and the condition of the premises, the hazards present, the likelihood of these hazards materializing, and the damage likely to arise. The extent of the sophistication of the risk assessment will depend on the complexity of the processes. All employers and selfemployed people are required to make a risk assessment. The regulation also provides that employers with five or more employees must record the significant findings of their risk assessment. Regulation 4 says that when an employer implements any preventive and protective measures, he should do so on the basis of the principles specified in schedule 1 to the regulations. These are the general principles of prevention set out in article 6(2) of Council Directive 89/391/EEC. The general principles of prevention are “(a) Avoiding risk (b) Evaluating risks that cannot be avoided (c) Combating the risks at source (d) Adapting the work to the individual—especially as regards the design of work places, choice of work equipment, and choice of working and production methods—with a view, in particular, to alleviating monotonous work and work at predetermined work rate and reducing their effect on health (e) Adapting to technical progress (f) Replacing the danger by no danger or less danger (g) Developing a coherent overall prevention policy that covers technology; organization of work; working conditions; social relationships; and the influence of factors relating to the working environment (h) Giving collective protective measures priority over individual protective measures (i) Giving appropriate instructions to employees” Regulation 6 deals with health surveillance and reads: “Every employer shall ensure that his employees are provided with such health surveillance as is appropriate having regard to the risks to their health and safety which are identified by the assessment.” The primary benefit, and therefore objective of health surveillance, should be to detect adverse health effects at an early stage, thereby enabling further harm to be prevented (HSC, 2000). Regulation 10 provides that every employer shall provide his employees with comprehensible and relevant information on the risks to health and safety identified by the assessment and the preventive and protective measures. Regulation 13 deals with capabilities and training. Regulation 13(1) reads: “Every employer shall entrust a task to his employees taking into account their capabilities as regards health and safety.” Regulation 13(2) deals with the issue of training and says that every employer shall ensure that his employees are provided with adequate health and safety training when they are recruited into the employer undertaking and when they are exposed to a new or increased risk. The training shall


201

“(a) Be repeated periodically when appropriate (b) Be adapted to account for any new or changed risks to the health and safety of employees concerned (c) Take place during working hours” (regulation 13(3)) Under regulation 14, employees have a duty to cooperate with the employer to enable him to comply with statutory duties for health and safety and to notify him of any shortcomings in the health and safety arrangements so that he can take remedial action if needed. Manual Handling Regulations The Manual Handling Operations Regulations 1992 made under the Health and Safety at Work etc. Act 1974 implement Council Directive 90/269/EEC on the manual handling of loads and supplement the general duties placed on employers and others by the Health and Safety at Work etc. Act 1974 and the broader requirements of the Management Regulations (HSE, 1998). The Court of Appeal in Koonjul v Thameslink Health Care Trust [2000] considered that the purpose of the regulations was to place an obligation on the employers to look after the safety of their employees who might not behave with proper and full concern for their own safety. The Manual Handling Regulations carry civil liability and are qualified by reasonable practicability. Regulation 2(1) defines manual handling operations: “‘Manual handling operations’ means any transporting or supporting of a load (including the lifting, putting down, pushing, pulling, carrying or moving thereof) by hand or by bodily force.” “Load” includes any person or any animal. By way of definition, the directive says: For the purposes of this Directive, “manual handling of loads” means any transporting or supporting of a load, by one or more workers, including lifting, putting down, pushing, pulling, carrying, or moving of a load, which, by reason of its characteristics or of unfavourable ergonomic conditions, involves a risk particularly of back injury to workers. The various parts of regulation 4(1) establish a clear hierarchy of measures, avoidance, assessment, and reduction of risk. Regulation 4(1)(a) reads: “(1) each employer shall—so far as is reasonably practicable, avoid the need for his employees to undertake any manual handling operations at work which involve a risk to their being injured.” This is the primary duty on the employer. If the general risk assessment carried out under the Management of Health and Safety at Work Regulations 1999 indicates the possibility of injury from manual handling, then the first issue to consider is whether the manual handling operation can be avoided. Manual handling of loads can be avoided by elimination of the handling operation or, alternatively, automation or mechanization. Regulation 4(1)(b)(i) deals with assessment of risk when it is not reasonably practicable to avoid the need for manual handling operations that involve a risk of injury. In these circumstances, employers “shall make a suitable and sufficient assessment of all such manual handling operations to be undertaken by them.” This assessment must have


202

regard to the factors set out in schedule 1 attached to the regulations. This schedule sets out five issues: the task, the load, the working environment, individual capability, and other factors. Questions that the employer must consider in the assessment of manual handling operations are provided on each of these issues. As an appendix to the guidance on the regulations (L23), the HSE published a filter to screen out more straightforward cases and identify manual handling operations in which a more detailed risk assessment is necessary. The filter is based on a set of numerical guidelines developed from data published in scientific literature and on practical experience of assessing risks for manual handling. They are pragmatic, tried, and tested; they are not based on any precise scientific formula. The intention is to set out an approximate boundary within which the load is unlikely to create a risk of injury sufficient to warrant a detailed assessment (HSE, 1998). This filter indicates a maximum weight of 25 kg for men and 16 kg for women in operations involving lifting a load. Regulation 4(1)(b)(ii) deals with reducing the risk of injury when the manual handling operation cannot be avoided and the obligation on the employer is to “take appropriate steps to reduce the risk of injury to those employees arising out of their undertaking any such manual handling operation to the lowest level reasonably practicable.” Health, safety, and productivity are most likely to be optimized if an ergonomic approach is used to design the manual handling operations as a whole. Wherever possible, consideration should be given to the task, load, working environment, and individual capacity, as well as the relationship among them, with a view to fitting the operations to the individual rather than the other way around (HSE, 1998). Reduction of risk can be brought about through work place or job design; mechanical assistance; improving work routine; lightening the load and making it easier to manage; improving space constraints and footing; improving thermal environment, ventilation, and lighting; and providing information and training. Regulation 4(1)(b)(iii) deals with the provision of additional information in relation to the load. This says that precise information, in situations in which it is reasonably practicable to do so, should be given on the weight of each load and the heaviest side of any load where the center of gravity is not positioned centrally. Regulation 5 deals with duties of employees and reads: “Each employee while at work shall make full and proper use of any system of work provided for use by his employer in compliance with regulation 4(1)(b)(ii) of these regulations.” Employees are already under a duty under section 7 of the Health and Safety at Work etc. Act 1974 to take reasonable care for their own health and safety and that of others and to cooperate with their employers to enable them to comply with their duties. Application of Manual Handling Regulations Application of the regulations is not confined to injuries arising from carrying excessive loads, as illustrated by the following cases. In King v RCO Support Services Limited [2001], Kay LJ accepted that the regulations applied in circumstances in which the claimant sustained personal injuries when he slipped on ice while shoveling grit because the operation fell within the definition of “manual handling operations” in regulation 2. The fact that the primary cause of the accident was ice rather than the manual handling was irrelevant; the possibility of an icy


203

surface was a factor that should have been considered when assessing the risk of the manual handling operation (Hermer and Ford, 2001). The plaintiff in Hawkes v London Borough of Southwark (1998) was injured when he fell down stairs while carrying a door. Also, it was accepted in Purves v Buckingham County Council (1998) that a teacher grabbing an unruly child could fall within the regulations (Hermer and Ford, 2001). Work Equipment Regulations The Provision and Use of Work Equipment Regulations 1998 are intended to implement the Work Equipment Directive, 89/655/EEC, and amendments to the directives. The regulations adopt a wide definition of “equipment” and impose general obligations in relation to its safety. These regulations carry civil liability. Two regulations are of specific interest in relation to musculoskeletal disorders. Regulation 4(2) states “in selecting work equipment, every employer shall have regard to the working conditions and to the risks to the health and safety of persons which exist in the premises or undertakings in which that work equipment is to be used and any additional risk posed by the use of that work equipment.” The Approved Code of Practice (HSC, 1998), which accompanies the Provision and Use of Work Equipment Regulations 1998, says that the employer should take account of ergonomic risks in selecting the equipment. Regulation 5(1) requires every employer to ensure that work equipment is maintained in an efficient state, efficient working order, and good repair. The guidance accompanying the regulations warns that ergonomic design must take into account the size and shape of the whole body and should ensure that the design is compatible with human dimensions. Operators should not be expected to exert undue force or stretch or reach beyond their normal strength or physical reach limitation to carry out tasks. This is particularly important for highly repetitive work such as working on supermarket checkouts or high-speed “pick and place” operations. Thus, it can be seen that the ambit of regulation 5(1) is of very real significance in that it gives rise to strict liability if injury has been caused by a malfunctioning machine, irrespective of the system of maintenance and repair imposed by a conscientious employer. In Cadger v Vauxhall Motors (2000), regulation 5 was breached when pneumatic work equipment malfunctioned; arguments about the rea sonableness of the system of inspection and maintenance were irrelevant (Ford and Hermer, 2001). Display Screen Equipment (or Visual Display Unit [VDU]) The Health and Safety (Display Screen Equipment) Regulations 1992 give rise to civil liability and are not supported by an approved code of practice. However, there is guidance on the regulations (HSE, 2001). The regulations define a work station and set out the duties of the employer in relation to assessment of risks, requirements for


204

workstations, daily work routine of users, provision of eye and eye-sight tests, training, and information. Regulation 1(2) (a) says that “display screen equipment” means any alphanumeric or graphic display screen, regardless of the display process involved. A “user” means an employee who habitually uses display screen equipment as a significant part of his normal work (regulation 1(2)(d)). In regulation 1(2)(e), “work station” means an assembly comprising: • Display screen equipment • Any optional accessories to the display screen equipment • Any disk drive, telephone, modem, printer, document holder, work chair, work desk, work surface, or other items peripheral to the display screen equipment • The immediate work environment around display screen equipment Regulation 2 deals with risk assessment of work stations and the updating of the assessment. The obligation of the employer under regulation 2(3) is to “reduce the risks identified in consequence of an assessment to the lowest extent reasonably practicable.” The guidance note at paragraph 21 says a suitable and sufficient analysis should be systematic; be appropriate to the likely degree of risk; be comprehensive, covering organization, job, work place, and individual factors; and incorporate information provided by employer and worker. Possible risks that have been associated with display screen equipment workers are summarized at Annex B to the Guidance on the Display Screen Regulations. Paragraph 1, Annex B reads: “The introduction of VDUs and other display screen equipment has been associated with a range of symptoms related to the visual system and working posture. These often reflect bodily fatigue. They can readily be prevented by applying ergonomic principles to the design, selection and installation of display screen equipment, the design of the work place and the organisation of the task.” The main conditions are upper-limb pains and discomfort, eye and eyesight effects, fatigue, and stress. Other concerns are epilepsy, facial dermatitis, electromagnetic radiation, and effects on pregnant women (HSE, 2001). Regulation 3 sets out the duty of the employer to ensure that work stations meet the requirements laid down in the Schedule to the Regulations. This schedule sets out the minimum requirements for work stations, which are contained in the annex to Council Directive 90/270/EEC on the minimum safety and health requirements for work with display screen equipment. An employer should ensure that a work station meets the requirements laid down in the schedule to the extent that: • These requirements relate to the component present in the work station concerned • These requirements have an effect with a view to securing the health, safety, and welfare of persons at work • The inherent characteristics of a given task make compliance with these requirements appropriate with respect to work station concerned


205

The requirements deal with equipment (display screen, keyboard, work desk or work surface, work chair); environment; space requirements; lighting; reflections and glare; noise; heat; radiation; humidity; and the interface between the computer and the operator/user where the general principles of software ergonomics should be taken into account (HSE, 2001). Regulation 4, dealing with daily work routine of users, reads: “Every employer shall so plan the activities of users at work in his undertaking that their daily work on display screen equipment is periodically interrupted by such breaks or changes of the activity as to reduce their work load at the equipment.” The guidance on the regulations states that when the display screen work involves intensive use of the keyboard, any activity that would demand broadly similar use of the arms or hands should be avoided during breaks. Similarly, if the display screen work is visually demanding, any activities during breaks should be of a different visual character. Breaks must also allow users to vary their posture. Exercise routines that include blinking, stretching, and focusing eyes at distant objects can be helpful and could be covered in training programs (HSE, 2001). Regulation 6 deals with the provision of training and says that the employer should ensure that users are provided with adequate health and safety training in the use of any work station upon which they are required to work. Health and safety training should be aimed at reducing or minimizing the risk areas set out in Annex B for each individual user and take account of guidance in Annex A. Information provided under regulation 7 must deal with all aspects of health and safety relating to work stations and the measures taken by the employer under the regulations that relate to the workers and their work. Minor amendments have been made by the Health and Safety Executive (Miscellaneous) (Amendments) Regulations 2002 to regulations 3, 5 and 6 relating to workstations, eye testing and training. Additional guidance on the regulations (L26, 2nd ed., HSE 2003) has been issued. Consultation The Safety Representative and Safety Committee’s Regulations 1977 and the Health and Safety (Consultation with Employees) Regulations 1996 deal with consultation with employees and their representatives. These regulations do not carry a civil right of action. These regulations ensure that all employees, whether represented by a trade union or not, must be consulted on the introduction of measures in the place of work that may substantially affect the health and safety of employees, the planning and organization of any health and safety training as required to be provided, and the health and safety consequences for employees of the introduction of new technology into the work place (HSE, 1996). There is also a duty on the employer’s part to make available all relevant information to enable employees to participate fully and effectively in the consultation process; these should cover the provision of accident reports and risk assessments that have been carried out (Zindani, 1998). The Guidance (HSE, 1998) accompanying the Manual Handling Regulations stresses the importance of consultation and the contribution that workers can make:


206

…the views of staff can be particularly valuable in identifying manual handling problems and practical solutions to them. Encourage employees, their safety representatives and safety committees to play a positive part in the assessment process. They can assist the employer by highlighting difficulties from such things as the size or shape of loads, how often they are handled, or the circumstances in which the handling operations are carried out. European Union Directives A claimant can bring an action against an emanation of the state based on a directive. In the leading case of Foster v British Gas Plc (1991), the European Court of Justice stated that direct effect could be relied upon against: “A body, whatever its legal form, which has been made responsible, pursuant to a measure adopted by the state, for providing a public service under the control of the state and has for that purpose special powers beyond those which result from the normal rules applicable in relations between individuals.” Under this ruling, an independent police authority is an emanation of the state, as are public health bodies; local and regional authorities; tax authorities; a nationalized corporation; a privatized water company; and the governing body of a voluntary aided school (Meade and Langstaff, 2001). Meade and Langstaff (2001) suggest that European jurisprudence has no obvious counterpart to that of the U.K. requirement of “reasonable practicability” as a qualification to the performance of an employer’s duty in many regulations. However, in the Court of Appeal referring to the Manual Handling Directive (90/269/EEC), which includes the terms “appropriate organizational measures” and “appropriate means,” Hale L.J, in King v Sussex Ambulance NHS Trust (2002), said, “The Directive does not refer to what is ‘reasonably practicable’ but that must be what it means by taking ‘appropriate measures’ or using ‘appropriate means’ to ‘reduce’ the risk.” 10.4 Examples of Case Law In a personal injury civil action, the claimant must show that he is injured and that the injury was caused, wholly or partly, by risks to which he was exposed at his work. If the action is brought in negligence, the claimant must show that the defendant was in breach of his duty of care in that the risk to which he was exposed was reasonably foreseeable and that it would have been reasonably practicable to circumvent the risk. If the action is brought under a breach of statutory duty, the claimant must show that he was protected by the statute, which placed the duty on the defendant and was breached by the defendant causing the injury. How the courts approach these issues is illustrated by the following examples of case law.


207

In relation to the existence or not of an injury, Mughal v Reuters Ltd. (1993) is a much-publicized case concerning musculoskeletal disorders. In this case, Judge Prosser said, “I believe that the main stream view is that there is no pathology, no clinical symptoms that can be pointed to as confirming a patient having RSI,” and he agreed with the view “that RSI has no place in the medical books.” An injury must be shown to be related to work. In Alexander v Midland Bank PLC (1999), the judge accepted that the condition from which the claimants suffered, fibromyalgia, was physical and caused by factors of repetitive work under intense pressure with insufficient work breaks and sustained bad posture. The Court of Appeal found that the judge had a choice between two alternative explanations. The five claimants had to prove that the physical explanation for their upper-limb disorders was more probable than the psychogenic cause suggested by the defendants. On the evidence, the judge had been entitled to conclude that the fibromyalgia was physically based and arising from the work conditions. In McPherson v London Borough of Camden (1999), it was held that De Quervain’s syndrome could be caused occupationally by repeated movements of the wrist or the thumb. The use of keyboards, on occasion, could cause the necessary stenosis and inflammation. The claimant had used her thumb excessively and her condition had been caused by the continual use of the keyboard in the poor ergonomic and working-time conditions that prevailed from June 1993 to January 1994. On the issue of foreseeability, the Court of Appeal in Koonjul v Thameslink Health Care Trust (2000) held that there must be a real risk—a foreseeable possibility of injury—and certainly nothing approaching a probability. Furthermore, Hale LJ stated that, in making such assessments, an element of realism must be present. The Court of Appeal in Alsop v Sheffield City Council (2002) adopted this reasoning. The facts of this case were that the claimant, a refuse collector, was injured pulling a wheelie bin up a concrete ramp with a gradient of 30°. There were no complaints by dustmen or refuse collectors and there were no recorded accidents of the type suffered by the claimant. The Court held that the council could do no more than to tell experienced dustmen to use their common sense. The claimant should have made use of the options open to him, which included the possibility of wheeling the bin to the end of the gradient where the road met the pavement at an even level. The Court of Appeal in Koonjul v Thameslink Health Care Trust (2000) considered the issue of reasonable practicability in making an assessment of manual handling operations under regulation 4(1) (b). The question of what involved a risk of injury was context-based and, accordingly, it was necessary to look at the particular operation in context; this included the context of the particular place of employment and also the particular employees involved. This case involved an employee in a small residential home with a small number of employees injuring her back pulling a bed out from a wall. The bed was against the wall to prevent children falling out of the bed. The Court held that the employer was entitled to take the fact that the employee had been carrying out the task for a very long time into account in assessing the risk of injury. In considering the duties of the employer, the court found that the first obligation was to prevent the risk but, in this case, there was good reason for having the bed against the wall to prevent the children falling out.


208

In these circumstances, the employer’s second obligation was to reduce the risk of injury to the lowest level reasonably practicable. In this case, the employment involved a number of everyday tasks and the Court considered the idea that the level of risk involved should be met by a precise evaluation of each of those tasks and precise warnings to each employee as to how each was to be carried out took the case way beyond the realms of practicability. In these circumstances, the Court held that there was no breach of the regulations. In O’Neill v DSG Retail Limited (2002), the claimant was supporting a microwave oven when he was called by a colleague; he turned by twisting his body toward the direction of the call without moving his feet. It was conceded by the defendants that, in breach of their own policy, the claimant had not been given practical training, nor had he been shown a video. The purpose of the training was to highlight the risks involved in lifting and to make employees aware of them by watching demonstrations of safe lifting and then practicing such techniques. The video contained demonstrations of safe lifting and also material designed to train people out of the instinct to twist when carrying a load. It was held by the Court of Appeal that the defendants’ failure to give the claimant the desired practical training and video gave rise to a foreseeable possibility of injury in relation to the particular task he had been carrying out and that failure to provide training was a cause of the claimant’s injury. By not giving the claimant training and not showing him the training video, the defendants failed to reduce the risk of injury to the lowest level reasonably practicable and were in breach of regulation 4(1)(b)(ii) of the Manual Handling Operation Regulations 1992. The 36-year-old claimant in Knott v Newham Health Care NHS Trust (2002) suffered a back injury and alleged that the defendants had no adequate or proper system for manually handling patients, which meant that she was exposed to risk of back injury. The Court held that, on the evidence, there was only one hoist for use in two wards and that hoist was often inoperable. The consequence of the inadequacy of mechanical aids was that the claimant would habitually use a drag lift to move heavy patients. The drag lift was an inherently unsafe method that carried with it a real risk of injury. The defendants did not operate an appropriate system for lifting patients and no real steps had been taken to reduce the risk of injury to its employees to the lowest level reasonably practicable during the relevant period; the defendants were therefore in breach of the regulations. It was also held that the claimant’s degenerative disease rendered her particularly vulnerable to disc prolapse, and the lifting of patients during the period of her employment with the defendants was likely to have damaged the annulus of the claimant’s disc posteriorily. The disc prolapse and neural damage were the eventual result of the process. It followed that the defendants’ breach of duty caused (and, at the very least, materially contributed to) the claimant’s injury. In King v Sussex Ambulance Health Trust (2002), the claimant was injured while carrying a patient in a carry-chair with a fellow worker down a stairway that was narrow, steep, and had a bend in it. The Court of Appeal judgment referred to it as an awkward lift with a not particularly heavy patient who needed a response within an hour. The only alternative was to call the fire brigade and have the patient taken out through a window.


209

The claimant’s case was brought on the basis of a breach of the Manual Handling Directive, a breach of the Manual Handling Regulations, and in negligence. In the County Court, the judge held that there was a breach of article 3.2 of the directive and said that calling the fire brigade should have been given more serious consideration than it was. The Court of Appeal held against the claimant on all grounds. Lady Justice Hale said that it was clear that if there were no liability under the directive then there was no liability under the regulations and, in her view, there was no liability under either. There was nothing to suggest that calling the fire brigade would have been an appropriate measure in this case to avoid the need for carrying patients down the stairs or to reduce risk of injury in so doing. If calling the fire brigade was not appropriate or reasonably practicable for the purposes of the directive or the regulations, it cannot be a lack of reasonable care to fail to do so. In practice, the claimant’s damages are reduced by a percentage comparable to the degree of fault that the court finds against the claimant (Barrett and Howells, 1997). The courts can impose substantial reductions on claimant damages due to contributory negligence. In Mearns v Lothian Regional Council (1991), the worker’s damages were reduced by 50% because help was available even though not readily available. In McCaffery v Datta (1997), a finding of one-third contributory negligence was upheld by the Court of Appeal. The Court considered that, given the claimant’s previous history of back problems, she could have arranged the patient’s bed to minimize the risk. However, a court will be slow to criticize mere inadvertence or inattention on the part of a worker (Zindani, 1998). In Caswell v Powell Duffryn Associated Collieries Ltd. [1940], Lord Porter said: The skill gained by a workman may enable him to take risks and do acts which in an unskilled man would be negligence, and on the other hand the fatigue and repetition of the same work may make a man incapable of the same care, and therefore not guilty of negligence, in doing or failing to do an act which a man less fatigued would do or leave undone. In a patient-lifting case, Munrow v Plymouth Health Authority (1991), the judge said, I think if one looks at the facts of this case in the light of the training which she received, this plaintiff could have chosen a hoist, but that she received nothing in the way of training nor assistance from any nursing plan which would have steered her towards a hoist and though I think she made the wrong decision I do not think that the wrong decision would be sound by way of contributory negligence. The cases show that the courts consider that the application of regulations and the principles of negligence must involve common sense and an element of realism while taking into account the particular circumstances of each individual case.


210

10.5 Checklist of Main Factors to Consider in a Legal Action in Relation to Work-Related Musculoskeletal Injury Primary factors Considerations Personal details

Age Previous injury Work history Gender Medication Out-of-work activities Selection Evidence of selection procedures Medical examination Suitability for the task Instruction and Evidence of instruction and training training Job-specific instruction and training Warning of the hazards, including unfavorable ergonomic conditions Evaluation of training effectiveness Revision of training Supervision Extent of supervision Extent of supervisory support for training Health surveillance Provision of health surveillance Methods of health surveillance Action on the results of health surveillance Consultation and Evidence of the consultation process complaints Evidence of complaints procedure Evidence of action on consultation or complaints Activity Description of the work system Description of the incident/injury Amplitude, frequency, and duration of exposure Subjective assessment Employee’s involvement in determining the workload Capacity to meet workload requirements Avoidance Possible avoidance of the injury-related task by job redesign or automation, work organization, or layout measures

Primary factors Risk assessment

Risk reduction

Considerations Generic risk assessments Can risk be avoided Detailed risk assessment Update of risk assessments and evaluation of interventions Use of an ergonomic approach including consideration of the task, equipment, job rotation, rest breaks, environment, work organization, job design, and the individual


211

References Barrett, B. and Howells, R. (1997). Occupational Health and Safety Law . London: Financial Times Pitman Publishing. Buckle, P.W. and Devereux, J.J. (2002). Work-related neck and upper limb musculoskeletal disorders: reaching a consensus view across the European Union. Appl. Ergonomics 33:207– 217. Carayon, P., Smith, M.J., and Haims, M.C. (1999). Work organization, job stress, and work-related musculoskeletal disorders. Hum. Factors 41:644–663. Carter, G.T. and Howard, J.S. (1995). Legal aspects on services for the disabled. In: R.A.F.Cox, F.C. Edwards, and R.I.M.McCallun (Eds.) Fitness for Work Medical Aspects , 2nd ed. Oxford: Oxford University Press, de Navarro, M., Ford, M., and Hendy, J. (2001). The general principles of negligence. In J.Hendy and M.Ford (Eds.) Munkman on Employer’s Liability , 13th ed. London: Butterworth, 33–65. European Agency for Safety and Health at Work (2000). Repetitive Strain Injuries in the Member States of the European Union: the Results of an Information Request . Luxembourg: Office for Official Publications of the European Communities. Ford, M. and de Navarro, M. (2001). Breach of statutory duty. In: J.Hendy and M.Ford (Eds.) Munkman on Employer’s Liability , 13th ed. London: Butterworth, 229–255. Ford, M. and Hermer, R. (2001). Machinery and equipment. In: J.Hendy and M.Ford (Eds.) Munkman on Employer’s Liability , 13th ed. London: Butterworth, 317–332. Health and Safety Commission (1998). Safe Use of Work Equipment: Approved Code of Practice and Guidance . London: HMSO. Health and Safety Commission (2000). Management of Health and Safety at Work: Approved Code of Practice and Guidance . London: HMSO. Health and Safety Executive (1996). A Guide to the Health and Safety (Consultation with Employees) Regulations . Norwich: HMSO. Health and Safety Executive (1998). Manual Handling: Guidance on Regulations . London: HMSO. Health and Safety Executive (2001). Display Screen Equipment Work: Guidance on Regulations . London: HMSO. Health and Safety Executive (2002). Upper Limb Disorders in the Workplace HSG60 (rev) . Sudbury: HSE Books. Hendy, J. and Ford, M. (2001). The employer’s duty of care. In: J.Hendy and M.Ford (Eds.) Munkman on Employer’s Liability , 13th ed. London: Butterworth, 91–116. Hermer, R. and Ford, M. (2001). Manual handling. In: J.Hendy and M.Ford (Eds.) Munkman on Employer’s Liability , 13th ed. London: Butterworth, 333–344. Kelly, V. (1998). Development of a tool for use in civil action involving a cumulative back injury. Unpublished M.Sc. thesis, University of Surrey. Meade, P. and Langstaff, B. (2001). European law. In: J.Hendy and M.Ford (Eds.) M unkman on Employer’s Liability , 13th ed. London: Butterworth, 257–283. Munkman, J. (1990). Employer’s Liability at Common Law , 11th ed. London: Butterworth. Pheasant, S. (1991). Ergonomics, Work and Health . London: Macmillan. Pheasant, S. (1996). Bodyspace—Anthropometry, Ergonomics and the Design of Work . London: Taylor & Francis. Scott, I. and Langstaff, B. (2001). Industrial diseases. In: J.Hendy and M.Ford (Eds.) Munkman on Employer’s Liability , 13th ed. London: Butterworth, 193–221.


212

Smith, I. and Ford, M. (2001). The general legislation: the Framework Directive, the Temporary Workers Directives and the Management of Health and Safety at Work Regulations. In: J.Hendy and M. Ford (Eds.) Munkman on Employer’s Liability , 13th ed. London: Butterworth, 285–301. Smith, M.J. and Sainforth, P.C. (1989). A balance theory of job design for stress reduction. Int. J. Ind. Ergonomics 4:67–79. Stranks, J. (1999). Health and Safety Law , 3rd ed. London: Financial Times Pitman Publishing. White, J. (1993). Civil Liability of Industrial Accidents . Dublin: Oaktree Press. Zindani, J. (1998). Manual Handling: Law and Litigation . Birmingham, England: CLT Professional Publishing.

Cases Alexander v Midland Bank PLC (1999) IRLR 723. Allsop v Sheffield City Council (2002) EWCA civ429, LTL 05/03/2002. Bankstown Foundry Pty Limited v Braistina (1986) 160 CLR 301. Blyth v Bermingham Water Works Company (1856), (1843/60) All ER Rep 478. Bolton v Stone [1951] AC 850, (1951) 1 All ER 1078. Bux v Slough Metals Ltd [1947] 1 All ER 262, CA. Cadger v Vauxhall Motors (2000) 6 CLD. Cartwright v G.K.N. Sankley Ltd (1973), 14 KIR 349, CA. Caswell v Powell Duffryn Associated Collieries Ltd [1940] AC 152, (1939) 3 All ER722. Clifford v Charles H. Challen & Son Limited (1951) 1 KB 495, (1951) 1 All ER 72. Cole v De Trafford (No 2) [1918] 2 KB 523, (1918/19) All ER Rep 290. Colfar v Coggins and Griffiths (Liverpool) Ltd. [1945] AC 197, (1945) 1 All ER 326. Davidson v Handley Page Limited [1945] 1 All ER 235. Edwards v National Coal Board [1949] 1 KB 704, (1949) 1 All ER 743. Foster v British Gas PLC (1991) FCR 84. Franklin v Gramophone Co. Ltd. [1948] 1 KB 542, (1948) 1 All ER 353. Garcia v Hartland and Wolfe Limited [1943] 1 KB 731, (1943) 2 All ER 477. General Cleaning Contractors v Christmas [1953] AC 180, (1952) 2 All ER 1110. Glasgow Corporation v Muir (1943) AC 448, (1943) 2 All ER 44. Hawkes v London Borough of Southwark (20 February 1998, unreported), CA. Hayes v Pilkington Glass Limited [1998] PIQR P 313, CA. Hewett v Alf Brown’s Transport Ltd. [1991] ICR 471, (1992) 15 LS Gaz R33, AC Hudson v Ridge Manufacturing Company Limited [1957] 2 QB 348, (1957) 2 All ER 299. Johnstone v Bloomsbury Health Authority (1992) QB 333, (1991) 2 All ER 293. Jones v Livox Quarries Limited [1952] 2 QB 608, 96 Sol Jo 344. Joseph v Ministry of Defence (1980) Times, 4 March, CA. King v RCO Support Services Ltd. [2001], ICR 608, AC. King v Sussex Ambulance NHS Trust (2002) EWCA civ 953, LTL 05/07/2002. Knott v Newham Health Care NHS Trust (2002), (2002) EWHC2091 (QB), LTL 16/10/2002. Koonjul v Thameslink Healthcare Services NHS Trust [2000] LTL 28/03/2000. Marshall v Golham Co. Ltd. [1954] AC360. McCaffery v Datta (1997) CLR 142.


213

McCann v J.R. McKeller (Alloys) Limited [1969] SC (HL) 1. McPherson v London Borough of Camden (1999) QBD, Itl 24/05/1999. McQuilter v Goulandris Bros. Limited [1951] SLT (notes) 75. McSherry v British Telecommunications PLC [1992] 3 Med LR 129. Mearns v Lothian Regional Council (1991) SLT 338. Morris v West Hartlepool Steam Navigation Company Limited [1956] AC 552, (1956) 1 All ER 385. Mughal v Reuters Ltd. (1993) IRLR 571, 16 BMLR 127. Munrow v Plymouth Health Authority (1991) (unreported 5/12/91) H.C. Exeter. National Coal Board v England [1954] AC 403, (1954) 1 All ER 546. O’Neill v DSG Retail Ltd. (2002), LTL 31/07/2002. Paris v Stepney Borough Council [1951] AC 367, (1951) 1 All ER 42. Purves v Buckingham County Council, (20 Nov. 1998, unreported) QBD. Reed v Ellis (1916) 38 OLR 123. Russell v Criterion Film Productions Limited [1936] 3 All ER 627. Schwalb v H. Fass and Son Limited (1946) 90 Sol Jo 394. Smith v Cammell Laird & Co. Ltd. [1940] AC 242. Smith v Leech Brain & Company Limited [1962] 2 QB 405, (1961) 3 All ER 1159. Staveley Iron and Chemical Company Limited v Jones [1956] AC 627, (1956) 1 All ER 403. Stokes v Guest Keane Nettlefold (Bolts and Nuts) Limited [1968] 1 WLR 1776, 5 KIR 401. Thompson v Smith Ship Repairers (Northshields) Limited [1984] QB 405, (1984) 1 All ER 881. Toronto Power Co. Ltd. v Paskwan [1915] AC 734. Wagon Mould (no 2) [1967]—Overseas Tankership (U.K.) Ltd. v Miller SS Co. Pty. (1967) 1 AC 617, (1967) ALR 97. Walker v Wabco Automotive U.K. Limited (1999, unreported), CA. White v Hallbrooke Precision Castings Limited [1985] IRLR 215, CA. Wilson v Tyneside Window Cleaning Company [1958] 2 QB 110, () 2 All ER 265. Wilsons and Clyde Coal Company Limited v English [1938] AC 57, (1937) 3 All ER 628. Wright v Dunlop Rubber Company Limited and ICI Limited (1972) 11 KIR 311, 115 Sol Jo 366.

11 Age and Functioning in the Legal System: Victims, Witnesses, and Jurors Deborah Davis University of Nevada, Reno Elizabeth F.Loftus University of California, Irvine 0–415–28870–3/05/$0.00+$ 1.50 © 2005 by CRC Press

11.1 Objectives and Organization of the Chapter As the population of elderly citizens continues to increase in the United States, greater numbers of older adults will become victims of crimes and later report their experiences to police, attorneys, and juries, and perhaps attempt to identify the perpetrators (National Institute on Aging, 1996). In fact, already we know that roughly 2 million elderly individuals become victims of crime each year (U.S. Bureau of Justice Statistics, 2002). Likewise, older Americans will become disproportionately represented among accident victims, witnesses in civil and criminal trials, and juries. Thus, aging has an impact upon the legal system in myriad ways, from its role in generating situations subsequently litigated in the courts, to the vagaries of memory among aging witnesses, to the role of age in juror judgments. To provide context for understanding age-related declines in cognitive functioning, we begin this chapter with a brief review of changes in the brain and their consequences for basic mechanisms of cognitive processing. Subsequently, we consider specific agerelated deficits in cognitive functioning that (1) cause or contribute to incidents generating torts; (2) impair accuracy of testimony regarding these incidents; and (3) affect jurors’ processing of evidence and, thus, verdicts. 11.2 Sources of Age-Related Declines in Cognitive Functioning The Aging Brain Modern cognitive neuroscience has begun to identify changes in the brain that in turn appear to cause changes in fundamental cognitive processes such as information processing, memory, performance, and judgment (for reviews of the following, see Glisky, 2001; Grady, 2001; Raz, 2000; Reuter-Lorenz, 2000; Schacter, 1996; and the


216

special 2002 issue of Neuroscience & Biobehavioral Reviews Vol. 6, Issue 7, on the aging brain). Structurally, for example, there is widespread shrinkage of brain matter volume (and thus enhanced volume of cerebrospinal fluid) stemming from such underlying causes as shrinkage of the brain tissue, cell loss in some regions, and reduced dendritic branching. Enhanced atrophy is one of the hallmarks of Alzheimer’s disease, but is also characteristic of older adults without such a diagnosis. Atrophy is also greater in those with high blood pressure or declining diastolic blood pressure (de Heijer et al., 2003). Brain mass steadily shrinks after the 50s at the rate of roughly 5 to 10% per decade, and brain volume is related to cognitive performance and dementia (Anstey and Maller, 2003; Mingus et al., 2002; Pantel et al., 2003; Scahill et al., 2003; Sullivan and Ruffman, 2004). Indeed, those who begin with smaller volume (i.e., smaller head circumference) and have less education are four times as likely to suffer dementia as others (Mortimer et al., 2003). Perhaps related to atrophy of tissues and cell volume, brain metabolism also changes so that cerebral blood volume and flow and cerebral metabolic rate of oxygen utilization decline. Age is negatively related to cerebral blood flow, which in turn predicts memory performance (Santens et al., 2003). Neuropathology also tends to escalate with age. A yellowish-brown lipid lipofuscin, called “wear and tear pigment,” accumulates in cells throughout the cerebellum and cerebral cortex, typically accompanied by decrease in myelination of nerve axons and increased numbers of vascular lesions. In Alzheimer’s patients, the brain accumulates neuritic amyloid plaques and tau protein tangles, which hamper neuronal function and eventually kill the cells. The extent of these plaques and tangles in neural tissue (although the extent of tau tangles is more predictive), combined with the number of areas of the brain in which they are present and the presence of stroke damage, directly predicts cognitive symptoms (Schneider et al., 2003; Snowdon, 2001; Guillozet et al., 2003). Neuronal communication is impaired by reduced dendritic branching and demyelination of axons (Bartzokis, 2004), as well as decline in neurotransmitters such as dopamine (which contributes to frontal lobe functions) and acetylcholine (which plays a role in learning and memory). One of the most prominent changes in the white matter of the brain (myelinated axons connecting nerve cells) is known as white matter hyper intensities (WMHs), which can be observed in MR images (or called leukoaraiosis when seen on CT scans). These WMH changes are assumed to reflect damage from vascular as well as neuronal causes and, particularly in some regions, they predict cognitive impairment (Bigler et al., 2003; Deary et al., 2003; de Groot et al., 2000). The effects of these abonormalities can be attenuated in those with greater education, however (Dufouil et al., 2003). The preceding changes in brain structure and pathology are not uniformly distributed throughout the brain. Some structures atrophy at greater rates, such as the hippocampal formation; the frontal, temporal, and parietal convexities; and the parasagittal region, whereas others, such as the occipital lobe, remain relatively spared. Changes in neuropathology are also somewhat region specific. Given this situation, it is not surprising that consensus exists among cognitive-aging researchers that although mental processes generally slow and become less efficient—and thereby become, in some cases, less accurate—arenas exist in which performance is age invariant or even positively

Age and functioning in the legal system

217

related to age (see reviews of age-related changes in a variety of cognitive processes in Craik and Salthouse, 2000; Salthouse, 2004; Verhaeghen and Salthouse, 1997). Generally, memory researchers have found greater declines in memory tasks that require a great deal of self-initiated or effortful processing (such as complex reasoning) but age invariance on memory tasks that require less effortful processing (such as word recognition) (see reviews in Craik and Salthouse, 2000; Light, 1991; Park and Schwartz, 2000). Generally, speed of processing, reasoning, and episodic memory (memory for events) appear to decline most rapidly (Salthouse, 2004; Verhaeghen and Salthouse, 1997). Selective decline in cognitive functioning appears to be related to selective decline in the brain. Unfortunately, some of the most vulnerable structures are also those most important for cognitive functioning. For example, the hippocampus (which contributes to formation and maintenance of memories) and the amygdala (considered particularly important for emotional memories) exhibit loss of number and volume of neurons, as well as synaptic loss. Medial-temporal dysfunction has been related to losses in long-term memory. Furthermore, the frontal lobes, and particularly the prefrontal cortex, are widely thought to suffer decline sooner and more extensively than other regions. Losses in volume, cells, and blood flow are greater in frontal regions, and some evidence indicates selectively greater decline in neurotransmitters and myelin (see the review in Rabbitt et al., 2001). Frontal lobe function has been shown to predict performance on a variety of tests of memory, cognitive functioning, and executive functions such as planning, initiating, and carrying out goal-directed behaviors—and inhibition of inappropriate behaviors. Age-related deficits in frontal functioning are related to a variety of failures of perception, memory, and decision making in forensic contexts. These will be reviewed in subsequent sections. Finally, the corpus callosum, which connects the various regions of the brain, suffers enhanced atrophy relative to some other regions of the brain, thus suggesting that it may be particularly susceptible to the effects of aging. Because a variety of processing and memory functions require communication between regions, dysfunction of the corpus callosum will contribute to age-related cognitive deficits. Indeed, age-related differences in processing efficiency have been shown to be greater for tasks requiring communication through the corpus callosum (see the review in Reuter-Lorenz, 2000). Indeed, some have argued that neural networks and larger systems of interconnected brain regions and the functional activity in these circuits may be generally more important than specific cortical regions in predicting cognitive functioning (Tisserand and Jolles, 2003) and that the frontal cortex may exert its effect upon executive functioning in combination with processes involving links between different brain regions (Andres, 2003). Raz (2000) noted that the magnitude of dysfunction is related to the order of appearance of structures in mammalian history such that “last in is first out” (p. 37). Those that evolved in more ancient species and those that appear earlier in human development are less affected by age than those that evolved or appear later. Thus, structures such as the prefrontal cortex, which evolved last and is presumed to be the seat of the intelligence that separates humans from lower animals, suffers the earliest and greatest decline with age. Structures that evolved earlier, such as the pons or occipital cortex, suffer much less damage with age. It should be noted, however, that even though


218

these regions maintain greater structural integrity, functioning may decline. For example, the primary visual cortex exhibits decline in sensitivity and increases in response time (Crognale et al., 2001). In apparent response to increasing decline in structure and function, the brains of older adults appear to accomplish cognitive functions somewhat differently than those of younger adults, as indicated by brain imaging during task performance. Thus, the aging brain appears to compensate for changing physiology and structure by rerouting or reorganizing functions, sometimes recruiting more of the brain to accomplish tasks that, in younger years, required fewer regions. The multitude of changes in the underlying structure, neuropathology, and function of the brain are presumed to cause changes in four general cognitive mechanisms widely considered responsible for age-related declines in a wide range of specific cognitive tasks. These are (1) sensory function, (2) speed of information processing, (3) working memory functioning, and (4) inhibitory processes. Each of these is conceptualized as a form of “cognitive resource” or “processing resource,” with the combination viewed by some as the best index of overall cognitive resources (Salthouse, 1991). The nature of these changes is reviewed in the following sections. Perception and the Aging Senses As the preceding section made clear, age-related changes in cognitive functioning are clearly linked to underlying changes in brain physiology. Among the most fundamental of these functions is sensory perception. External information can only enter the brain for further processing through the gateway of the senses. When these senses fail, so will the accuracy of perception and, in turn, that of memories and judgments based on the flawed perceptions. To the extent that information is encoded poorly or incorrectly, it cannot later be remembered completely and accurately or used appropriately. In this sense, then, perception may be regarded as the most fundamental sense in which cognitive processing fails with age. Clinically significant and subclinical reductions in hearing and vision increase markedly with age, beginning in the 40s (see reviews in Kline and Scialfa, 1996; Schneider and Pichora-Fuller, 2000). Of particular importance, whereas loss in visual acuity can often be compensated for through corrective lenses, in most cases hearing aids are limited in effectiveness for correction of hearing loss (Lubinski and Higginbotham, 1997). Even so, uncorrectable losses in vision and hearing become more common with increasing age. Some have argued that sensory functioning is a crude proxy for brain integrity and suggested that sensory function should therefore mediate all remaining cognitive abilities (Lindenberger and Baltes, 1994, 1997). In two studies, the authors tested subjects ranging from 25 to 103 years of age. Simple tests of auditory and visual acuity were strongly correlated with a wide range of other cognitive abilities, such as speed of processing, reasoning, memory, world knowledge, and verbal fluency. Furthermore, sensory functioning appeared to function as a more fundamental cognitive resource, and/or as a better indicator of underlying brain functioning, than the other cognitive mechanisms to be discussed later, in that the sensory tests mediated age-related variance in the other measures of cognitive function. Taken together, vision and hearing accounted for 49.2%


219

of the total variance in the overall index of cognitive functioning and 93.1% of the agerelated variance. Other researchers (Mariske et al., 1997; Salthouse et al., 1996) have since obtained similar findings (see the review in Schneider and Pichor a-Fuller, 2000). Dulay and Murphy (2002) argued that such effects should obtain for all senses if the common cause hypothesis is valid. Consistent with this supposition, brain reactivity to olfactory stimuli declines (Cerf-Ducastel and Murphy, 2003), olfactory threshold rises, and odor detection and identification and odor memory decline across the lifespan (Lehrner et al., 1999; Larsson et al., 2000), and the quality of olfactory functioning predicts general cognitive decline (Dulay and Murphy, 2002). Although such correlational results do not indicate that decrements in cognitive functioning are caused by losses in sensory function, they do indicate that decrements in sensory functioning are powerfully predictive of decrements in other cognitive functions—most likely, as Lindenberger and Baltes (1994) suggested, because sensory function is a better proxy for underlying brain function than the other measures. Not surprisingly, other variables related to physical (and thus brain) decline have been related to cognitive performance. Interestingly, balance-gait accounted for as much variance in cognitive functioning as hearing/vision in the Lindenberger and Baltes studies and may be thus taken as another excellent predictor of cognitive function. Other studies have found a link between grip and lower-limb strength and cognitive performance (see Schneider and Pichora-Fuller, 2000, for review). Recent evidence has also begun to link metabolic processes such as glucoregulation (Messier et al., 2003) to cognitive performance. Hearing Older listeners tend to suffer decline in processing of auditory information (labeled “presbycusis”), the hallmark of which is elevated thresholds for detection of highfrequency (higher-pitched) sounds, due to damage to hair cells at the base of the cochlea (cells on the basilar membrane in the inner ear where high-frequency sounds are coded). In addition, the elderly tend to suffer difficulties in detection of simple, high- and lowintensity stimuli, discrimination of small changes in frequency (pitch) or intensity (loudness), temporal resolution abilities (detecting when one sound stops and another begins), filtering out of background noise, and precision in location of the source of sounds—and thereby to suffer deficits in correct encoding of auditory information in many everyday situations (see Pichora-Fuller and Carson, 2001, for review of physical causes of these declines). In practical terms, perhaps the greatest deficit resulting from hearing loss is accuracy in speech perception. Discrimination between consonants becomes difficult for the elderly, particularly the higher frequency consonants such as “f,” “g,” “s,” “t,” and “z” (Sataloff and Vassallo, 1966). Generally, speech begins to sound “fuzzy,” with words appearing to run together even when the speech is loud enough to hear. These problems are exacerbated by general decline in cognitive processing efficiency that slows processing of speech, sometimes to the point that earlier words or sentences have dropped from working memory before the rest are fully understood (see Nussbaum et al., 2000). Speech processing also becomes more difficult when others speak at a fast rate; when


220

there is external noise, distortion, or reverberation in the target speech; and when the listener is under stress (see Kline and Scialfa, 1996, for review). Problems with hearing in conversation require a louder and slower conversational style and can result in such poor outcomes as inaccuracy and misunderstanding, stress in both parties, inefficiency, slowness and repetition, and avoidance of further conversation on the part of both parties. Finally, because we also rely on hearing to produce our own speech sounds, hearing loss can result in declining clarity in production of speech (Nussbaum et al., 2000). Despite these problems, three quarters of those who might benefit from hearing aids fail to use them, and among those who do, many problems remain unsolved. Hearing aids function to aid in hearing low-intensity sounds, such as consonants in speech, but also tend to amplify irrelevant and unwanted background sounds that interfere with hearing of the target sounds. This leads users to experience difficulties processing specific sounds— such as speech—in auditorily complex situations involving multiple simultaneous sources of sound (see the review in Schneider and Pichora-Fuller, 2000). Even those who show hearing sensitivity within normal limits and hear well in quiet conditions may suffer age-related inability to hear well in noisier circumstances (see reviews in CHBB, 1988; Willott, 1991). Individuals suffering from cochlear pathologies may also find speech discrimination difficult even in quiet circumstances (see Humes, 1996, for a review); in addition, they have enhanced difficulties with frequency selectivity (ability to detect smaller differences in pitch) and abnormal growth of loudness as signal intensity increases (see reviews in Moore, 1989; Pickles, 1988; Schneider, 1997). Of particular importance, even when speech is understood under noisy or other perceptually difficult conditions, it is remembered less well (see Schneider and PichoraFuller, 2000, for review) by older as well as younger adults. Thus, it is important to consider the external context in which conversation was heard, along with the internal processing difficulties created by the age of the listener, when attempting to assess accuracy of processing of, and memory for, speech. As suggested by Rabbitt (1991), difficulty in hearing may affect memory because the more resources one must devote to decoding the speech signals, the fewer are available to think about what is heard—i.e., to rehearse the material or to process it deeply and elaboratively—which would otherwise facilitate memory (see below). Specific hearing deficits may derive from degeneration of the ear (such as losses in threshold sensitivity to various frequencies and the cochlear pathologies noted earlier) or brain functioning. Even when a person has a normal audiogram assessing threshold sensitivities primarily determined by the ear, age-related deficits detected by auditory brainstem responses (ABRs) may independently cause decline in auditory functioning under more difficult circumstances. For example, deficits of synchrony in neural firing may cause problems with frequency discrimination, temporal discrimination, localization, binaural unmasking (ability to focus on a particular sound when multiple sounds enter each ear from multiple but different origins), or speech perception (see Hellstrom and Schmiedt, 1990) Again, such losses may be independent of threshold sensitivity. For example, difficulty in temporal discrimination (e.g., detecting when a sound has stopped and another begun) may cause difficulties in speech perception, even though the person hears loudness adequately. Older adults experience greater difficulty with gap detection


221

and with detection of differences in duration of sounds (see Fozard and Gordon-Salant, 2001; Schneider, 1997; Schneider and Pichora-Fuller, 2000; Strouse et al., 1998). Finally, difficulties in speech perception can be compensated for in some degree by lip-reading and general use of nonverbal cues to clarify meaning. However, because vision also tends to decline with age, the older person is often further hampered by inability to read and employ nonverbal information effectively (Nussbaum et al., 2000). Some have argued that difficulties in comprehension of speech and conversation are caused by the increased load imposed on working memory by the enhanced effort that older listeners must devote to accurate hearing (Rabbitt, 1991). The more working memory resources must be devoted to detection and identification of the sounds, the less working memory capacity is available for processing and storage of meaning. If so, removing the burden of sound detection and identification should improve cognitive processing of the speech. Therefore, several researchers have addressed the question of whether age-related declines in word recognition, immediate recall of verbal materials, and discourse comprehension can be eliminated or reduced by improvements in hearing. These studies addressed this issue by adjusting volume to exceed the threshold sensitivities of young and old adults by a constant amount—in quiet and noisy conditions—so that each person could recognize the words accurately, or by hampering the hearing of younger adults to simulate the sensory difficulty of older adults. Generally, these results indicate that when the sheer difficulty of hearing the words adequately is equated between young and old, age-related differences in comprehension and memory are minimized or eliminated (see the review in Schneider and Pichora-Fuller, 2000). Summary Clearly, age is associated with a variety of hear ing-related losses. Perhaps most important for the forensic arena are losses in accuracy of speech perception, elevation of thresholds for sound detection, and increasing difficulty in sound localization. However, when the possible role of hearing in forensic circumstances is considered, it is vital to remember that the specific deficit of interest must be directly tested. As shown by a host of studies (see reviews in Moore, 1989; Pickles, 1988; Schneider, 1997; Schneider and Pichor a-Fuller, 2000; Hellstrom and Schmiedt, 1990; Humes, 1996), a person may have normal or reasonably normal threshold sensitivity as measured by standard hearing tests while suffering several serious losses in other auditory functions or, indeed, in threshold sensitivity under circumstances more complex than those presented by a standard hearing test. Of particular interest would be the effects of aging on speech perception. Although the person possessing normal threshold sensitivity may hear the speech as loud enough, some of the fine detail encoded at higher frequencies would be lost even in quiet circumstances. In louder circumstances, low frequencies tend to be masked by the background noise, and gap perception becomes more difficult, thereby blurring distinctions between words. Gap perception also becomes more difficult for older listeners with faster speakers. Difficulties with binaural processing can impair ability to localize the source of sound and to focus clearly on one speaker amid a crowded noisy room. Thus, for any given older person of normal threshold sensitivity or not, specific deficits of relevance to the situation in question must be individually established.


222

Vision Correctable refractive errors constitute the most commonly recognized age-related form of decline in visual functioning. However, even with such correction and among those with both good and bad general health of the eyes, visual acuity (per the “Snellen” measure in which subjects must identify letters of varying size at a specified distance) begins to decline after 45 years of age. Moreover, declines occur in other aspects of visual functioning, such as contrast sensitivity (ability to discern luminance differences for targets of various sizes) and contrast discrimination (ability to discern differences between objects in amount of contrast to background). Older adults may need two to three times more contrast to be able to see small- or medium-sized targets (such as the characters that comprise most text). Other aspects of visual functioning decline as well, including dark adaptation (speed with which rod-based black-and-white vision functions clearly after light is reduced or eliminated), perimetric fields (gradient of differential ability to see (threshold sensitivity) in the foveal vs. peripheral regions of the retina), color vision, and stereopsis (use of both eyes for depth perception). Pathologies of the eye also increase with age, including cataracts, glaucoma, and macular degeneration (see Michaels, 1993, for a review). Overall, visual representations tend to become less precise or accurate—more so under conditions of poor lighting or poor stimulus contrast (for a review, see Faubert, 2002; Schneider and Pichora-Fuller, 2000; Pichora-Fuller and Carson, 2001; also see Enoch et al., 1999, for discussion of some visual functions that do not decline with age). Visual perception also becomes slower, so longer viewing times are necessary for older adults to perceive and recognize a visual image (Jacoby and Debner [2001], discussed in Jacoby et al., 2001). As with hearing, some age-related changes in vision are the result of changes in the eye: • Retinal image quality (visual acuity) is reduced and hence vision is blurred by changes in light diffraction through the cornea. • Thresholds for rod (low light level) and cone (high light level) vision are raised as less light passes through yellowed and more opaque lenses and narrowed pupils. • Poorer rod vision (including poorer vision under low illumination and restricted effective field of view) results from decreasing numbers of rods (Curcio, 2001). • Poorer cone (high light level) vision results from decreased sensitivity of the cones to the light that does pass through to the retina (Werner, 1998). • Presbyopic difficulties in adjusting focus to accommodate distance follow from increases in stiffness of the lenses. • Inefficient transduction of light to neural impulses results from narrowing of the arteries of the retina. • Decrease in retinal contrast is caused by increased light scatter by the cornea (resulting in decreased ability to detect and recognize common objects, including traffic signs). • Blue-yellow color confusions are caused by greater absorption of and therefore reduced sensitivity to shorter (blue) wavelengths (and discrimination of colors of the same hue and desaturated colors such as pastels becomes more difficult).


223

Generally, less efficiency and precision in processing visual input results. (See Faubert, 2002; Fozard and Gordon-Salant, 2001; Kline and Scialfa, 1996, 1997; Schneider and Pichora-Fuller, 2000; Scialfa, 2002; Spear, 1993.) Essentially every structure in the primary visual pathway changes with age. Density in the retinal ganglion cell layer decreases, particularly outside the macular region. Similarly, cell and synapse density, number of synapses, and myelin sheath integrity decrease in the visual cortex. Similar changes occur throughout the brain and predict a variety of specific indices of cognitive decline (Faubert, 2002; Peters et al., 1996, 2000; Nielsen and Peters, 2000; Scialfa, 2002). Great individual differences exist within age groups, with poorer vision associated with greater exposure to sunlight, which is associated with aging of the lens and the retina (see Werner, 1998, for review). There is substantial evidence of the practical effects of declining visual functioning of the eye on activities of daily living. Difficulties with visual acuity, depth perception, motion perception, and vision in low illumination or shadowed areas cause difficulty in managing daily routines. We illustrate a number of these effects in the context of driving. Visual Acuity Visual acuity (the clarity of an image) is especially important for night driving (Eby et al., 1998; Panek et al., 1977), but declines substantially with age, particularly in conditions of low illumination. Thus, visual acuity—at least for drivers (and especially older drivers)—tends to be worst when it is most needed. The previously referenced changes in lens color and opaqueness, combined with narrowing of the iris and often cataracts on the corneas, reduce the amount of light entering the eye. Because the sharpness of an image is dependent upon illumination and, to some degree, on color and because color vision drops out with sufficient drop in illumination, overall visual acuity will decline with decline in illumination entering the eye. Furthermore, neural processing mechanisms adapt to lower illumination by, in essence, trading temporal and spatial acuity for sensitivity. The aging eye causes some degree of reduction in illumination under all circumstances. However, this essential reduction renders the aging person more vulnerable to loss of acuity in external conditions of low illumination, such as nighttime. Indeed, older persons have substantially greater difficulty with night vision (and night driving), as well as more difficulty (time needed) in recovering from glare (see Klavora and Heslegrave, 2002). Motion Perception Motion perception is dependent upon a variety of visual factors, including • Ocular musculature adequate to track objects smoothly • Contrast sensitivity (ability to see differences in luminance between target and background) • Stereopsis (use of both eyes for depth perception) • Temporal contrast sensitivity (ability of the eye to see changes occurring over time at different rates, such as fast vs. slow flickering patterns) • Backward masking (the ability of new stimuli to obscure the afterimage of previously seen objects)


224

These factors, as well as others, decline with age. Thus, not surprisingly, motion perception (dynamic vision) tends to be impaired in the elderly (Tran et al, 1998; Willis and Anderson, 2000; Wist et al, 2000, Wojciechowski et al., 1995) in foveal and peripheral vision, although more so in the foveal area (see reviews in Schneider and Pichora-Fuller, 2000; Faubert, 2002; Scialfa, 2002). Because the oculomotor system (eye movement) is slowed with age, older adults suffer impairments of tracking speed—i.e., greater difficulty tracking objects moving at high speeds and when there is a high degree of relative motion between the target and observer (Scialfa et al., 1988). Furthermore, older adults suffer deficits in visual marking for moving stimuli. Visual marking refers to the top-down inhibition of attention to or processing of old stimuli already in the visual field (such as cars already in the stream of traffic) so that processing of new visual information (such as a car entering the traffic stream) can be facilitated (Watson and Humphreys, 1997). This visual marking is assumed to be advantageous, in that it facilitates the observer’s ability to maintain an upto-date representation of the world, which in turn would provide the most useful basis for action decisions. As we shortly review, older adults suffer deficits in attentional control, particularly in controlled inhibition of attention to undesirable or irrelevant stimuli. Consistent with this general deficit, older adults manifest deficits in visual marking for moving stimuli (although they do so less consistently for stationary stimuli; see Kramer and Atchley, 2000; Watson and Maylor, 2002). That is, they show greater difficulty locating new stimuli amid a constellation of already present stimuli. Age-related declines in marking ability are relevant for a variety of everyday human factors issues (see Rogers and Fisk, 2000), such as driving, instrument monitoring, air traffic control, and so on). In practical terms, deficits in motion perception play a substantial role in automobile accidents among the elderly (Staplin and Lyles, 1991). Difficulties in motion perception render the elderly subject to errors in speed, distance, time-to-collision, and gapacceptability judgments, which tend to result in misjudgments in complex traffic situations, including failure to yield the right of way. Such failures, in combination with generally slower responses, are responsible, for example, for the tendency of the elderly to be involved in accidents in which they turn left in front of an oncoming vehicle (see Klavora and Heslegrave, 2002, for review). Older persons also become less sensitive to (less likely to notice changes in) temporally modulated (changing across time) stationary patterns such as flickering lights, as well as to spatial movement of objects (see Schneider and Pichora-Fuller, 2000). While a person is driving, for example, this insensitivity can result in failure to notice a vehicle begin to move, changes in traffic lights, or other important changes in position. Visual Search Visual search refers to the ability to identify a target object or objects in a field of distractors. Depending upon the complexity of the scene, the number of distractors, the nature of features distinguishing the target from distractors, and features of the context such as fog, mist, or poor illumination, accuracy and speed with which the target is located can be compromised. Correct discrimination between stimuli requires comparing a given stimulus with internal models of correctness and incorrectness. Reduction in


225

short-term visual memory capacity will limit ability to discriminate accurately and rapidly. Generally, age is associated with poorer overall performance and greater reactivity to features that tend to compromise performance (Burton-Danner et al., 2001; Scialfa and Joffe, 1997; Madden et al, 2002; Speranza et al., 2001). Indeed, older adults report difficulties with everyday activities relying on visual search, including such problems as distracting objects and events, cluttered visual scenes, and insufficient time to find target objects (Kosnik et al., 1988; Kline et al., 1992). Safe driving often requires the ability to search complex traffic scenes rapidly in order to locate critical objects—such as signs, traffic lights, pedestrians, turnoffs, and address markers—and respond appropriately. Therefore, age-related slowing and increased susceptibility to distractors render older drivers less able to locate critical objects. For example, older drivers are less able to locate traffic signs successfully, particularly in visually cluttered scenes (Ho et al., 2001; Schieber and Goodspeed, 1997) and less able to read them successfully, particularly at night (Sivak et al., 1981). Older adults tend to scan traffic scenes differently, using more fixations, less evenly distributed fixations, and more repeat fixations to a specific target (Maltz and Shinar, 1999) and require more eye movements to find a target such as a sign (Ho et al., 2001). Thus, older adults are less efficient and less accurate in traffic-related visual searches. Peripheral Vision Peripheral vision declines with age, resulting in less likelihood of detection of objects in the peripheral field, less accurate discrimination between objects in the peripheral field, and less accurate motion detection. One study (Johnson, 1986; see also Panek et al., 1977) showed that the range of the visual field declined from 180° among younger adults to 140° beyond age 70. A related index of peripheral vision, the “useful field of view,” or UFV, comprises the limits at which the observer can no longer locate or identify objects or targets in the visual field (tested under conditions in which eye movement is prohibited). (Coeckelbergh and colleagues [2004] discuss age differences for an alternate measure, AFOV, or attended field of view, which tests for target search in circumstances permitting eye movement.) UFV has become popular among many investigators, particularly those concerned with predicting driving performance and accidents (Owsley et al., 1991; Sekuler et al., 2000). Useful field of view is assumed to decrease with age as a function of the three underlying cognitive processes of decreased ability to divide attention, decreased ability to ignore distractors, and reduced processing speed (Ball et al., 1990). UFV has proven to be a strong predictor of accident involvement. Those with poor peripheral vision are believed to have two to six times the accident rate of those with normal peripheral vision (Morgan and King, 1995; Ball et al., 1993; see reviews in De Raedt and PonjaertKristoffersen, 2001; Klavora and Heslegrave, 2002; Park and Gutchess, 2000). However, this predictive value appears to be due to the divided attention component and is unrelated to the other two (Owsley et al., 1998). In light of the role of capacity for divided attention in determining UFV, it is not surprising that the presence of distractors effectively reduces UFV (Sekuler et al., 2000), particularly for older adults in whom capacity for divided attention is reduced (see below). Such distractors can include irrelevant objects in the peripheral field, greater


226

similarity between relevant and irrelevant objects, and engagement in multitasking (such as eating or talking on the phone while driving). Depending upon circumstances, older adults’ UFV can become restricted to only one third of that of younger adults in similar circumstances (see Schneider and Pichora-Fuller, 2000, for review), with obvious consequences for detection of pertinent traffic signs, obstructions, and approaching vehicles. Fortunately, declines in UFV can be successfully decreased with normal elderly clients and stroke victims with a program called “visual-motor useful field of view” (VM-UFOV) using the “Dynavision” apparatus (a device used for training athletes and others to enhance dynamic vision, reaction time, movement time, hand-eye coordination, peripheral vision, and the ability to divide attention and make decisions under stress). The device is in use in a number of rehabilitation and medical centers across the country (see the review in Klavora and Heslegrave, 2002). Summary As with hearing, a number of age-related declines occur in the visual system. Also, as with hearing, in forensic contexts it is vital to test the specific suspected deficits relevant to the situation at hand. The most common vision test is the Snellen test of visual acuity (identification of letters of varying sizes from fixed distances). However, Snellen acuity is measured in conditions involving no motion but full illumination. For this reason, it is a poor predictor of other aspects of processing, including impaired processing of motion, tracking of high-speed objects, vision in low illumination, and the UFV index that is so predictive of peripheral vision performance and automobile crashes. For traffic accidents involving the elderly, impairment in one or more visual skills may well play a causal role. The elderly appear to be aware of visual deficits to some degree. They drive shorter distances; drive more slowly; and decrease their night, highway, and rush-hour driving. However, they are still at greater risk of accidents, per mile driven, at a level comparable to that of 15- to 25-year-olds (see Klavora and Heslegrave, 2002; De Raedt and Ponjaert-Kristoffersen, 2001). Although thus far we have illustrated with examples regarding driving, impairments in vision contribute widely to a variety of accidents and are relevant for ergonomic design in a number of situations (see Echt, 2002, for an excellent review of the effects of agerelated declines in vision on reading and the use of computer screens) and to issues of perception and memory in forensic contexts involving witnesses and jurors (explored later). Processing Speed Substantial evidence exists in support of the proposition that decreased speed of processing accounts for age-related decline in performance of all cognitive tasks. Salthouse (1991,1996) proposed that two general mechanisms underlie the relationship between speed of processing and cognitive performance. First, Salthouse suggested that “the time to perform later operations is greatly restricted when a large proportion of the available time is occupied by the execution of earlier operations” (p. 404), which he referred to as “the limited time mechanism.” Second, he suggested that “the products of


227

earlier processing may be lost by the time that later processing is completed” (p. 405); he referred to this as the “simultaneity mechanism.” These two mechanisms can together result in slower, but still accurate, performance or performance failures. Earlier stages of processing may be slowed but accurate, but can result in performance failure if this slowing causes the person to fail to reach later stages. Salthouse (1991, 1996) believed the effects of slow processing to affect all cognitive tasks, including those with no obvious speed component. Salthouse (1996) reviewed an impressive array of evidence that performance on “perceptual speed tasks” is an excellent predictor of performance on a wide range of other specific cognitive tasks. Perceptual speed tasks require the person to make rapid perceptual same-different judgments about pairs of digit or letter strings, or two similar symbols. Speed of processing is measured by the number of correct same-different comparisons made in a fixed period of time. Generally, Salthouse (1996) demonstrated that age-related variance on other specific cognitive tasks is reduced or eliminated when controlling for variance in processing speed. Despite the success of perceptual speed tasks as predictors of performance on other, even apparently non-speed-related, cognitive tasks, the processing speed theory of agerelated decline has not been universally accepted, as suggested by Bashore et al. (1997), Fisk and Fisher (1994), Hartley (1992), and Perfect (1994). These authors suggest that no strong tests have been conducted regarding the issue of whether age differences on the wide range of specific tasks are attributable to the general factor of processing speed vs. process- or task-specific factors and no agreement on how best to test the relative influences of general vs. specific factors has been reached. Notwithstanding such unresolved issues, however, it is clear that speed of processing does decline with age (Salthouse, 1996). Processing Speed and Older Drivers A person trying to follow directions or a map can often be required to make quick choices and execute them in a complex and rapidly moving traffic situation—frequently in the midst of complex and rapidly changing contexts such as shopping districts replete with shops, pedestrians, and cyclists. Thus, impairments of processing and response speed can seriously compromise the driving safety of older adults, who tend to perform best under simple conditions, when not rushed. Increased difficulties of older adults in perception, choosing responses, preparing and executing responses, and so on tend to result in slower reaction times that, in turn, can cause accidents. Stelmach and Nahom (1992) reviewed changes in all aspects of motor performance, including selection, preparation, and execution of responses, among others. Consistent with a large body of research showing stronger age-related declines of all kinds under more complex or demanding conditions, decrements in speed were particularly strong under conditions of task complexity and complex arrays of stimuli—for example, during complex driving situations or responding to pending collisions. Increased decrements under demanding conditions are presumably due to increased demands upon working memory (see below). In this light, not surprisingly, response selection (which presumably places the greatest demands upon working memory) was the most age-sensitive factor studied.


228

In addition to slowing of cognitive processing with age, physical behaviors and reactions are slowed as well. Older drivers exhibit slower movements, due in part to reduced joint flexibility, joint deterioration, arthritis, etc. The U.S. Department of Transportation maintains a web site with a variety of materials related to how physical condition and cognitive functioning are related to driving safety (http://www.nhtsa.dot.gov/people/injury/olddrive/), which includes an excellent research report (April, 1999) of the data documenting these relationships (http://www.nhtsa.dot.gov/people/injury/olddrive/safe/). Also see a recent meta-analysis of the relationship between neuropsychological functioning and driving ability by Reger and colleagues (2004). Working Memory Craik and Byrd (1982) suggested that age-related deficits in processing resources are best measured by working memory tasks, rather than perceptual speed tasks. Working memory is conceptualized as the total processing resources available at any given moment to perform online cognitive operations (Baddeley, 1986) and is measured through tasks in which the subject must store and process information simultaneously. For example, a person might be asked to perform a computational span task, in which he or she must solve a series of equations while remembering the second number in each equation. Working memory is measured by how many equations the subject solves correctly while accurately remembering the target number. Well-documented age-related declines in working memory reflect decreasing ability to hold multiple items of information in mind while performing one or more cognitive tasks (see a recent meta-analysis by Verhaeghen et al., 2003) Age-related deficits appear most strongly in cognitive tasks imposing stronger processing demands on working memory, such as tasks requiring simultaneous consideration of multiple items of information, difficult manipulation of information (e.g., reasoning, problem-solving, mathematics), multitasking requiring divided attention, or working in distracting conditions that require filtering of irrelevant information. These deficits are negligible for those requiring less effortful processing (Craik and Jennings, 1992; Light, 1991). Furthermore, effective processing load or capacity is affected by age-related physical conditions. For example, chronic pain (e.g., Grigsby et al., 1995) is related to deficits in information processing, as are such conditions as hypertension, diabetes, impaired thyroid function, and others (see the review in Nilsson and Soderlund, 2001). In practical terms, declining capacity of working memory is reflected in age-related declines in performance of more processing-intensive tasks, in multitasking (see the review in Kemper et al., 2003), and in memory for information encountered during processing-intensive tasks or in complex or distracting circumstances. Of particular pertinence in legal settings, deficits in working memory are widely considered to result in failures of “associative learning” or “binding” (learning associations between items such as between word pairs or, more practically, between a person’s face and the context in which that person was encountered). Failures of this kind are considered to contribute to such dramatic memory errors as mistaken eyewitness identifications (Wells et al., 1998) or development of false memories of sexual abuse (Loftus and Ketcham, 1994). Furthermore, failures of working memory can compromise the ability of older jurors to


229

process trial input and to remember it and reason adequately to arrive at an appropriate verdict. Fortunately, declining working memory can be compensated for through “environmental supports” (Cherry et al., 1996; Craik and Byrd, 1982; Park et al., 1990). Essentially, such supports reduce the load on working memory by providing external cues (such as pictures or lists) to make it unnecessary to hold all information in working memory in order to use it for processing or judgment. Working Memory and Older Drivers Decline in working memory capacity is widely considered to contribute substantially to age-related accidents. As working memory declines, the person cannot manage the demands for divided attention and complex decision making that tend to characterize complex driving situations. Therefore, as noted earlier, speed of response slows most dramatically for the elderly under complex, stimulus-intensive driving conditions. In these conditions, many distractors place demands on allocation of attention, and rapid responses to multiple moving stimuli must be selected, planned, and executed rapidly. Older drivers are prone to experiencing cognitive overload on the road and fail to divide attention successfully in order to process the full range of relevant stimuli such as other vehicles, pedestrians, traffic lights and signs, road hazards, and so on. They encounter problems in discriminating between relevant and irrelevant stimuli in complex situations and have difficulty ignoring irrelevant or meaningless information in favor of focus on relevant and important objects. Thus, older drivers have difficulty changing lanes, turning, passing, driving in reverse, and comprehending traffic sign symbols. They tend to commit such attention-related errors as running red lights or stop signs, failure to yield right-of-way, and committing traffic violations during turning maneuvers. Perhaps because the driving situation is more complex at intersections, older drivers are more likely to experience multivehicle, side-impact collisions at intersections, particularly when turning left. They also experience accidents due to tendencies to turn mistakenly prior to the intersection, commit illegal turns, disregard traffic signals, and rear-end others, due to failure to notice change in motion (Stamatiadis et al., 1991). Finally, all of the consequences of limitations in working memory are enhanced by anxiety and stress, which further limit working memory capacity, narrow the focus of attention, and thereby enhance accident potential (see De Raedt and PonjaertKristoffersen, 2001; Klavora and Heslegrave, 2002; Preusser et al., 1998). Generally, Hakamies-Blomqvist (1994) found errors of attention to be the most important cause of fatal accidents involving older drivers. Reflecting failures of attention and processing,. 44% of older drivers in another of this author’s studies (Hakamies-Blomqvist, 1993) had not noticed any danger prior to their accident, whereas only 26% of younger drivers failed to notice danger. Inhibitory Functions A particularly interesting proposition was introduced by Hasher and Zacks (1988; see also Lustig et al., 2001; Persad et al., 2002), who argued that optimal performance requires control over attention to irrelevant information. Furthermore, they provided


230

evidence that age is associated with declining ability to focus on target information while inhibiting attention to irrelevant information. Others have since provided neuroimaging and neuropsychological evidence to suggest that age-related declines in attentional control may be the result of age-related changes in the frontal lobes of the brain (Moscovitch and Winocur, 1995; Shimamura and Jurica, 1994; West, 2000). Successful inhibition in older adults is associated with activation of areas of the brain beyond those activated by younger adults, suggesting that older adults must compensate for age-related difficulties in inhibition with supplementary recruitment of brain resources (Nielson et al., 2002). In contrast to theorists emphasizing age-related reductions in speed of processing, the inhibitory control framework emphasizes changes in efficiency of processing. Efficient processing is presumed to require strong attentional control so that attention is occupied only by goal-relevant information. Inefficient inhibitory processes are assumed to permit goal-irrelevant information to command attention and enter working memory, thereby detracting from processing of goal-relevant information. Moreover, impaired inhibitory processes allow the irrelevant information to persist in inhabiting working memory, thereby affecting cognitive performance for some time after exposure. As the attentional draw of the irrelevant information becomes stronger, so does impairment of the focal task (Zacks and Hasher, 1997), particularly when inhibitory processes are compromised by age or other factors. Hasher and colleagues suggest that failures of inhibitory control affect present performance and future memory. Present performance depends upon the ability to concentrate attention on task-relevant information and to ignore irrelevant distractions. Furthermore, memory is dependent upon attention (Schacter, 2001). Memory follows the focus of attention and depends upon the amount and quality of attention. Thus, to the extent that failing inhibitory processes alter the nature of attentional processes, they will alter performance and memory. Proponents of the inhibitory deficit framework view a wide variety of age-related declines in cognitive processing as the result of failing inhibitory processes. Although substantial debate exists regarding whether these deficits are the result of generalized age-related slowing vs. specific deficits in inhibitory capacity (McDowd et al., 1995; Verhaeghen and Meersman, 1998), strong evidence exists of age-related decline in ability to control attention and to engage more generally in controlled rather than automatic processing strategies (see reviews by McDowd and Shaw, 2000; Rogers, 2000). Older adults are less able to sustain attention or maintain focus on relevant information. Problems of this nature are reflected in poorer performance on “vigilance tasks,” such as those of air traffic controllers, radar monitors, or product defect screeners, in which the person must sustain attention for long periods and successfully focus on relevant targets only. Older adults are also more distracted by irrelevant information (current distractors or irrelevant information from long-term memory). This age-related susceptibility to current distractors results in poor immediate performance, as well as shallow encoding and later poorer memory. Memory and performance are further impaired by age-related failures to inhibit attention to irrelevant past information. For example, Lustig and colleagues (2001; see also Bowles and Salthouse, 2003) reviewed evidence that older adults are more susceptible to “proactive interference” effects. That is, previously learned information is


231

more likely to interfere with learning (memory for) new information. Presumably, this difficulty in learning the new information is caused by the age-related inability to suppress activation of (and therefore interference from) the old information (see Hedden and Park, 2001, for similar age-related enhancement of “retroactive interference” effects). Similarly, failure to inhibit attention to irrelevant information hampers problemsolving and decision making. Finally, similar inhibition deficits are reflected in agerelated differences in automatic vs. controlled processing in that older adults are more susceptible to the effects of positive priming and more likely to engage in schematic and automatic processing (see reviews by McDowd and Shaw, 2000; Rogers, 2000). In later sections, we explore the implications of these processes for older witnesses and jurors. Circadian Patterns and Inhibitory Capacity Hasher and her colleagues have shown that inhibitory abilities vary with the time of day in older and younger adults. However, whereas younger adults’ abilities are lowest in the morning and peak in late afternoon, those of older adults peak in the morning and wane throughout the day (May and Hasher, 1998; May et al., 1993, Yoon et al., 2000b). May (1999), for example, showed that distractors had greater effects on performance for young and old adults who were tested at their own nonoptimal time of day, compared to those tested at their optimal time. These established differences in circadian arousal patterns can contribute to hourly variation in accidents, performance in job-related settings, or memory and information processing in witnesses and jurors. Stereotype Vulnerability, Stereotype Threat, and Enactment of AgeStereotyped Behaviors A final explanation for age-related decrements in memory and performance has proposed that age differences may be caused in some degree by age stereotypes. Research examining stereotypes of age has revealed positive as well as negative attitudes toward aging and the old. However, in contexts in which cognitive competence is at issue, the stereotypes tend to be overwhelmingly negative (Hertzog et al, 1999; Kite and Johnson, 1988; Levy, 2003), including expectations of declining cognitive performance and memory (see reviews in Levy and Langer, 1994; Nelson, 2002; Stone and Stone, 1997). Ryan (1992) suggested that these stereotype-based expectations of poor cognitive performance may become a self-fulfilling prophecy through indirect impact on decreased effort, lesser use of adaptive strategies, avoidance of challenging situations, or failure to seek medical help as a result of improper attribution of cognitive failures to age, rather than to underlying disease or drug side effects. Similar processes may cause other stereotype-based self-fulfilling prophecies (see Wheeler and Petty, 2001, for review of mechanisms by which stereotypes affect performance). Indeed, research has suggested that childhood exposure to the negative images of old age present in fairy tales, television, and everyday conversations in America can influence one’s level of activity and alertness in old age (Langer et al., 1988; Rodin and Langer, 1980). A series of subsequent studies has supported Ryan’s (1992) reasoning. Levy and Langer (1994) reasoned that age-related deficits in memory performance would be


232

predicted by beliefs in age-related memory stereotypes. Therefore, they compared the performance of Chinese mainland participants to that of hearing and deaf Americans. Based on research indicating that those from the Chinese mainland and deaf Americans would have less exposure to negative age stereotypes, they predicted (and found) that, although younger participants from all groups performed equivalently, older American hearing participants performed more poorly than older American deaf or older Chinese participants. In China, no age differences were found on any of the four separate memory tests that commonly show age-related deficits in American participants. Overall, positive views of age also predicted better memory performance, and path analyses indicated that memory performance across cultures was mediated by views of aging. In a subsequent conceptual replication, Yoon and colleagues (2000a) found a difference on two of four memory tasks between Anglophone Canadians and Chinese Canadians who had recently emigrated from Hong Kong. However, younger adults outperformed older adults on all tasks. This age difference was reduced for Chinese Canadians on two tasks, but not eliminated. Moreover, unlike Levy and Langer (1994), views of aging did not mediate memory performance. The authors concluded on the basis of these findings, as well as various methodological issues in their own and the Levy and Langer work, that cultural differences in age-related stereotypes cannot yet be considered a clear contributor to age-related decline in memory. Levy (1996) later employed a subliminal priming methodology to examine the effects of activation of age-related stereotypes on four memory tasks. Older adults consistently performed worse than younger adults. However, exposure of older participants to negative age-related stereotype primes (e.g., senile) resulted in poorer memory performance, whereas exposure to positive (e.g., wise) primes enhanced performance. This difference was not obtained with younger adults, for whom such stereotypes presumably did not seem personally relevant. In a later conceptual replication of Levy’s priming research, Stein and colleagues (2002) found that, although priming negative stereotypes impaired performance of older adults who were unaware of the primes, priming positive age stereotypes did not enhance performance. Like Levy (1996), the authors found no priming effects in younger adults. Levy et al. (2000) argued that exposure to negative age stereotypes can cause cardiovascular stress (which could in turn affect cognitive functioning). The authors exposed older individuals to subliminal presentations of positive or negative age stereotypes, followed by mathematical and verbal challenging tasks (stressors). Those exposed to negative primes exhibited increased physiological responses to stress— including skin conductance, systolic and diastolic blood pressure, and heart rate— compared to those exposed to positive primes; these effects persisted at a second measurement one half-hour after the interventions. Negative primes also induced poorer performance on the math test than positive primes did. Levy and her colleagues went on to explore the effects of subliminal positive and negative age-related stereotypical primes on other age-related behaviors such as measures of walking performance of gait speed and swing time (time with one foot in the air during walking; Hausdorff et al., 1999; see also Bargh et al., 1996, for effects of negative age stereotype primes on walking speed of young adults) and handwriting (Levy, 2000). In each study, performance was shown to conform to the nature of the primes (but in varying degree).


233

In perhaps the most dramatic of her studies, Levy et al. (1999–2000) examined whether aging self-stereotypes could influence the will to live. The authors exposed young and old participants to positive or negative age-stereotype primes presented subliminally. They then asked participants to read a series of hypothetical medical situations involving fatal conditions and interventions that could prolong life. The scenarios included presentations of disadvantages of the treatment involving financial cost or care-giving responsibilities of family members. Older persons who were exposed to positive aging stereotypes tended to accept the life-prolonging medical interventions, regardless of financial or caregiving costs, whereas those exposed to negative aging primes tended to reject such interventions although no effects of the primes were observed among younger participants. Levy subsequently demonstrated that positive self-stereotypes of aging measured earlier in life appear to predict the will to live and actual health and mortality in older age (Levy et al., 2002a, b). Finally, borrowing on the stereotype threat paradigm (Steele, 1997), Rahhal and Hasher (1998) examined age-related differences in performance of tasks that participants had been led to believe were memory related (i.e., relevant to the negative age-cognitive functioning stereotypes) or memory unrelated (unrelated to negative age-related stereotypes). Presumably, older adults perform worse when they believe the task reflects abilities (i.e., memory) for which negative age-related stereotypes exist than if it reflects one for which there are no such stereotypes. Indeed, although the tasks were held constant, with memory instructions, younger adults outperformed older adults, whereas with “knowledge” instructions, there were no age differences. Similarly, Earles and Kersten (1998) found that younger adults performing a series of memory tasks tended to remember the items they found most difficult, whereas older adults tended to remember those they found relatively easy. Unfortunately, the effects of memory-related stereotype threat appear to be strongest for older persons who place more value on their memory abilities. Hess and his colleagues (2003) had older and younger adults perform memory tasks under high or low conditions of stereotype threat. High threat affected the performance of older more than that of younger subjects, and this effect was strongest for older adults who valued memory performance the most. In part, stereotype threat appears to impair performance through its effects upon working memory. Although this has not yet been shown with respect to age stereotypes, Schmader and Johns (2003) demonstrated that stereotype threat regarding gender and ethnicity impaired working memory capacity, and that reduction in working memory capacity mediated the effect of stereotype threat on women’s math performance. Overall, the results of these stereotyping studies suggest that older adults can react to salient age-related stereotypes by behaving in stereotype-consistent ways (see the review in Levy, 2003). Thus, negative age-related stereotypes appear to contribute to, but not to explain fully, age-related performance deficits. Although salient positive age-related stereotypes may sometimes reduce age-related performance decrements, it is clear that they cannot eliminate them. It is also clear that older adults can suffer greatly from exposure to (or belief in) negative age stereotypes.


234

11.3 Applications to Aging Victims and Witnesses Age-related problems with forensic implications occur at all stages of cognitive processing. We review some of these next. We emphasize the aging witnesses but also discuss implications for generation of accidents. Problems of Perception In order to process and remember information accurately and therefore use it appropriately, the perceiver must first perceive it correctly. Age exerts a major influence at this stage. We have previously reviewed the nature of decline in auditory, visual, and olfactory processes with age and noted some of their effects for aging drivers. We now review several areas of perception that rely on these senses and, therefore, suffer agerelated declines as well. We focus on those that are relevant in forensic contexts. Estimating Time Time estimation is central to everyday life. One must prospectively project time to schedule work and leisure activities such as time to arrive at a destination, cook dinner, perform tasks at work (write a chapter), see a movie, and so on. One must also prospectively estimate much shorter time intervals, such as in those involved in drivingrelated judgments such as time to impact, gap-acceptability judgments, and so on, correctly. Such short-interval judgments are also important for sports-related abilities such as catching a ball. Witnesses are most often confronted with the task of estimating actual time passed— the duration of a particular sound or event (e.g., “How long were you able to observe the perpetrator?”) or the elapsed time between two events (e.g., “How long was the defendant out of the room?” or “How long between when he left and when you heard the scream?”). Generally, estimates of duration tend to be inaccurate and to overestimate actual duration, particularly when the observer is under stress (Loftus et al., 1987; see Block and Rakay for a review). Aging is associated with enhanced inaccuracy (McCormack et al., 2002) and enhanced verbal overestimation of duration (see a meta-analysis by Block et al., 1998). Witnesses are also frequently asked to report the order in which events occurred across short and long intervals. Again, memory for order is often poor across the lifespan, but older adults suffer enhanced impairment (Schmitter-Edgecombe and Simpson, 2001), perhaps because memory for order is dependent upon frontal region functioning (Schacter, 1996). Visuospatial Processing Processing of visuospatial information appears to be particularly susceptible to agerelated decline. Converging evidence indicates that visuospatial cognition is more age sensitive than verbal cognition (Craik, 2000; Jenkins et al., 2000; Lawrence et al., 1998; Myerson et al., 2003a, b; Park et al., 2002; Verhaeghen, 2002). Furthermore, like all aspects of information processing, visuospatial deficits are manifest more strongly as


235

demands upon working memory increase (e.g., Chen et al., 2003), for example, in complex visual environments or when multitasking. This decline is reflected in a number of age-related deficits in driving skills, as we outlined in the earlier sections on visual decline. However, visuospatial processing is central to a wide variety of additional daily living skills, as well as in forensic contexts. First, elderly witnesses suffer the same difficulties in processing of speed, motion, distance, time to impact, and so on as elderly drivers. Thus, whether they are the parties involved in an accident or those witnessing it, older persons are less likely to encode such information accurately or to report it accurately later as witnesses. Second, processing of spatial information is central to object identification, including face identification, as we discuss shortly. In addition to facial identification, object identification is crucial to many professions, from military personnel attempting to distinguish friend from foe to medical personnel reading x-rays or slides. Age-related declines in visual marking, target detection, or visual matching are reflected in performance of these and other identification tasks. For example, Dollinger and Hoyer (1996) found age-related declines between younger (mean age 26.5 years) and middle-aged (mean age 45.7 years) matched novices and medical technologists on domain-specific and general visual recognition performance. Younger participants were faster than older participants on both tasks. Moreover, those tested under dual task conditions were slower and less accurate than those tested under single task conditions—a difference that was greater for older adults. However, expertise compensated to some degree for age-related declines, in that medical technologists suffered fewer age-related deficits for the domain-specific tasks. Other studies have similarly indicated that performance differences between young and older experts are substantially smaller than those between young and older novices (Rogers and Fisk, 1996; Salthouse, 1991). Poor identification of shapes and objects is associated with poor memory for them. Thus, in addition to declines in facial identification and recognition, similar age-related declines occur in object recognition (Park et al, 2002)—which would in turn be reflected in witness failures to identify accurately such objects as vehicles, clothes, guns, and other items relevant in forensic contexts. Older adults also experience greater difficulty with memory for the location of objects. Older persons cite difficulties in remembering object locations in everyday life among their memory concerns (Reese et al., 1999), and substantial evidence exists to support the beliefs that spatial location memory declines with age (see the review in Cherry and Jones, 1999). Caldwell and Masson (2001), for example, had subjects work with drawings of household objects and rooms of a house depicted on a computer monitor to simulate placing objects in various places. Older subjects demonstrated poorer memory for object locations. In a real-life study of memory for spatial layout, Uttl and Graf (1993) tested museum visitors of ages 15 to 74 on memory for the spatial layout of recently visited museum displays. Those over age 55 showed reduced accuracy. Finally, even when reporting on familiar well-known spatial locations, older adults perform more poorly (see the review in Lipman and Caplan, 1992). For example, Evans and colleagues (1984) found that older adults were less able to recall, or place accurately on a map, buildings from a familiar part of their city. These results and others suggest that older witnesses may suffer a disadvantage when asked to testify regarding spatial


236

layout or object location. Furthermore, evidence demonstrating that older adults tend to overestimate their ability to recall object location (West et al., 2002) suggests that they will not be aware of their inaccuracy, perhaps becoming very confident, but inaccurate, eyewitnesses. Spatial processing is also important for encoding location or relative location and for navigation from one point to another, as well as for memory for that route. In a creative use of virtual environment (VE) technology, Moffat and colleagues (2001) examined the navigation skills of adults from 22 to 91 years of age. The VE included a richly textured series of interconnected maze-like hallways, complete with dead ends and others leading to the goal. Older participants took longer to solve each trial, traveled longer distances, and made significantly more spatial memory errors. More of the younger (86%) than older (24%) adults achieved error-free performance after five trials. Furthermore, performance was related to standard measures of verbal and visual processing. Similar results for place navigation have also been obtained using a virtual Morris water maze (Moffat and Resnick, 2002; see also Shukitt-Hale et al., 2004). The navigation skills that depend upon spatial processing are reflected in everyday use of the environment. For example, Simon and colleagues (1992) examined the relationship between spatial processing and neighborhood use among older adults. Spatial memory was related to neighborhood knowledge, which in turn was predictive of neighborhood use (e.g., visits to shops, services, restaurants) and more predictive than mobility or years living in the neighborhood. Finally, such skills are also reflected in memory for routes taken, which also has been shown to decline with age (see the review in Lipman and Caplan, 1992). Problems of Attention and Encoding Successful encoding of information is dependent upon adequate attention (Craik and Lockhart, 1972). Memory is, in turn, dependent upon adequate encoding, such that inadequate attention and depth of processing at encoding may cause a person later to commit errors of omission (failure to remember things that did happen) and errors of commission (“remembering” things that did not happen, or remembering inaccurately what did happen; Schacter, 1999). Age-related decline in attentional control and resources renders older adults more susceptible to both errors, such that they remember less, and less accurately, than younger adults (see reviews in Craik, 2000; Koutstaall and Schacter, 2001). These age-related limitations also render older adults more susceptible to disruption of attentional capacities at encoding. For example, encoding is more impaired by distraction for older than younger adults. Therefore, although complex events are encoded more poorly than simpler events for those of all ages, complexity will cause greater encoding deficits in older adults. Thus, across a variety of tasks, undistracted older adults often perform similarly to distracted younger adults, and distracted older adults perform more poorly than distracted younger adults (see Zacks et al., 2000).


237

Bias in Interpretation Information may be perceived accurately by the senses and nevertheless be encoded incorrectly due to failures of interpretation. One source of such failures is “schematic processing.” The influence of schemas on information processing is pervasive (for reviews of the following, see Fisk and Taylor, 1991; Hastie, 1981; Kunda, 1999). They allow us to recognize and categorize objects, people, events, situations, and other features of the environment. They form the basis of expectations that tell us what to do with objects or persons, or how to behave in specific circumstances. They offer standards for evaluation for what we witness or experience, direct attention to schema relevant features of what we witness, and direct interpretation of it. Schemas generally direct attention and information processing so that attention is selectively focused on schema-relevant aspects of the situation. Thus, memory for schema relevant information is generally better than for schema-irrelevant information. Schema-inconsistent information is often processed more carefully, in order to try to integrate it into the schema-driven impression and, hence, this enhanced processing leads to better memory. Schemas exert a biasing effect on initial interpretation and later memory. Information is interpreted in light of salient schemas. Hence, interpretation is often distorted toward consistency with the schema, e.g., when activation of a racial stereotype can lead to interpretation of success as luck rather than ability. Furthermore, memory is distorted so that schema-relevant information is retained at the expense of schema-irrelevant information. In addition, schema-based “constructive” memory processes can cause a person to fill in additional schema-consistent information. Generally, memory is biased toward retention of true schema-relevant information and addition of false, but schemaconsistent, information. Older adults are more susceptible to the biasing effects of schematic processing with respect to social information (Hess, 1999) and memory for nonsocial objects (Hess and Slaughter, 1990). We discuss this in greater detail in the section on jury decision making. Failure of Retention Age is related to more rapid forgetting in a variety of specific domains, including spatial location (Rutledge et al., 1997). When tested for recall immediately, older adults show substantially less impairment relative to younger adults than when tested after delay (see reviews in Craik, 2000; Zacks et al., 2000). Thus, it is particularly important to interview older witnesses as soon as possible after the event in question. This differential slope of forgetting over time also has implications for the performance of older jurors, particularly in long trials in which evidence and witness testimony must be remembered for weeks or months before the jurors reach a verdict. Accuracy of Retrieval Generally, older adults are unable to remember events as completely or as accurately as younger adults (see reviews in Craik, 2000; Rubin, 2000). “Episodic memory,” or memory for autobiographical events that have happened comparatively recently (see reviews in Craik, 2000; Wingfield and Kahana, 2002), is among the cognitive abilities


238

shown to decline most sharply with age. Thus, older adults will, on average, be less able to retrieve elements of events they have witnessed, whether peripheral details or the core of the event. However, a particular deficit of episodic memory—memory for the context of an event (often referred to as “source monitoring”)—has been of interest in legal settings. Therefore, we will focus our discussion on this problem. Age and General Problems of Source Monitoring Older persons experience greater difficulty in “associative learning”—that is, in forming and maintaining connections between mental events. This is referred to as a deficit in “binding,” or integration of the various elements of an event, or of the event with the context in which it occurs (see reviews in Craik, 2000; Glisky, 2001; Koutstaal and Schacter, 2001). “Source memory” or “source monitoring” refers to the latter instance of associative learning (or binding), in which the memory of the core or gist of an event is bound successfully with the memory of the context in which it occurred. Substantial evidence has indicated that source or contextual memory is particularly dependent upon frontal lobe function and that the frontal lobes are preferentially affected by aging (see Glisky, 2001; Koutstaal and Schacter, 2001; Raz, 2000). Memory for context is considered to be more difficult than simple memory for the core of an event, as well as more susceptible to age-related decline, in that it requires greater attention (and divided attention) and processing of content and the spatiotemporal link between elements at encoding—capacities known to decline with age. Older adults may have problems dividing attention to encode the core event, along with the contextual cues and links between them, and therefore narrow attention to the core at the expense of context. Schacter and his colleagues (see Koustaal and Schacter, 2001) have suggested that elderly adults tend to encode information in a less elaborative and distinct manner (such that cues that would distinguish the target information from other potentially similar information are encoded and bound with the target), which tends to render them more susceptible to source-monitoring errors. Indeed, a number of studies have shown that fact memory is substantially less affected by aging than source memory (see the review in Glisky, 2001). Furthermore, although source errors can be reduced for younger and older adults through use of more elaborative encoding strategies that increase the distinctiveness of target information, age differences still remain (see Koustaal and Schacter, 2001). Difficulty with adequate encoding of contextual cues has been shown to result in a disadvantage for older adults in memory for, among others, • Perceptual details such as color, case, or font of written stimuli • Locations • Temporal sequence • The sex or specific identity of a speaker • Whether an item was presented in video or photo format • Whether something was presented auditorily or visually • Whether the individual personally simply said something or actually did it • Whether something was seen or heard vs. simply inferred on the basis of other things that were seen or heard


239

• Whether something was imagined or actually happened • Why a face seems familiar (i.e., where it is known from) • Which person did or said something • Things one is supposed to do at a particular time or place (termed “prospective memory”) • The source of medical information See reviews in Glisky (2001); Zacks et al. (2000); and Koutstaal and Schacter (2001). Furthermore, older adults tend to rely more on schematic knowledge in attempts at source memory, which may enhance their source-monitoring errors when the actual source is inconsistent with their schemas (Mather et al., 1999). In part, these deficits in source monitoring are reflected in problems we have already reviewed, such as less detail and poor memory for time and spatial features such as durations, sequence, location, layout, or route. However, of particular interest in forensic settings, problems in source memory have been implicated as the cause of inaccuracy in witness testimony. Three general arenas of source confusion in witness testimony have been extensively studied: • The “misinformation effect” (Loftus, 1979), in which information suggested during an interview replaces or becomes integrated with the memory of the actual event • Confusion between ideas or “memories” developed in therapeutic procedures and actual events (Loftus and Ketcham, 1994) • Misidentification of an innocent suspect due to misunderstanding of the source of the sense of familiarity of the face We review research relevant to age differences in susceptibility to these three forms of source monitoring errors in the following subsections. Later, we consider the implications of age-related decline in source monitoring for jurors. The Misinformation Effect A popular paradigm for studying memory distortion that is applicable to the forensic arena has been the “misinformation” paradigm developed by Loftus and her colleagues (Loftus, 1979; 1992; Loftus and Hoffman, 1989; Loftus et al., 1978). Participants are exposed to pictures or films depicting objects or action sequences and, subsequently, experience misleading information regarding the contents. For example, the experimenter may ask a question that presumes the existence of a fact or object not depicted in the original materials (e.g., “Did another car pass the red Datsun while it was at the stop sign?”). When later asked whether the nonpresented object (the stop sign) was present in the original depiction, many participants will nonetheless report that it was and proceed to describe it in detail. Older subjects have reliably shown greater susceptibility to this misinformation effect (Cohen and Faulkner, 1989; Karpel et al., 2001; Loftus et al., 1992; Searcy et al., 2000; see Coxon and Valentine, 1997, for less clear differences) and greater confidence in their false memories (Cohen and Faulkner, 1989; Karpel et al., 2001; Mitchell et al., 2003). Underwood and Pezdek (1998) investigated relative susceptibility to misinformation provided by credible vs. incredible sources. Although misinformation from a credible source induced more errors immediately after the misinformation, 1 month later errors consistent with the misinformation were high for both high- and low-credibility sources.


240

Presumably, as memory for the source of information decreased, discounting of information from a low-credibility source also decreased. Greater susceptibility to source forgetting in the elderly would be expected to render them relatively more susceptible to misinformation from low-credibility sources—particularly as the delay between the misinformation and retrieval increases. Multhaup and colleagues (1999) investigated the potential of a “source-monitoring” retrieval procedure to reduce the misinformation effect in older adults. This procedure had previously been shown to eliminate the misinformation effect in younger adults (Lindsay and Johnson, 1989). Lindsay and Johnson gave half of the participants in a misinformation experiment a yes/no recognition test for items in the original event. The other half was given a source-monitoring test, in which they were asked to identify the origin of each item as the original picture, the (misleading) text, neither, or both. The authors found the standard misinformation effect with the yes/no procedure, but not for the source-monitoring procedure. Multhaup et al. (1999) replicated this design with older adults and, again, the source-monitoring test format eliminated the misinformation effect. At least in their sample, instructions requiring conscious effort to retrieve the contextual source of information were effective for the older adults. A conceptually similar pattern of results was obtained by Multhaup (1995), who showed that the “false fame” effect was eliminated by a similar source-monitoring retrieval task. Subjects in the false-fame task pronounce a list of nonfamous names. Later, they judge as famous or nonfamous a list of names including famous names, pronounced nonfamous names, and new nonfamous names. During the fame judgment task, the pronounced nonfamous names seem familiar. If the person fails to monitor the source of that feeling of familiarity accurately as the previous pronunciation, he may falsely conclude the name is famous. As a result, previously pronounced nonfamous names will more often be judged famous than those not previously pronounced. Multhaup’s (1995) source-monitoring retrieval task led participants to indicate whether each test name was a famous name, a nonfamous name pronounced earlier, or a new nonfamous name. Although older adults were more susceptible to the false fame effect than younger adults in a standard famous/nonfamous recognition task, the false-fame effect was eliminated for both groups in the sourcemonitoring retrieval task. For the misinformation and false-fame studies, the authors attributed their results to the adoption of stricter decision criteria for assigning a particular item to a particular source to perform the source-monitoring task. In the absence of such strict criteria, the person may rely on a sense of familiarity to decide whether the item was present in the picture, whereas the stricter criterion presumably requires reliance on memory for more specific perceptual and spatial details to establish context. Multhaup et al. (1999) suggested that older adults may benefit from listing possible sources when they are trying to retrieve the context of events or information. (See Zacks et al., 2000, for other factors and strategies that moderate age differences in source monitoring failures.) A promising line of research showing that elderly people can learn to reduce source monitoring errors involves the use of a procedure known as the DRM paradigm. Here, subjects study lists of semantically related words (such as dream, tired, bed, snore) and then try to recall or to recognize what they previously heard. A central finding is that all adults, regardless of age, are prone to remember nonpresented but associated words (e.g.,


241

sleep). However, older adults are more likely than younger adults to falsely recognize the critical nonpresented lures. This increase in false recognition has been seen in a number of paradigms (Dodson and Schacter, 2002). On a more positive note, older adults were able to learn to use a retrieval strategy called the “distinctiveness heuristic” to reduce their rate of error. If pictures accompanied the words, an especially large reduction in error rate occurred for the elderly, making their performance more comparable to that of young adults. Source Confusion and Therapeutic Process Recent years have witnessed an explosion of interest in the extent to which common therapeutic techniques such as hypnosis, guided imagery, dream interpretation, and participation in “survivor groups” might lead a patient to develop false memories—in particular, false memories of sexual abuse (see Loftus and Ketcham, 1994). A wealth of data has now accumulated to demonstrate that these techniques can, indeed, produce false memories (see reviews in Davis and Follette, 2001; Loftus and Ketcham, 1994). Essentially, the person is exposed to suggestions from the therapist that a particular event may have happened, followed by a variety of procedures designed to uncover the memories that are at first allegedly inaccessible. These procedures involve continuing suggestion during such activities as dream interpretation, guided imagery, or hypnosis. For example, the person may be asked to imagine the event in question, try to remember it, or expose himself to situations that might trigger the memory (such as survivor group discussions). In this context, the person may develop vivid images or dreams of the presumed events, even though they never occurred. The combination of suggestion and imagery results in what feels like a “memory.” However, the person mistakenly believes the source of this memory to be an actual event, rather than the mental processes and images generated by the therapeutic process. Empirical research has demonstrated the capacity of suggestion, guided imagery, hypnosis, and other procedures to generate false memories. However, although there is not a great deal of research on aging and the potential of such procedures to induce false memories, we did locate some work using imagination with older participants. A number of studies have shown that older adults more easily mistake events they have imagined as memories for actual events. This difference applies with respect to confusion between imagining and actually doing something, and imagining vs. seeing or hearing something done or presented by others (see reviews in Koustaal and Schacter, 2001; Zacks et al., 2000). Given these specific results in combination with the general literature showing age-related decline in accurate source monitoring, it is reasonable to hypothesize that older adults will be more susceptible to development of false memories through common therapeutic procedures. Face Perception and Eyewitness Identification Cognitive researchers have identified age-related declines in fundamental face perception processes, which ultimately result in greater rates of inaccurate eyewitness identification and, in particular, higher rates of false identification. We first consider age-related changes in basic face perception processes and then turn to research specifically on eyewitness identification.


242

Age-Related Declines in Face Perception and Face Memory. Aging is generally associated with decline in spatial/nonverbal ability (see reviews in Chalfonte and Johnson, 1996; Jenkins et al., 2000). Consistent with this general decline in visuospatial abilities, research on face recognition has documented age-related declines in recognition performance for unfamiliar (once before seen) faces (Bartlett, 1993; Bartlett and Leslie, 1986; Bartlett et al., 1989; Crook and Larrabee, 1992; Ferris et al., 1980; Smith and Winograd, 1978). Older adults exhibit greater susceptibility to false recognition of faces not actually previously seen (Koustaal et al., 2001), although consistent age differences in accurate recognition of once-seen faces has not been observed (see Searcy et al., 1999, for review). In addition, older adults report less confidence when they do correctly reject foils (Yarmey, 1984; Yarmey and Kent, 1980). Older adults seem generally more susceptible to false recognition because they exhibit greater tendency to select false information when recognizing details of a previously seen event as well. Several processes appear to underlie age-related differences in facial recognition performance. First, many face recognition tasks rely at least in part on long term episodic memory—for example, in the standard eyewitness identification paradigm to be discussed later. Because episodic memory shows one of the steepest declines with age, age-related declines in facial memory may reflect general deterioration of long-term episodic memory. Consistent with problems with long-term episodic memory, Memon and colleagues (2003) found the effect of age on perpetrator lineup identification to be greater after 1 week than after 35 minutes. Second, face recognition relies in part on accurate source-monitoring capabilities. The person must recognize that the face is known and retrieve the context in which it was encountered. Some have suggested that older adults rely more on heuristic strategies such as “familiarity,” “feeling of knowing,” or “availability,” in the absence of more consciously controlled processes such as contextual recollection and source monitoring (Bartlett and Fulton, 1991; Jacoby, 1991; Searcy et al., 1999, 2000). Thus, they will tend to choose faces that seem familiar, whether or not the source of familiarity is correct. Indeed, Memon et al. (2003) showed that correct perpetrator lineup identifications were related to an independent measure of source recollection ability, which in turn was negatively related to age. Third, advancing age may result in declining face perception abilities. Indeed, evidence from the standard Benton Facial Recognition Test (Benton et al., 1994) has indicated that older adults become less able to match a target face to the correct alternative among six candidates shown immediately below (and simultaneously with) the target. This non-memory-dependent task has been shown to predict time-delayed memory-dependent witness lineup performance (Searcy et al., 1999). Therefore, age appears to be related to decline in perceptual feature matching, which is in turn related to poorer recognition performance. Furthermore, Memon and Bartlett (2002) have offered evidence that older witnesses reported greater reliance on feature matching as a strategy for identification of culprits from a lineup and, in turn, those who reported relying on feature matching were less accurate (13%) than those who reported other strategies (44%). Interestingly, the reverse was true for younger witnesses (38 vs. 15%) Thus, it appears that older adults may be more reliant on feature matching as a strategy for recognition and less accurate in using their preferred strategy.


243

These and other potential mechanisms underlying decline in face perception and memory are the likely results of age-related changes in brain structure and function. For example, Schretlen and his colleagues (2001) examined the influence of age, sex, education, perceptual comparison speed, frontal lobe volume, nonfrontal volume, and ventricle-to-brain ratio (VBR) to Benton Facial Recognition Test performance. Age was the strongest negative predictor of performance. However, VBR and processing speed alone accounted for almost 34% of the variance, with age in the equation. Frontal and nonfrontal lobe volume also contributed significantly to the equation, but added only slightly more than 1% to the explained variance. Such results are consistent with longstanding findings that lesions, injuries, or lobectomies to the frontal lobes create impairment in face processing and recognition (see the summary in Schretlen et al, 2001). Others have directly examined brain function during facial processing tasks through use of event-related brain potentials (ERP) (Grady, 2002; Pfutze et al., 2002) or PET (positron emission tomography; Grady et al., 2002) in attempts to identify the location and nature of age-related differences in brain function and their relationship to deficits in specific stages of face processing and retrieval under varying task conditions. A detailed review of these results is beyond the scope of this chapter. However, Grady et al. (2002) and Pfutze et al. (2002) provide excellent summaries of recent relevant literature. Generally, such studies of brain function have shown some age-invariant functions but a greater variety of age-related changes in brain functions, which can be empirically linked to face processing/recognition performance. Furthermore, they have shown that older adults appear to use different areas of the brain than younger adults for face processing under some conditions, indicating that loss of function in some areas of the brain may be compensated for by alternative processing routes. Together with studies linking structural changes in the brain to performance, the ERP and imaging studies clearly demonstrate that brain changes underlie age-related declines in face processing. Finally, several studies have shown that older adults less accurately identify facial displays of emotion in photographs and video clips (Sullivan and Ruffman, 2004; McDowell et al., 1994). As with other visual skills, older adults appear to use different areas of the brain for emotion identification than younger adults (Gunning-Dixon et al., 2003). Although not directly relevant to eyewitness identification, these results suggest that older adults may be less accurate in reporting the emotional tone of interchanges and thus less accurate in their reports of the meaning of the exchange. Specific Studies of Age and Eyewitness Identification. Compared to the enormous number of studies that have been done on children as eyewitnesses, there is a relative paucity of research involving older eyewitnesses. A few studies have shown that elderly eyewitnesses are more prone to inaccuracy in eyewitness identifications (List, 1986; Memon and Bartlett, 2002; Memon and Gabbert, 2003a, b; Memon et al, 2002; O’Rourke et al., 1989; Yarmey, 1985, 1993, 1996; Yarmey et al., 1984), with elderly witnesses making from 25 to 50% more identifications than their younger counterparts—most of them false. A few studies have failed to replicate this age difference (Memon et al., 2003; Wright and Stroud, 2002; Yarmey and Kent, 1980). However, Yarmey and Kent (1980) did find poorer recognition of bystanders among older participants. Also, Memon et al. (2003) found no age differences on the Benton Facial Recognition Test in their sample. Thus, the lack of differences in false identifications in their study may be explained by their relatively high-functioning older sample.


244

As with basic face recognition studies, eyewitness identification studies have shown that older adults are particularly prone to false identifications of innocent foils, whereas there are often smaller or no age differences in accurate identification of the perpetrator in target-present lineups (Adams-Price, 1992a, b; Searcy et al., 1999,2000; Memon and Bartlett, 2002; Memon et al., 2003; Memon and Gabbert, 2003a, b), although Searcy et al. (2001) found the opposite pattern and, in another study, found an age-related reduction in hit rates (Searcy et al., 1999, lineup 3). Generally, the tendency of older adults to make false identifications in target-present, as well as in target-absent, lineups is robust and widely interpreted as the result of older adults’ greater use of “gist-like” encoding and failures of source memory (Koutstaall and Schacter, 1997; Schacter et al., 1998; Tun et al., 1998). The Own-Age Effect. An unfortunate characteristic of much of the research on age and eyewitness memory is that, although witness age is varied systematically, perpetrator age is often held constant within younger age groups, thereby confounding witness age with witness-perpetrator relative age. This presents a problem of potential significance. That is, just as eyewitness researchers have documented a “cross-race” effect in eyewitness identification (Sporer, 2001), there appears to be a “cross-age” effect, such that witnesses are better able to identify targets of their own age group. In two experiments, Wright and Stroud (2002) found that participants from two age groups (18 to 25 vs. 35 to 55), who had previously viewed videos of a car or a television theft, were more accurate at identifying the culprit when viewing culprit-present lineups of people their own age. (This effect occurred in those making the identification after only 1 day, but was very weak in those making the identification after 1 week and was weaker in older than in young adults.) No same-age effect was found for misidentification of innocent suspects in target-absent lineups. Similarly, an earlier study by List (1986) had 10-year-olds, college-age students, and adults 65 to 70 view shoplifting videos. Culprits were college-age or middle-age females. Older adults exhibited poorer recognition memory for the college-age culprits than either of the two other groups but better memory for the middle-age culprit. Memon et al. (2003) investigated accuracy in identification of perpetrators who were old or young among old and young participants. Younger participants generally outperformed older participants. Older participants made more false identifications in perpetrator-present and perpetrator-absent lineups, particularly for younger perpetrators and after delays of 1 week rather than 35 minutes. Younger participants were also more prone to misidentify someone in older lineups, but were not less likely to identify the perpetrator. Finally, own-age recognition biases have been observed outside the eyewitness identification paradigm. Several studies using standard face recognition paradigms, for example, have found superior performance for own-age targets (Bartlett and Leslie, 1986; Fulton and Bartlett, 1991; Mason, 1986). Furthermore, in their initial demonstration of the “change-blindness” phenomenon, Simons and Levin (1998) obtained what was apparently an own-age effect. College-aged experimenters approached an unsuspecting pedestrian on campus, asking for directions to a campus building. While the pedestrian attempted to provide directions, two other confederates appeared to interrupt rudely by passing between the experimenter and pedestrian, carrying a large door. As the door passed between them, the first


245

experimenter changed places with a second, who emerged from behind the door and continued the exchange with the pedestrian as if he had been in the conversation all along. The second experimenter wore different clothing and differed in height and voice. Yet, less than half of participants noticed the change and older adults were less likely to notice than younger adults. Simons and Levin interpreted this age difference as an instance of the broader owngroup bias in face processing. A replication of the door experiment revealed that even younger people are more likely to notice the change for members of their own social group (other college students) than for those of a different social group (the same persons, but dressed as construction workers; Simons and Levin, 1998, experiment 2), perhaps due to greater specificity in processing for in-group members. The growing evidence of an own-age bias in face recognition suggests that the results of many earlier studies of eyewitness accuracy might be due not to age-related general decline in accuracy but rather to the particular age combinations used in the study. Further research is needed to examine the relationship between witness and culprit age that employs adults throughout the age span, including into old age. Meanwhile, the results are nonetheless useful in practice for a wide variety of witnessing situations. Because crime is much more prevalent among the young, elderly witnesses will often be in the position of attempting identification of a younger culprit. Interaction of Age with Conditions of Encoding or Retrieval. To the extent that age renders the witness, more susceptible to failures of encoding, storage, or retrieval, it is relevant to ask whether the aging witness will suffer enhanced impairment as a result of poor witnessing conditions, longer retention intervals, or biasing influences at retrieval. To date, however, only a few studies have addressed these issues. Witnessing Conditions. Two studies examined the effects of “weapons focus,” violence of the crime, perpetrator disguise (O’Rourke et al., 1989), and “cross-race” targets (Brigham and Williamson, 1979) as a function of age, but found no interaction of these variables with age. Memon et al. (2003) examined the effect of duration of exposure to the perpetrator’s face, and whether the lineup was target present or target absent, on accuracy in young and old adults. They also found no interactions with age (although as noted earlier, the sample of older adults scored equivalently to the young adults on the Benton Facial Recognition Test). We were unable to locate other studies examining the interaction of age with effects of such variables as lighting, arousal, and so on, even though the documented declines in perceptual processes would suggest such age-related interactions. Thus, a number of issues remain open for future research to pursue. Attention at Encoding. The Negative Relationship between Memory for Detail and Eyewitness Accuracy. Some previous research has indicated that memory for details of some aspects of an event is negatively associated with memory for others (Wells and Leippe, 1981). If attention is focused on some things, attention to others will suffer. To the extent that older adults suffer decrements in capacity for divided attention (as reviewed earlier), they might be expected to show enhanced negative relationships between recall of peripheral detail and identification accuracy. Indeed, Searcy et al. (2000) found that higher levels of recall of the target event and of other nonfacial features of the target were associated with higher levels of false identifications for seniors, but not for younger adults.


246

Lineup Procedures. Several studies have attempted to assess the differential impact of variations in identification procedures for younger and older witnesses. Eyewitness researchers have adopted the position that witness identification accuracy is enhanced through use of sequential lineups in which one potential culprit is present at a time, as opposed to simultaneous lineups in which (usually) six potential suspects are presented simultaneously (Steblay et al., 2001). More specifically, as shown by Steblay’s (Steblay et al., 2001) meta-analysis of these effects, sequential lineups appear to reduce the overall likelihood of any choice, correct or incorrect. Thus, although the procedure reduces the proportion of false identifications of innocent suspects, it also reduces true identifications of the perpetrator (hits). Self-reports of those tested with a sequential lineup suggest they tend to use an “absolute”—as opposed to a “relative”—judgment strategy, whereby the target is judged for absolute fit to the witness’s memory of the culprit rather than for the relative fit of the target to the memory (i.e., the fit relative to other members of the lineup) (Dysart and Lindsay, 2001; Wells et al., 1998). This stricter decision criterion would naturally have the effect of reducing witness willingness to make any choice, whether correct or incorrect. Memon and her colleagues have investigated the potential interaction of age with lineup procedure in a series of studies. Using a target-present lineup only, Memon and Bartlett (2002) investigated the effects of simultaneous vs. sequential procedures for young (18 to 30 years) vs. old (60 to 80 years) adults. Generally, older adults made more false identifications, but lineup procedure did not interact with age. Instead, sequential lineups decreased accurate identifications (hits) for both age groups. Similar results were obtained by Memon and Gabbert (2003b) for target-present lineups. However, these researchers (2003a) did find an interaction between age and lineup procedure in target-present lineups (but not target-absent lineups), such that sequential testing reduced false alarm rates for younger but not for older adults. For target-absent lineups, the authors found main effects for both age and testing conditions, such that both groups made fewer false identifications under sequential testing conditions, but older adults made more false identifications than their younger counterparts in sequential and simultaneous target-absent lineups Memon and Gabbert (2003b) further investigated the effects of lineup procedure specifically for TP (target-present) lineups. However, in this study, they also varied whether the appearance (hairstyle) of the perpetrator had changed subsequent to the crime. This variation was expected to have a greater effect under conditions in which the witness adopts stricter criteria of absolute fit for identification (i.e., the sequential procedure) than under those in which he might adopt a more lenient criterion of best fit (i.e., the simultaneous procedure). Younger (17 to 33 years) and older (58 to 80 years) adults viewed a short film depicting a female target stealing money from a car. Later, participants described the perpetrator and subsequently viewed a lineup, in either simultaneous or sequential format, in which the perpetrator’s hairstyle had or had not changed. Across conditions, younger adults recalled more details of the target’s appearance correctly and fewer incorrectly. Older adults made fewer hits and more false identifications than younger adults. Across age groups, fewer choices of all kinds were made under sequential (58%)


247

than under simultaneous (75%) lineup procedures. This drop was due to a drop in hit rates in the sequential lineups because there was no drop in false identifications. Interestingly, although the rate of misses increased dramatically for perpetrators whose appearance had changed, it did so particularly for sequential lineups, and particularly for younger adults. For older adults, although the pattern was similar, no significant main effects or interactions were found. Memon and Gabbert suggested that change in appearance was particularly detrimental to younger adults, who they presumed would be relying on attempts to remember the face in context specifically, whereas it would be less detrimental to older adults, who they presumed would be more likely to rely on a contextfree sense of familiarity. Memon and Gabbert’s (2003b) results suggest that, in practical circumstances, older adults may suffer a lesser drop in hit rates when viewing sequential, as opposed to simultaneous, lineups, specifically under conditions in which the perpetrator’s appearance may have changed. Changes in clothing, hair, facial hair, the presence of a hat, removal of elements of a disguise, and so on are arguably quite common between the crime and arrest. To summarize, for target-present lineups, age effects are somewhat inconsistent. For simultaneous lineups, some have found reduced hit rates among older adults (Searcy et al., 1999, lineup 3; 2001), and some have found no age differences (Searcy et al., 1999, lineup 1; Memon et al., 2003). Older adults have been consistently more susceptible to false alarms in target-present and target-absent lineups, whether administered simultaneously or sequentially. Sequential vs. simultaneous testing did not eliminate agerelated declines in eyewitness performance. Instead, under conditions in which the perpetrator’s appearance (hairstyle) was different in the lineup than when he or she had committed the crime, the hit rate of younger adults was selectively impaired by the sequential procedure, whereas the rate of false identifications of older adults was not improved. Overall, then, there appear to be no reliable age by lineup procedure interactions. Source Monitoring and Exposure to Mugshots. In some cases, the witness is asked to examine mug books to attempt identification of a suspect. This exposure can lead to a sense of familiarity, which forms the basis of a source-monitoring error that can produce a false identification when the witness later sees the same person in a lineup. If the witness actually identifies the innocent person in the mugbook, he may become committed to that identification in a way that further encourages him later to identify the innocent falsely in a lineup (see the review in Memon et al., 2002). Memon and colleagues (2002) exposed participants to a videotaped car theft. Subsequently, they were asked to examine a target-absent mug book to attempt identification of the suspect (although control subjects did not inspect the book). All participants were given a target-absent lineup 48 hours later. Older adults were significantly more likely to identify someone from the mug book falsely and to choose a foil from the lineup. However, although all subjects who viewed a mug book were more likely to pick the foil who appeared in both the mug book and lineup, there was no age difference in this tendency. The Verbal Overshadowing Effect. Memon and Bartlett (2002) also attempted to assess age-related susceptibility to the “verbal overshadowing” effect (Schooler and Engstler-Schooler, 1990). Verbal over-shadowing refers to a situation in which verbal


248

descriptions of a target interfere with later target recognition. However, the authors reported only a small effect of verbal descriptions prior to attempted identification of a culprit, and no evidence of greater susceptibility in older jurors was found. Interviewing Procedure: Effects of Context Reinstatement. Because false identifications are considered to result in part from inaccurate source monitoring, a number of investigators have examined the effects of context reinstatement on identification accuracy (see the review in Searcy et al., 2001). The results of these investigations have been mixed. Nevertheless, Searcy and colleagues (2001) tested the effects of context reinstatement on identification accuracy of young and old adults, 1 month after a personal interaction with the target. No beneficial effects were obtained in either population. Similarly, Memon and colleagues (2002) found no benefit from verbal reinstatement of the context of a previously viewed crime on identification accuracy. Although other procedures for context reinstatement may prove more effective, there is currently no overall support for the effectiveness of context reinstatement. The “cognitive interview” (Fisher and Geiselman, 1992) includes such techniques as context reinstatement, minimizing background noise, asking open-ended questions, and encouraging use of nonverbal responding. This interview method has been shown to enhance the amount of correct information recalled. Although they did not examine accuracy of eyewitness identification, Mello and Fisher (1996) found that, whereas the cognitive interview enhanced accuracy for young and old adult witnesses to a videotaped crime, enhanced recall was greater for older adults. However, McMahon (2000) found no overall or age-specific effects of the cognitive interview (vs. a structured interview) but did find a significant age effect on amount of true (but not false) information recalled. Overall, research investigating remedies for age-related declines in eyewitness accuracy has been disappointing, yielding conflicting findings that make firm conclusions premature. Given the importance of accuracy in eyewitness testimony, this search for improvements should not be abandoned. Eyewitness Metamemory: Are Aging Eyewitnesses Aware of Their Inaccuracy? Eyewitness researchers have shown that the relationship between confidence in accuracy and actual accuracy is often weak (Bornstein and Zickafoose, 1999; Brewer et al., 2002; Olsson, 2000). Some researchers have examined the relative strength and direction of this relationship in older and younger witnesses. Yarmey (1985) reported that, although elderly witnesses were less confident than younger witnesses, there were no age differences in the relationship between confidence and accuracy. In contrast, although replicating the main effect of age on accuracy, four studies found positive relationships between confidence and accuracy in young adults, and no significant relationships among older adults (Memon and Bartlett, 2002; Memon et al., 2002, 2003; Searcy et al., 2000). Searcy and colleagues (2001) further found that seniors’ self-rated memory self-efficacy was positively associated with false identifications in target-absent lineups. It appears, then, that seniors may less accurately monitor the quality of their identification performances. A New Area for Expert Testimony? Stereotypes of older eyewitness are mixed. That is, although older witnesses are seen as less accurate (Groth, 1979; Kwong See et al., 2001), they are also seen as more honest (Brimacombe et al., 1997; Kwong See et al., 2001). Thus, juror intuition is consistent with actual overall age differences—at least with


249

respect to accuracy. Even so, aging witnesses are not always seen as less credible. Nunez and colleagues (1999) presented a case involving an aggravated assault, in which the victim identified the defendant. The victim was described as a generic adult victim or as one of five variations of older adults varying in apparent abilities. Ratings of believability were not significantly lower for older witnesses. Thus, in some cases, jurors may not be sensitive to age-related inaccuracies. It is also important to remember that age does not exert uniform effects on all measures of accuracy (e.g., hits in target-present lineups are comparable, whereas false identifications are greater), and age-related decline is widely variable across individuals. Thus, jurors would benefit from some mechanism for distinguishing between impaired and relatively unimpaired older witnesses. Reflecting this need, Geiselman and his colleagues (2001) examined the effects of disclosing the results of a witness’s Benton Facial Recognition Test scores on juror verdicts. Witnesses described as scoring lower on the BERT were seen as less credible. When the testimony included an explanation of the conditions under which BERT scores are most predictive of accuracy, jurors correctly placed more weight on the BERT scores when witnessing conditions for the reported crime were poor. These results indicate that testimony regarding individual witness abilities (including age-related abilities) can significantly affect witness credibility. Voice Perception and Earwitness Identification Forensic research addressing “earwitness identification” is far less extensive than that on eyewitness identification but has received increasing attention in recent years (see the review by Yarmey, 1995). Like eyewitness identification, evidence suggests that earwitness identification accuracy will decline with age. In line with the general agerelated decline in auditory acuity, pitch discrimination, and so on, older adults suffer declining ability to recognize specific speakers (Bull and Clifford, 1984; Maylor, 1997). In part, this difference is due to encoding difficulties. For example, older adults are less accurate at initial discrimination between male and female voices (Kausler and Puckett, 1981). Furthermore, older adults—particularly those with relatively poor frontal lobe functioning—show poorer source memory for voices (Glisky et al., 1995). After hearing several sentences spoken in one of two voices, those with poor frontal function showed impaired memory for the voice, but not for the sentences. Although there is a burgeoning literature on earwitness identification (Yarmey, 1995), we were unable to locate any studies of the relationship between speaker identification and age conducted within a witness identification paradigm. Therefore, much remains to be investigated regarding the elderly earwitness. 11.4 Age-Related Judgment and Decision-Making Processes in Jurors Stereotypes of the cognitive status of older adults appear to be both realistic and overly optimistic. Several studies of beliefs regarding age and cognitive function among German adults (Heckhausen et al., 1989; Heckhausen and Krueger, 1993; Hummert et al., 1994) found that, although age is viewed as associated with memory problems (e.g., slow, forgetful, absentminded) and inflexibility (e.g., moralistic, overcautious, obstinate, headstrong, stubborn, inflexible), it is also viewed as associated with knowledge (e.g.,


250

knowledgeable, smart, experienced, well read, educated, well informed, etc.) and wisdom (e.g., open minded, reasonable, levelheaded, wise, intelligent, etc.). Thus, stereotypes of aged thinkers do include forgetfulness, but also appear to include confidence in more rational and wise decision making. How, if at all, do such stereotypes reflect actual differences in juror decision-making processes? Thus far, we have summarized a number of differences in cognitive/informationprocessing abilities between younger and older persons. In this section, we first consider some implications of age-related declines in perception and memory for jury information processing. Second, we briefly review what is known about age-related differences in judgment and decision processes that would affect how jurors use the information they acquire to come to verdict/sentencing/damage decisions. As we consider these issues, we offer recommendations for effective presentation of information to older jurors if they are seated on the trial jury. Age-Related Differences in Information Processing Perhaps most relevant for aging jurors are age-related declines in perceptual abilities (see Faubert, 2002; Scialfa, 2002) that will compromise ability to process speech, nonverbal cues to meaning and deceit, and visual and auditory exhibits successfully. The most central of these, of course, are speech and discourse processing/understanding, which unquestionably decline with age. However, much of this decline is due to simple perceptual difficulty, and can be reduced by slower and louder speech (see the review in Schneider et al., 2002). Thus, attorneys facing older jurors should take care not to speak too quickly, to face jurors and have witnesses face jurors when possible (to allow the jurors to clarify unclear speech through lip reading and nonverbal cues), and to maintain adequate volume at all times. Unfortunately, even when aging jurors do understand the content and the point of the presented evidence, they will (on the average) retain less of that information, particularly across the long retention intervals required in lengthy trials. For example, Fitzgerald (2000) examined memory for evidence among younger vs. older jurors. In a measure of free recall, jurors were asked to write an account of a 2-hour toxic tort trial they had just witnessed, written “so that someone who had not watched the trial would know what had taken place.” Overall, the accounts of older jurors were judged as significantly less “cohesive” than those of younger jurors. Older jurors’ accounts also included fewer statements of probative facts, but more evaluative statements regarding parties to the case or lines of argument (although these differences were attenuated when legal instructions were provided before, rather than after, the evidence). Older jurors did not, however, include more erroneous statements or reports of evidence that was not actually presented. Recognition measures of memory for evidentiary statements and testimony were similarly influenced by age, in that older jurors recognized fewer actual statements, but did not show greater false recognition for nonpresented statements. Clearly, older jurors will be at greater risk of failure to understand and remember evidence. Thus, these jurors may benefit more than younger ones from standard presentation tactics designed to enhance memory, such as redundancy, vividness in language and visual exhibits, exhibits designed to reinforce points, enumeration of points


251

of evidence and drawing connections between them and trial issues, and so on. Fitzgerald (2000) also varied whether jurors were permitted to take notes, showing that older and younger jurors benefited. Thus, note-taking may be another valuable support for older jurors. Implications of Age-Related Failures in Source Monitoring We have already documented the age-related relative inability to monitor the contextual source of information or events accurately. In jurors, failures of source monitoring would be expected to result in failure to track the source of testimony and evidence accurately, as well as to track the association between information and other cues relevant to the accuracy of that information. Indeed, older persons are less able to remember accurately which person gave them specific items of information (Ferguson et al., 1992; Hashtroudi et al., 1989; Schacter et al., 1991) and less able to remember the source of conversational contributions accurately (Brown et al., 1995; see Davis et al. [Chapter 12 in this volume], Davis and Friedman, in press for reviews). This failure to track the source of information accurately may contribute to failure to discount low-credibility evidence or testimony properly. For example, research on the “sleeper effect” (Priester et al., 1999; Underwood and Pezdek, 1998) has shown that although information associated with a source lacking in credibility may be initially discounted, as time passes the information becomes more credible. As the low-credibility source is forgotten or becomes dissociated from the information, the information ceases to be discounted, becomes more credible, and thus exerts greater influence on judgment. As noted earlier, using the misinformation paradigm, Underwood and Pezdek (1998) found that subjects became relatively more influenced by misinformation from a lowcredibility source over time. To the extent that aging jurors will be more susceptible to confusion regarding the source of evidence, the testimony of witnesses who have been discredited may likewise become more influential with older jurors because they fail to connect the discrediting information with the discredited witness’s testimony. Although the effects of the passage of time have not been directly tested, Chen and Blanchard-Fields (2000) studied age differences in reactions to true and false information, using the “false information paradigm” developed by Gilbert and his colleagues (1990, 1993). Participants read police reports containing true statements printed in black and false statements printed in red. They were told that the red statements had gotten into the report by accident and actually belonged to another crime report or an unrelated incident. Presumably, source-monitoring errors would be more prevalent among older adults—and among younger adults subject to distraction during their review of the report—and later cause them to include the discredited information in their judgments. As expected, undistracted older adults and distracted younger adults were more influenced by the false statements. Thus, even when judgment immediately followed the presentation of information, older adults were more likely to lose the link between a fact and the credibility of the fact than younger adults. A related failure may occur with respect to attorney claims or arguments. People tend to rate repeated statements or claims as more valid and believable than those presented less often—an effect known in the marketing and advertising literature as the “truth effect.” Older persons are more susceptible to this effect (Law et al., 1998), suggesting


252

that older jurors may be more persuaded by oft-repeated claims, whether backed by good evidence or not. This tendency is assumed to result from the increasing sense of familiarity resulting from repetition, combined with failure of source memory. Although we know of no direct tests of our suggestion that source forgetting may selectively enhance the credibility of discredited witnesses or attorney arguments in aging jurors, one study directly tested source memory for plaintiff, defense, and neutral facts (i.e., correct linking of statements to witnesses and other participants in the trial). Older jurors made fewer correct links than younger jurors (Fitzgerald, 2000). (See also Davis et al. [Chapter 12 in this volume] and Davis and Friedman, in press for reviews of age differences in source monitoring in conversation.) A final kind of source monitoring is potentially more important that those we have cited thus far; that is, maintaining the mental link between an item of evidence and the implications of that evidence for the central issues of the case. Such links are often a problem for jurors because attorneys do not always take care to tell jurors why a particular fact is important and to link it to the case issues—often because they believe it is so obvious as not to need connection. In both senses, however, older jurors may be at a disadvantage. Cohen (1979) suggested that even when elderly people are able to understand what was said, they are less able to draw inferences from the literal content, so others often view them as failing to “get the point.” This point has been supported in studies of comprehension of medical information. Older patients answer questions concerning the literal content of drug information somewhat more inaccurately than younger adults (e.g., “What are the side effects of this drug?”), but perform much more poorly when answering inferential questions requiring searching for information in several places and putting it together sensibly (e.g., “How long will this medication last?”—which requires use of information concerning number of times daily the medication is to be taken, along with the number of pills in the bottle). The age difference became even stronger for more complex inferential questions, although both groups performed more poorly on more complex questions (Finucane et al., 2002; Park et al., 1994). This age-related deficit in inferential processes would suggest that attorneys must go to greater lengths to connect the information they provide to the inferences and conclusions they wish the jurors to draw. Furthermore, given that older jurors may also be more susceptible to forgetting such connections, attorneys should reinforce these connections more frequently for them. Age-Related Differences in Judgment/Decision Processes Once the evidence is in, jurors must consider the nature and weight of the evidence for each side, compare that evidence to legal standards for verdict and damage decisions, and arrive at their final decision. Again, some evidence suggests that older adults face greater difficulties in reasoning, judgment, and decision making. For example, Salthouse (1993; Verhaeghen and Salthouse, 1997) found that younger adults outperformed older adults on a variety of tests of reasoning ability. In fact, age-related declines in matrix reasoning and analytical reasoning (which test such abilities as perception of relationships between information, integration of information to reach a conclusion, abstraction, verification, and so on) have shown some of the largest correlations with age of any behavioral


253

variable (Salthouse, 2001). Such deficits present a clear challenge for jurors faced with the lengthy presentations, substantial processing demands, and complex decisions necessitated by their role. Gathering and Combining Information In part, older adults seem to shortchange the information-gathering stage of decision making, relying on less complete evidence gathering and comparisons between fewer alternatives prior to making a decision. Some have suggested (Cole and Balasubramanian, 1993) that restrictions on working memory capacity may lead aging decision-makers to reach a point of information overload with less information than their younger counterpoints. Beyond this point, additional search can yield dysfunctional consequences and incorrect decisions (Malhotra, 1982). Zwahr and colleagues (1999) studied the implications of limitations in working memory for information use in medical decisions. Younger and older women studied detailed information regarding the pros and cons of hormone replacement therapy (HRT) for menopause and recommended a decision for a hypothetical woman in a vignette. Although age was unrelated to the recommended decision to take HRT, the manner in which the participants arrived at their decisions did vary with age. Older women sought out less information, made fewer comparative judgments, and considered fewer alternatives to HRT prior to arriving at their decisions (see also Meyer et al., 1995, for age and decisions regarding breast cancer). Path analyses indicated that cognitive abilities (including text memory, vocabulary, working memory, and perceptual speed) directly affected decision-making processes, whereas age exerted only indirect effects due to its strong negative association with cognitive abilities. Studies of decision making in arenas such as product choice (Beatty and Smith, 1987; Cole and Balasubramanian, 1993; Furse et al., 1984; Johnson, 1990, 1993; Schaninger and Sciglimpaglia, 1981) have shown a similar age-related restriction in information search and review of alternatives prior to a decision, and/or poorer final choices. Although not directly analogous to information processing in court, studies of computer search and retrieval tasks, such as searches for information on the Internet, library, or customer service databases, have shown that older adults have less efficient search strategies and greater difficulty remembering previously followed links or information on previously viewed pages (see the review in Czaja et al., 2001). In a particularly elaborate design, Streufert and colleagues (1990) studied decision making among four-person young, middle-aged, and older teams. Participants studied background materials about a fictitious nation and simulated the performance of a “national security council” required to manage the nation over a simulated compressed time scale of events across several months. The simulation posed management, governance, diplomatic, and emergency problems. As with the individual decisionmaking research reviewed previously, older teams considered less information, came to fewer decisions, and were less responsive to incoming information. Furthermore, whereas younger and middle-aged teams had more focused discussions and reached decisions more quickly, older teams talked more but were more diffuse and less task oriented and “engaged in many seemingly endless discussions or arguments, often about minor details” (Streufert et al., 1990, p. 556). However, they did respond to


254

unexpected positive and negative (emergency) information as well as the younger teams. Older teams also expressed satisfaction with their performance, apparently unaware of their failure to handle a wide range of events unsuccessfully. The literature on aging and decision making appears to converge in indicating that older jurors are likely to give less complete consideration to the evidence before arriving at an individual verdict. To the extent that the jury contains mostly older people, the Streufert et al. (1990) study suggests that the entire group may collectively review the evidence less completely. At this point, however, it is premature to conclude that older jurors will ultimately vote differently than younger jurors because they will, perforce, be exposed to the full jury’s deliberations. Schematic Processing Evidence also suggests that older jurors may use the information they acquire differently because they are more susceptible to biases due to personal attitudes and values and, more generally, to schematic processing (see the review in Hess, 1999). Hess (1999) suggested that aging will affect judgment through three components. The first, a processing component, refers to the cognitive resources brought to bear, including such components as speed, working memory, and inhibition that are known to decline with age. This decline should result in fewer resource-consuming modes of processing, including greater use of more automatic and effortless “top-down,” or schematic, processing, and less use of more effortful “bottom-up,” or systematic, piecemeal processing. Hess (1999) explicated the implications of greater use of schematic processing for construction of mental models. Generally, mental models refer to representations of specific situations constructed by individuals to reflect their current level of understanding (Johnson-Laird, 1983; 1989). They include an integrated view of the available information regarding the situation at hand, in the context of the person’s general world knowledge and world view, which results in a coherent conceptual representation independent of the surface structure of incoming information. In the trial context, construction of mental models may include those of specific individuals, products, situations, etc., or of the juror’s overall trial story. It is well known within the social cognition literature that more automatic and schematic processing occurs under conditions of limited attentional resources or heavy processing load (Bargh, 1994). Thus, in line with reduced processing capacity views of aging, studies of the relationship of age to construction of mental models have shown that age-related resource limitations posed by working memory and reduced efficiency due to failing inhibitory capacity combine to impair construction, updating, and access to mental models (see the review in Hess, 1999). Essentially, schematic, top-down-driven processing results in more rapid categorization, less systematic review of new information, and therefore less updating and revision of impressions as new input comes in. Along with increased schematicity, age-related enhanced susceptibility to the effects of priming on judgment (Hess et al., 1998) is consistent with the notion that the initial impression a person forms becomes, in effect, the schema that controls further processing. Information relevant to that schema


255

(or initial trial story) will be noticed and remembered better and will later become more influential in the final impression. To the extent that older adults are more susceptible to schematic processing, for example, they should also show greater selective memory for schema-relevant information. Consistent with this expectation, Hess and his colleagues (see the review in Hess, 1999) showed that older adults attempting to remember a narrative showed poorer memory for items that were atypical for, or irrelevant to, the type of situation depicted than younger adults. Essentially, script-inconsistent or -irrelevant information that could not be linked to the operative script was simply not processed sufficiently to promote memory. Generally, research on reproduction of recently learned stories has shown that older adults remember less information and are more likely to impose schematic narrative themes, morals, or lessons on their recall of story elements, thus remembering more highly integrative and succinct representations of the story’s essential meanings than younger adults (Adams et al., 1990, 1997; Pratt et al., 1989; Mergler et al., 1984/1985— see the review in Isaacowitz et al., 2000). Hess and his colleagues also showed that older adults are less likely to process apparently schema-inconsistent information carefully in order to try to understand the inconsistency. In research with younger adults, schema-inconsistent information is typically remembered better than consistent information (Hastie, 1984). Presumably, the effort to understand how the inconsistent information can be explained in light of the overall mental model causes it to be processed more deeply, leading to superior memory for it. However, Hess (1999) showed that this effect occurred for younger but not older adults. In line with Craik’s (2000) argument that self-initiated, effortful processing becomes less likely with age, the author interpreted these results to mean that older adults simply did not choose to engage in the effort to try to integrate the apparently contradictory information and therefore did not remember it better (or sometimes at all). In turn, failure to process apparently contradictory information carefully renders it less influential—hence, the lesser tendency of older adults to update and revise mental models as indicated by incoming information. In addition to its effects on memory for information that is presented, as noted earlier, schematic processing can lead to pseudomemories for information that is not presented, but consistent with (or implied by) the information that was presented. For example, in the false recognition paradigm developed by Roediger and McDermott (1995), participants study a list of words (such as symphony, song, lyrics, piano, jazz, etc.) semantically connected to a particular theme (e.g., music). Although the theme word (music, in this example) is implied by the other words, it is never presented. Nevertheless, many participants falsely recognize “music” as a word they previously learned. Older adults are reliably more susceptible to this effect (see the review in Koutstaal and Schacter, 2001). Thus, older adults may be susceptible to developing pseudomemories for nonpresented evidence or testimony implied by or consistent with the trial stories they have developed based on the actual presentation. Schematic processing can also influence the nature of source monitoring errors. For example, Mather and colleagues (1999) found that those who had previously watched a videotaped discussion tended to misattribute a liberal statement to a Democrat (rather than to the Republican who actually made it) or a statement regarding working out to an


256

athlete (rather than to the writer who actual made it) and that older adults were more susceptible to such schema-driven misattributions. In other words, source memory was reconstructed so that the statements matched the person from the social category schematically associated with the behavior or attitude reflected in the statement. Similar results have been obtained by others (Bayen et al., 2000; Sherman and Bessenoff, 1999). Generally, schematic influences on memory tend to be greater when veridical memory is poorer, when processing demands are high in comparison to processing resources, or when processing capacity is restricted in some way (Bayen et al., 2000; Sherman and Bessenoff, 1999; Spaniol and Bayen, 2002)—all conditions more likely to occur for older adults. Such schematic influences can include, among other processing biases: • Selective attention • Biased interpretation of evidence • Selective memory for evidence • Schema-consistent source misattributions and other schema-driven memory distortions • False memories for nonpresented evidence Thus, memory is trimmed of schema-irrelevant material and shaped by schema-driven distortions and additions to form a generally schema-consistent account of what has occurred. Finally, Hess and his colleagues (Hess, 1999) found that when coming to a judgment based on information in memory (such as would be the case for jurors arriving at a verdict at the end of trial), older jurors were less systematic in their use of that information. It should be noted that a particularly unfortunate result of schematic processing and failing inhibitory control is increased prejudice among older adults. In an influential model of prejudice, Devine (1989) proposed that what differentiates prejudiced from nonprejudiced people is their voluntary inhibition of negative and stereotype-related thoughts. She argued that because our culture is suffused with negative stereotypes concerning race, social class, sexual orientation, gender, age, and other social categories, we cannot fail to think of them when confronted with a member of such a category. The unprejudiced person is presumed to reject and inhibit such thoughts voluntarily, whereas the prejudiced person willingly accepts them. Von Hippel and colleagues (2000) proposed that older people may become more prejudiced unwillingly, due to failing capacity to inhibit stereotype-based thinking and response. In support of their hypothesis, the authors found that older adults were more prejudiced, even though they reported stronger desire than younger adults to control these reactions. Furthermore, their judgments relied on stereotypes even when instructed not to, whereas those of younger adults did not. Finally, age differences in stereotyping and prejudice were mediated by age differences in inhibitory capacity. Thus, older jurors are likely to suffer stronger effects of stereotype-based processing and to reflect greater impact of prejudice in their verdicts. Hess (1999) also pointed to a second “knowledge-based” age-related influence on processing. That is, he suggested that the nature of the person’s experiential and knowledge base changes with age and that older persons tend to rely more on their own experience to arrive at judgments. For example, Erber and colleagues (2001) found that, although younger adults were equally forgiving of a younger and older target who forgot


257

she was wearing a hat and left the store without paying, older adults were more forgiving of the older target. Arguably, the older adults’ experiences told them the older target had a better excuse for forgetting. Older persons also employ different social rules than younger persons when attributing blame or causality to a person’s behavior or to a social outcome (see the review in Blanchard-Fields, 1999). To the extent that social or other domain-specific knowledge increases or becomes more accurate, age may lead to greater accuracy in judgment. Hess and Auman (2001), for example, reviewed evidence suggesting that older adults may become more sophisticated in judgments of people and their behavior. In two studies, the authors found that relative to younger adults, older adults tended to give greater relative weight to more diagnostic information regarding a person than to less diagnostic information. Thus, at least when the evidence is primarily social, older adults may suffer less disadvantage in processing trial input. In this vein, Yates and Patalano (1999) suggested that aging jurors may develop more and more strongly held story schemas for a wide variety of circumstances that they may bring to bear on judgments of a particular case. If accurate, these strongly held schemas and stories among older jurors may facilitate accurate decision making. However, it should also be noted that relatively automatic application of this social knowledge base can become problematic if the juror fails to consider all information that might suggest an exception to general social-behavioral rules. Indeed, impaired executive function in younger adults has been shown to result in a pervasive memory bias favoring schema-consistent over schema-inconsistent material, whereas in unimpaired persons the bias is exactly the opposite (Macrae et al., 1999).The authors suggest that executive function is necessary to facilitate enhanced processing of (and thus better memory for) unexpected or schema-inconsistent material, and that failing executive function renders the mind less Firmly established personal opinions and values can also bias the manner in which a person processes evidence. For example, Klaczynski and Robinson (2001) illustrated the importance of Hess’s (1999) third age-related change in processing—that of motivation. The authors examined age differences in use of analytical vs. heuristic reasoning to support personal beliefs. Participants read vignettes with arguments supporting or attacking the value of their social groups (e.g., religion, social class). Both age groups tended to use heuristic reasoning to support belief-consistent arguments or evidence, whereas they tended to use scientific reasoning to support rejection of arguments contradicting their beliefs. Biased reasoning, however, was more common among middle-aged and older participants than among younger ones. Thus, particularly among older adults, motivation to discredit a particular point of view led to more systematic processing. A final relevant line of research has concerned the “theory of mind” (TOM) performance of older adults (reasoning about mental states and how they predict and explain behavior). Participants read stories requiring an inference about the characters’ thoughts and feelings and how they drive behavior. For example, in one story developed by Happe and colleagues (1998), a burglar has just robbed a shop. As he leaves the scene of the crime, a policeman sees him drop his glove. Not knowing of the burglary, the policeman wants to tell the burglar he has dropped his glove; however, when the policeman shouts out, the burglar gives himself up.


258

The participant is asked to explain why the burglar does this and thus must judge what the burglar thought about the policeman’s shout. Two of three studies on this issue have found poorer performance among older adults (Sullivan and Ruffman, 2004, and Maylor et al., 2002, vs. Happe et al, 1998). Sullivan and Ruffman further found that performance in older (but not younger) adults was predicted by fluid as well as crystallized abilities, and that when these abilities were controlled for, age differences were eliminated. The latter authors also found poorer memory for the stories among older adults. Both findings would suggest poorer memory and judgment among older jurors. Summary The literature reviewed here suggests several preliminary conclusions regarding differences that might be expected between younger and older jurors. On the average, older jurors will experience greater difficulty with hearing, seeing, understanding, and integrating trial input to form their judgments. These jurors are likely to engage in more schematic processing of trial input, using existing knowledge structures to guide processing and using less detail in specific testimony and evidence to guide their judgments. Once an initial impression is formed, it may exert greater influence on subsequent information processing for older than for younger jurors, causing them to use less critical judgment for evaluation of later evidence or to be more biased in processing that evidence. Throughout, older jurors will exert less effort to process and integrate apparently contradictory information carefully. In addition to general schematic processing, older jurors are specifically more prone to stereotype-based judgments. Thus, race and other heavily stereotyped social categories and issues can be expected to bias the judgments of older jurors to a greater degree than those of younger jurors. Furthermore, the specific stereotypes and social beliefs possessed by older adults are in many instances different from those of younger adults, so efforts to identify and take into account these strongly held values and beliefs will be important. Older jurors will also tend to remember less trial input overall and will be more likely to make source-monitoring errors regarding the source or credibility of statements or evidence and the difference between evidence actually presented and schema-related and other inferences they have drawn from it. In particular, older adults may be less likely to correctly remember false or discredited information as false rather than true, and thereby more likely to be influenced by poor-quality evidence. Moreover, because of failing inhibitory capacities, older adults may be less able to follow judicial instructions to ignore inappropriate statements or evidence. Perhaps in light of these difficulties, older persons report less confidence in their ability to serve well as jurors than younger persons do (Boatright, 2001). Despite at least some relative disadvantages in quality of processing of trial input, older jurors may be more committed to their judgments via choice-supportive biases in memory for choice-relevant evidence (Mather and Johnson, 2000; Mather et al, 2000) and more resistant to influence from other jurors (Pasupathi, 1999). Thus, their decisions may be relatively unresponsive to correction in deliberation. These conclusions must be considered in light of considerable individual variability in processing capacity among older jurors. Within older jurors, schematicity and reductions


259

in range of information use are related to working memory capacity and other measures of cognitive functioning. Also, even in the context of these limitations on processing capacity, older jurors may possess enhanced ability to recognize more diagnostic information—particularly regarding domains (such as human behavior) in which their life experiences have provided them with greater expertise than their younger counterparts. Finally, older adults can engage in deeper processing when highly motivated and in possession of sufficient cognitive resources. Thus, many older adults will behave very comparably to their younger counterparts. The study of age-related differences in juror behavior is still in its infancy. The cognitive aging literature provides evidence of age-related differences in cognitive processing and is a rich source of hypotheses regarding differences to be expected in juror behavior. However, it remains to conduct further research specifically with relevant trial-related stimulus materials and judgment tasks. Meanwhile, the consumer literature is also a rich source of findings of potentially great relevance to juror behavior. The arena of advertising effectiveness bears at least superficial resemblance to persuasion in trial because, in both cases, the recipient of persuasive messages must evaluate the relative credibility of competing claims for validity. 11.5 Overall Conclusions The literature reviewed herein reveals that age is crucial to functioning in the legal system. Age-related declines in cognitive functioning can cause, or contribute to causing, incidents that are later litigated in the legal system and can affect the performance of witnesses and jurors. However, although substantial empirical evidence has directly linked age-related decline in cognitive functions to accidents, such as those related to driving, very little research has directly examined the influence of age on witness testimony or juror decision making. The basic research on memory and decision making clearly documents age-related changes in these processes that would predict a number of specific interactions of age with particular witnessing variables and with variations in evidence, trial procedures, etc. that might affect jury decision making. Eyewitness researchers have begun to examine the effects of age on eyewitness identification and on susceptibility to the misinformation effect. However, other aspects of witness memory remain largely unaddressed. Moreover, the general area of age and juror decision making remains completely unexplored, with perhaps the single exception of the Fitzgerald (2000) study cited earlier. Finally, few studies of reactions to aging perpetrators or other case parties exist. Given the rapid aging of America, there is a clear and present need to develop better understanding of the many and varied ways in which aging victims, witnesses, and jurors can be expected to function differently than their younger counterparts. 11.6 Checklist: Relationships among Underlying Causes, Effects on Cognitive Processing, and Consequences for Victims, Witnesses, and Jurors Causes of Effects on cognitive processing Examples of


260

impaired processing

practical difficulties

Brain structure Reduced efficiency of neuronal transmission; impaired and pathology perceptual abilities; impaired formation and retention of memories; impaired efficiency and complexity of cognitive processing; impaired self-control and executive capacity

Causes of impaired processing Hearing

Effects on cognitive processing

Victims: Failure to smell leaking gas Witnesses: Reduced capacity to form long-term memories Jurors: Less accurate memory and failure to remember until deliberation

Examples of practical difficulties

Elevated sound detection thresholds; failure to Victims: understand speech; difficulty in localization Failure to understand instructions of sound leads to accident Witnesses: Inaccurate report of witnessed conversations Jurors: Failure to understand courtroom instructions, arguments, and testimony Vision Failure to see clearly, particularly in poor Victims: light; restriction in range of peripheral vision; Failure to see cars approaching difficulty in motion perception, tracking from side at intersection, failure to objects, visual marking, etc.; difficulty estimate speed of oncoming reading nonverbal cues during conversation vehicles Witnesses: Failure to report accurately on things seen accurately; failure to report correct object locations, failure to identify persons or objects correctly Jurors: Difficulty seeing courtroom exhibits, reading nonverbal cues to interpret speech or reading cues to deception Victims: Reduced speed Slowed performance overall; impaired of cognitive performance of complex tasks Slowed reaction times in processes emergency situations Witnesses: Enhanced impairment of perception and memory in complex circumstances Jurors: Reduction in elaborative


Reduced capacity of working memory

Failures of inhibition

Stereotype vulnerability

261

processing of trial input Victims: Victimization by scam artists due to failing ability to reason and critically evaluate proposals Witnesses: Less complete memory; failure of source monitoring such as susceptibility to misinformation effect Jurors: Greater use of relatively effortless schematic processing; failure to track source and credibility of testimony Inability to control attention; greater Victims: distractibility; reduced ability to divide Accidents due to enhanced attention; greater susceptibility to priming distractibility; failure to focus attention on most important tasks of driving Witnesses: Less memory for peripheral detail due to inability to divide attention between core event and context Jurors: Stronger effects of racial prejudice on judgment of minority defendant; inability to follow instructions to ignore inadmissible evidence Greater anxiety when stereotype salient; Victims: belief in applicability of negative stereotypes Elderly victim believes claims that to self; self-fulfilling prophecy he promised to send check to scam artist because of belief in fallibility of own memory Witnesses: Witness loses confidence in own testimony under cross examination because of belief in age-related memory impairment suggested by attorney Jurors: Failure to offer memories of evidence due to low confidence Reduced ability to hold multiple items of information in mind; reduced ability to carry out complex cognitive operations such as reasoning, math, etc.; reduced ability to multitask; failures of binding and source memory

Acknowledgment The authors would like to thank Michael Webster, Michael Crognale, and William P.Wallace for their helpful comments on previous drafts of this chapter.


262

Defining Terms Episodic memory—Memory for recently experienced events. Inhibitory processes—The ability to control attention so that attention is focused on important or task-relevant information while attention to irrelevant information is inhibited. Presbycusis—Decline in processing of auditory information. Processing speed—The speed at which basic cognitive operations occur. Source memory (source monitoring)—Memory for the context in which a core event occurred (such as memory for where a person was encountered, where or from whom information was obtained, where a conversation took place, etc.). Stereotype vulnerability—The tendency to act in stereotype consistent manner when relevant stereo-types are salient. Visual acuity—Ability to see clearly. Visuospatial processing—Processing of spatial location and relationships through vision; includes object identification and perception of location and motion. Working memory—The ability to hold information in short-term memory while simultaneously using that information to form judgments or solve problems. References Adams, C., Labouvie-Vief, G., Hobart, C.J., and Dorosz, M. (1990). Adult age group differences in story recall style. J. Gerontol: Psychol Sci. , 45, 17–27. Adams, C., Smith, M.C., Nyquist, L., and Perlmutter, M. (1997). Adult age-group differences in recall for the literal and interpretive meanings of narrative text. J. Gerontol. Series B: Psychol. Sci. Social Sci. , 4, 187–195. Adams-Price, C. (1992a). Eyewitness memory and aging research: a case study in everyday memory. In R.L.West and Sinnott, J.D. (Eds.), Everyday Memory and Aging: Current Research and Methodology (246–258). New York: Springer-Verlag. Adams-Price, C. (1992b). Eyewitness memory and aging: predictors of accuracy in recall and person recognition. Psychol Aging , 7, 602–608. Andres, P. (2003). Frontal cortex as the central executive of working memory: time to revise our view. Cortex , 39, 871–895. Anstey, K.J. and Maller, J.J. (2003). The role of volumetric MRI in understanding mild cognitive impairment and similar classifications. Aging Mental Health , 7, 238–250. Baddeley, A. (1986). Working Memory . Oxford, England: Clarendon Press. Ball, K., Owsley, C., Sloane, M.E., Roenker, D.L., and Bruni, J.R. (1993). Visual attention problems as a predictor of vehicle crashes in older drivers. Invest. Opthalmol. Visual Sci. , 34, 3110–3123. Ball, K., Roenker, D.L., and Bruni, J.R. (1990). Developmental changes in attention and visual search throughout adulthood. In J.Enns (Ed.), Advances in Psychology (Vol. 69, 489–508). Amsterdam: North-Holland-Elsevier Science. Baltes, P.B. and Smith, J. (1997). Emergence of powerful connection between sensory and cognitive functions across the adult life span: a new window to the study of cognitive aging? Psychol. Aging , 12, 12–21. Bargh, J.A. (1994). The four horsemen of automaticity: awareness, intention, efficiency, and control in social cognition. In R.S.Wyer and T.K.Srull (Eds.), Handbook of Social Cognition (1– 40). Hillsdale, NJ: Erlbaum.


263

Bargh, J.A., Chen, M., and Burrows, L. (1996). Automaticity of social behavior: direct effects of trait construct and stereotype activation on action. J. Personality Social Psychol. , 71, 230–244. Bartlett, J.C. (1993). Limits on losses in face recognition. In J.Cerella, Rybash, J., Hoyer, W., and Commons, M. (Eds.), Adult Information Processing: Limits on Loss (351–379). San Diego, CA: Academic Press. Bartlett, J.C. and Fulton, A. (1991). Familiarity and recognition of faces in old age. Memory Cognit. , 19, 229–238. Bartlett, J.C. and Leslie, J.E. (1986). Aging and memory for faces versus single views of faces. Memory Cognit , 14, 371–381. Bartlett, J.C., Leslie, J.E., Tubbs, A., and Fulton, A. (1989). Aging and memory for pictures of faces. Psychol. Aging , 4, 276–283. Bartlett, J.C., Strater, L., and Fulton, A. (1991). False recognition and false fame of faces in young adulthood and old age. Memory Cognit ., 19. Bartzokis, G. (2004). Age-related myelin breakdown: a developmental model of cognitive decline and Alzheimer’s disease. Neurobiol. Aging , 25, 5–18. Bashore, T.R., van der Molen, M.W., and Ridderinkhof, K. (1997). Is the age-complexity effect mediated by reductions in a general processing resource? Biological Psychology , 45(1–3), 263– 282. Bayen, U.J., Nakamura, G.V., Dupuis, S.E., and Yang, C.L. (2000). The use of schematic knowledge about sources in source monitoring. Memory Cognit ., 28, 480–500. Beatty, S. and Smith, S. (1987). External search effort: an investigation across several product categories. J. Consumer Res. , 14, 83–95. Benton, A.L., Sivan, A., Hamsher, K., Varney, M.R., and Spreen, O. (1994). Facial Recognition: Stimulus and Multiple Choice Pictures: Contributions to Neuropsychological Assessment . New York: Oxford University Press. Bigler, E.D., Lowry, C.M., Kerr, B., Tate, D.F., Hessel, C.D., Earl, H.D. et al. (2003). Role of white matter lesions, cerebral atrophy, and APOE on cognition in older persons with and without dementia: the Cache County, Utah, study of memory and aging. Neuropsychology , 17, 339–352. Birren, J.E. (1965). Age changes in speed of behavior: its central nature and physiological correlates. In A.T.Welford and J.E.Birren (Eds.), Behavior, Aging, and the Nervous System (191–216). Springfield, IL: Charles C.Thomas. Blanchard-Fields, F. (1999). Social schematicity and causal attributions. In T.M.Hess and F.Blanchard-Fields (Eds.), Social Cognition and Aging (222–237). San Diego, CA: Academic Press. Block, R.A. and Zakay, D. (1997). Prospective and retrospective duration judgments: a metaanalytic review. Psychonomic Bulletin and Review , 4, 184–197. Block, R.A., Zakay, D., and Hancock, P.A. (1998). Human aging and duration judgments: a metaanalytic review. Psychology and Aging , 13, 584–596. Boatright, R.G. (2001). Generational and age-based differences in attitudes towards jury service. Behavioral Sci. Law , 19, 285–304. Bornstein, B.H. (1995). Memory processes in elderly eyewitnesses: what we know and what we don’t know. Behav. Sci. Law , 13, 337–348. Bornstein, B.H. and Zickafoose, D.J. (1999). “I know I know it, I know I saw it.” The stability of the confidence-accuracy relationship across domains. J. Exp. Psychol.: Appl. , 5, 75–88. Bowles, R.P. and Salthouse, T.A. (2003). Assessing the age-related effects of proactive interference on working memory tasks using the Rasch model. Psychol. Aging , 18, 608–615. Brewer, N., Keast, A., and Rishworth, A. (2002). The confidence-accuracy relationship in eyewitness identification: the effects of reflection and disconfirmation on correlation and calibration. J. Exp. Psychol: Appl. , 8, 44–56. Brigham, J.C. and Williamson, N.L. (1979). Cross-racial recognition and age: when you’re over 60, do they still “all look alike?” Personality Social Psychol. Bull ., 5, 218–222.


264

Brimacombe, C.A.E., Quinton, N., Nance, N., and Garrioch, L. (1997). Is age irrelevant? Perceptions of young and old adult eyewitnesses. Law Hum. Behav. , 21, 619–634. Brown, A.S., Jones, E.M., and Davis, T.L. (1995). Age differences in conversational source monitoring. Psychol Aging , 10, 111–122. Bull, R. and Clifford, B.R. (1984). Earwitness voice recognition accuracy. In G.L.Wells and E.Loftus (Eds.), Eyewitness Testimony: Psychological Perspectives (92–123). Cambridge, MA: Cambridge University Press. Burton-Danner, K., Owsley, C., and Hackson, G.R. (2001). Aging and feature search: the effect of search area. Exp. Aging Res. , 27, 1–18. Caldwell, J.I. and Masson, M.E.J. (2001). Conscious and unconscious influences of memory for object location. Memory Cognit. , 29, 285–295. Cavanaugh, J.C. (2000). Metamemory from a social-cognitive perspective. In D.C.Park and N.Schwarz (Eds.), Cognitive Aging: a Primer (115–130). Philadelphia, PA: Psychology Press. Cerf-Ducastel, B. and Murphy, C. (2003). FMRI brain activation in response to odors is reduced in primary olfactory areas of elderly subjects. Brain Res. , 986, 39–53. Chalfonte, B.L. and Johnson, M.K. (1996). Feature memory and binding in young and old adults. Memory Cognit. , 24, 403–416. Charness, N. and Bosman, E.A. (1995). Compensation through environmental modification. In R.A. Dixon and L.Backman (Eds.), Compensating for Psychological Deficits and Declines: Managing Losses and Promoting Gains (147–168). Mahwah, NJ: Erlbaum. Chen, J., Hale, S., and Myerson, J. (2003). Effects of domain, retention interval, and information load on young and older adults’ visuospatial working memory. Aging, Neuropsychology, and Cognition , 10, 122–133. Chen, Y. and Blanchard-Fields, F. (2000). Unwanted thought: age differences in the correction of social judgments. Psychol Aging , 15, 475–482. Cherry, K.E. and Jones, M.W. (1999). Age-related differences in spatial memory: effects of structural and organizational context. J. Gen. Psychol. , 126, 53–73. Cherry, K.E., Park, D.C., Frieske, D.A., and Smith, A.D. (1996). Verbal and pictorial elaborations enhance memory in young and older adults. Aging, Neuropsychol, Cognit. , 3, 15–29. Christianson, S.-A. and Loftus, E.F. (1987). Memory for traumatic events. Appl. Cognitive Psychol , 1, 225–239. Coeckelbergh, T.R.M., Cornelissen, F.W., Brouwer, W.H., and Kooijman, A.C. (2004). Agerelated changes in the functional visual field: Further evidence for an inverse age X eccentricity effect. J. Gerontol: Psychol Sci. , 59B, 11–18. Cohen, G. (1979). Language comprehension in old age. Cognitive Psychol , 11, 423–439. Cohen, G. and Faulkner, D. (1989). Age differences in source forgetting: effects on reality monitoring and on eyewitness testimony. Psychol Aging , 4, 10–17. Cole, C.A. and Balasubramanian, S.K. (1993). Age differences in consumers’ search for information: public policy implications. J. Consumer Res. , 20, 157–169. Committee of Hearing, B., and Biomechanics. (1988). Speech understanding and aging. J. Acoustical Soc. Am. , 83, 859–895. Coxon, P. and Valentine, T. (1997). The effects of the age of eyewitnesses on the accuracy and suggestibility of their testimony. Appl Cognitive Psychol , 11, 415–430. Craik, F.I.M. (2000). Age related changes in human memory. In D.C.Park and N.Schwarz (Eds.), Cognitive Aging: a Primer (75–92). Philadelphia, PA: Psychology Press. Craik, F.I.M. and Byrd, M. (1982). Aging and cognitive deficits: the role of attentional resources. In F.I.M. Craik and S.Trehub (Eds.), Aging and Cognitive Processes (191–211). New York: Plenum Press. Craik, F.I.M. and Jennings, J.M. (1992). Human memory. In F.I.M.Craik and T.A.Salthouse (Eds.), The Handbook of Aging and Cognition (51–10). Mahwah, NJ: Erlbaum. Craik, F.I.M. and Lockhart, R.S. (1972). Levels of processing: a framework for memory research. J. Verbal Learning Verbal Behav. , 11, 671–684.


265

Craik, F.I.M. and Salthouse, T.A. (Eds.). (2000). The Handbook of Aging and Cognition . Mahwah, NJ: Lawrence Erlbaum. Crognale, M.A., Page, J.W., and Fuhrel, A. (2001). Aging of the chromatic onset visual evoked potential. Optometry Vision Sci. , 78, 442–446. Crook, T.H. and Larrabee, G.J. (1992). Changes in facial recognition memory across the adult life span. J. Gerontol. , 47, 138–141. Curcio, C.A. (2001). Photoreceptor topography in ageing and age-related maculopathy. Eye , 15, 376–383. Czaja, S.J., Sharit, J., Ownby, R., Roth, D.L., and Nair, S. (2001). Examining age differences in performance of a complex information search and retrieval task. Psychol. Aging , 16, 564–579. Davis, D. and Follette, W.C. (2001). Foibles of witness memory for traumatic/high profile events. J. Air Law Commerce , 66, 1421–1549. Davis, D. and Friedman, R.D. (in press). Memory for conversation: the orphan child of witness memory researchers. In M.P.Toglia, J.D.Read, D.R.Ross, and R.C.L.Lindsay (Eds.), Handbook of Eyewitness Psychology ( Vol. 1 ): Memory for Events . Mahwah, N.J.: Erlbaum. Davis, D., Kemmelmeier, M., and Follette, W.C. (2005). Memory for conversation on trial. In I.Noy and W.Karwowski (Eds.), Handbook of Human Factors in Litigation , Boca Raton, FL: CRC Press. Davis, D. and Loftus, E.E (in press). Internal and external sources of distortion in adult witness memory. In M.P.Toglia, J.D.Read, D.R.Ross and R.C.L.Lindsay (Eds.), Handbook of Eyewitness Memory ( Vol. 1 ): Memory for Events . Mahwah, N.J.: Erlbaum. de Groot, J.C., de Leeuw, F.E., Oudkerk, M., van Gijn, J., Hoffman, A., Jolles, J. et al. (2000). Cerebral white matter lesions and cognitive function: the Rotterdam Scan Study. Ann. Neurol. , 47, 145–151. de Heijer, T., Skoog, I., Oudkerk, M., de Leeuw, F.-E., de Groot, J.C., Hofman, A. et al. (2003). Association between blood pressure levels over time and brain atrophy in the elderly. Neurobiol. Aging , 24, 307–313. De Raedt, R. and Ponjaert-Kristoffersen, I. (2000). The relationship between cognitive/neuropsychological factors and car driving performance in older adults. J. Am. Geriatrics Soc. , 48, 1664–1668. De Raedt, R. and Ponjaert-Kristoffersen, I. (2001). Predicting at-fault car accidents of older drivers. Accident Anal. Prev. , 33, 809–819. Deary, I.J., Leaper, S.A., Murray, A.D., Staff, R.T., and Whalley, L.J. (2003). Cerebral white matter abnormalities and lifetime cognitive change: a 67-year follow-up of the Scottish Mental Survey of 1932. Psychol. Aging , 18, 140–148. Devine, P.G. (1989). Stereotypes and prejudice: their automatic and controlled components. J. Personality Social Psychol. , 56, 5–18. Dodson, C.S., Johnson, M.K., and Schooler, J.W. (1997). The verbal overshadowing effect: why descriptions impair face recognition. Memory Cognit. , 25, 129–139. Dodson, C.S. and Schacter, D.L. (2002). Aging and strategic retrieval processes: reducing false memories with a distinctiveness heuristic. Psychol. Aging , 17, 405–415. Dollinger, S.M.C. and Hoyer, W.J. (1996). Age and skill differences in the processing demands of visual inspection. Appl Cognit. Psychol , 10, 225–239. Dufouil, C., Alperovitch, A., and Tzourio, C. (2003). Influence of education on the relationship between white matter lesions and cognition. Neurology , 60, 831–836. Dulay, M.F. and Murphy, C. (2002). Olfactory acuity and cognitive function converge in older adulthood: support for the common cause hypothesis. Psychol. Aging , 17, 392–404. Dysart, J.E. and Lindsay, R.C.L. (2001). A preidentification questioning effect: serendipitously increasing correct rejections. Law Hum. Behav. , 25, 155–166. Earles, J.L. and Kersten, A. (1998). Influences of age and perceived activity difficulty differences on activity recall. J. Gerontol.: Psychol. Sci. , 53B, 33–41.


266

Eby, D.W., Tronbley, D.W., Molnar, L.J., and Shope, J.R. (1998). The assessment of older drivers’ capabilities: a review of the literature (No. UMTRI-98–24). Ann Arbor: MI: University of Michigan Transportation Research Institute. Echt, K.V. (2002). Designing web-based health information for older adults: visual considerations and design directives. In R.W.Morrell (Ed.), Older Adults, Health Information and the World Wide Web (61–87). Mahwah, NJ: Erlbaum. Enoch, J.M., Werner, J.S., Haegerstrom-Portnoy, G., Lakshminarayanan, V., and Rynders, M. (1999). Forever young: visual functions not affected or minimally affected by aging: a review. J. Gerontol: Biol. Sci. , 54A, B336–B351. Erber, J.T., Szuchman, L.T., and Prager, I.G. (2001). Ain’t misbehavin’: the effects of age and intentionality on judgments about misconduct. Psychol Aging , 16, 85–95. Evans, G.W., Brennan, P.L., Skorpovich, M.A., and Held, D. (1984). Cognitive mapping and elderly adults: verbal and location memory for urban landmarks. J. Gerontol. , 39, 452–457. Faubert, J. (2002). Visual perception and aging. Can. J. Exp. Psychol , 56, 164–176. Ferguson, S.A., Hashtroudi, S., and Johnson, M.K. (1992). Age differences in using source-relevant cues. Psychol. Aging , 7, 443–452. Ferris, S.H., Crook, T., Clark, B., McCarthy, M., and Rae, D. (1980). Facial recognition memory deficits in normal aging and senile dementia. J. Gerontol ., 35, 707–714. Finucane, M.L., Slovic, P., Hibbard, J.H., Peters, E., Mertz, C.K., and MacGregor, D.G. (2002). Aging and decision-making competence: an analysis of comprehension and consistency skills in older vs. younger adults considering health-plan options. J. Behav. Decision Making , 15, 141– 164. Fisher, R.P. and Geiselman, R.E. (1992). Memory-enhancing techniques for investigative interviewing: the cognitive interview. Unpublished manuscript, Springfield. Fisk, S.T. and Taylor, S.E. (1991). Social Cognition . New York: McGraw-Hill. Fisk, A.D. and Fisher, D.L. (1994). Brinley plots and theories of aging: The explicit, muddled, and implicit debates. J. Gerontol. , 49(2), P81–P89. Fitzgerald, J.M. (2000). Younger and older jurors: the influence of environmental supports on memory performance and decision making in complex trials. J. Gerontol: Psychol. Sci. , 55B, 323–331. Fozard, J.L. and Gordon-Salant, S. (2001). Changes in vision and hearing with age. In J.E.Birren (Ed.), Handbook of the Psychology of Aging (5th ed., 241–266). San Diego, CA: Academic Press. Fulton, A. and Bartlett, J.C. (1991). Young and old faces in young and old heads: the factor of age in face recognition. Psychol. Aging , 6, 623–630. Furse, D., Girish, P., and Stewart, D. (1984). A typology of individual search strategies among purchasers of new automobiles. J. Consumer Res. , 10, 417–4431. Geiselman, R.E., Tubridy, A., Barroso, A., McClean, R., Mazafarian, M., and Zoumberakis, H. (2001). Benton Facial Recognition Test scores: their effect on juror verdicts. Am. J. Forensic Psychol. , 19, 89–97. Gilbert, D.T., Krull, D.S., and Malone, P.S. (1990). Unbelieving the unbelievable: some problems in the rejection of false information. J. Personality Soc. Psychol. , 59, 601–613. Gilbert, D.T., Tafarodi, R.W., and Malone, P.S. (1993). You can’t believe everything you read. J. Personality Soc. Psychol , 65, 221–233. Glisky, E.L. (2001). Source memory, aging, and the frontal lobes. In M.Naveh-Benjamin, Moscovitch, M., and Roediger III, H.L. (Eds.), Perspectives on Human Memory and Cogntivie Aging: Essays in Honour of Fergus Craik (265–276). New York: Psychology Press. Glisky, E.L., Polster, M.R., and Routhieaux, B.C. (1995). Double dissociation between item and source memory. Neuropsychology , 9, 229–235. Grady, C.L. (2001). Age-related changes in the functional neuroanatomy of memory. In M.NavehBenjamin, Moscovitch, M., and Roediger, H.L. III (Eds.), Perspectives on Human


267

Memory and Cognitive Aging: Essays in Honour of Fergus Craik (325–333). New York: Psychology Press. Grady, C.L. (2002). Age-related differences in face processing: a meta-analysis of three functional neuroimaging experiments. Can.J. Exp. Psychol , 56, 208–220. Grady, C.L., Bernstein, L.J., Beig, S., and Siegenthaler, A.L. (2002). The effects of encoding task on age-related differences in the functional neuroanatomy of face memory. Psychol. Aging , 17, 7–23. Grigsby, J., Rosenberg, N.L., and Busenbark, D. (1995). Chronic pain is associated with deficits in information processing. Perceptual Motor Skills , 81, 403–410. Groth, A.N. (1979). The older rape victim and her assailant. J. Geriatr. Psychiatry , 11, 203–215. Guillozet, A.L., Weintraub, S., Mash, D.C., and Mesulam, M.M. (2003). Arch. Neurol 60, 5, 729– 736. Gunning-Dixon, F.M., Gur, R.C., Perkins, A.C., Schroeder, L, Turner, T., Turetsky, B.I. et al. (2003). Age-related differences in brain activation during emotional face processing. Neurobiol Aging , 24, 285–295. Gunning-Dixon, F.M. and Raz, N. (2000). The cognitive correlates of white matter abnormalities in normal aging: a quantitative review. Neuropsychology , 14, 224–232. Hakamies-Blomqvist, L.E. (1993). Fatal accidents of older drivers . Accident Anal Prev. , 25, 19– 27. Hakamies-Blomqvist, L.E. (1994). Aging and fatal accidents in male and female drivers. J. Gerontol. , 4, 286–290. Hamm, B.P. and Hasher, L. (1992). Age and the availability of inferences. Psychol Aging , 7, 56– 64. Happe, F.G.E., Winner, E., and Brownell, H. (1998). The getting of wisdom: theory of mind in old age. Dev. Psychol , 34, 358–362. Hartley, A.A. (1992). In Craik, F.I.M. and Salthouse, T.A. (Eds.). Handbook of Aging and Cognition . Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Hasher, L., Stoltzfus, E.R., Zacks, R.T., and Rypma, B. (1991). Age and inhibition. J. Exp. Psychol: Learning, Memory, Cognit. , 17, 163–169. Hasher, L. and Zacks, R.T. (1988). Working memory, comprehension, and aging: a review and a new view. In G.H.Bower (Ed.), The Psychology of Learning and Motivation (193–225). San Diego, CA: Academic Press. Hasher, L., Zacks, R.T., and May, C.P. (1999). Inhibitory control, circadian arousal, and age. In D.Gopher and Koriat, A. (Eds.). Attention and Performance (Vol. XVII, 653–675). Cambridge, MA: MIT Press. Hashtroudi, S., Johnson, M.D., and Chrosniak, L.D. (1989). Aging and source monitoring. Psychol. Aging , 4,106–112. Hastie, R. (1981). Schematic principles in human memory. In E.T.Higgins, C.P.Herman and M.P.Zanna (Eds.). Social Cognition: The Ontario Symposium (Vol. 1, 39–88). Hillsdale, NJ: Erlbaum. Hastie, R. (1984). Causes and effects of causal attribution. J. Personality Social Psychol , 46, 44– 56. Hausdorff, J.M., Levy, B.R., and Wei, J.Y. (1999). The power of ageism on physical function of older persons: reversibility of age-related gait changes. J. Am. Geriatrics Soc. , 47, 1346–1349. Heckhausen, J., Dixon, R.A., and Baltes, P.B. (1989). Gains and losses in development throughout adulthood as perceived by different adult age groups. Dev. Psychol , 25, 109–121. Heckhausen, J. and Krueger, J. (1993). Developmental expectations for the self and most other people: age grading in three functions of social comparison. Dev. Psychol , 29, 539–548. Hedden, T. and Park, D.C. (2001). Aging and interference in verbal working memory. Psychol Aging , 16, 666–681. Hellstrom, L.I. and Schmiedt, R.A. (1990). Compound action potential input/output functions in young and quiet-aged gerbils. Hearing Res. , 50, 163–174.


268

Hertzog, C., Lineweaver, T.T., and McGuire, C.I. (1999). Beliefs about memory and aging. In T.M.Hess and F.Blanchard-Fields (Eds.), Social Cognition and Aging (43–69). San Diego, CA: Academic Press. Hess, T.M. (1999). Cognitive and knowledge-based influences on social representations. In T.M.Hess and F.Blanchard-Fields (Eds.), Social Cognition and Aging (239–263). San Diego, CA: Academic Press. Hess, T.M. and Auman, C. (2001). Aging and social expertise: the impact of trait-diagnostic information on impressions of others. Psychol. Aging , 16, 497–510. Hess, T.M., Auman, C., Colcombe, S.J., and Rahhal, T.A. (2003). The impact of stereotype threat on age differences in memory performance. J. Gerontol: Psychol. Sci. , 58B, 3–11. Hess, T.M., McGee, K.A., Woodburn, S.M., and Bolstad, C.A. (1998). Age-related priming effects in social judgments. Psychol Aging , 13, 127–137. Hess, T.M. and Slaughter, S.J. (1990). Schematic knowledge influences on memory for scene information in young and older adults. Dev. Psychol , 26, 855–865. Ho, G., Scialfa, C.T., Caird, J.K., and Graw, T. (2001). Visual search for traffic signs: the effects of clutter, luminance, and aging. Hum.Factors , 43, 194–207. Humes, L.E. (1996). Speech understanding in the elderly. J. Am. Acad. Audiol , 7, 161–167. Hummert, M.L., Garstka, T.A., Shaner, J.L., and Strahm, S. (1994). Stereotypes of the elderly held by young, middle-aged, and elderly adults. J. Gerontol.: Psychol Sci. , 49, 240–249. Isaacowitz, D.M., Charles, S.T., and Carstensen, L.L. (2000). Emotion and cognition. In F.I.M.Craik and T.A. Salthouse (Eds.), The Handbook of Aging and Cognition (2nd ed., 593– 631). Mahwah, NJ: Erlbaum. Jacoby, L.L. (1991). A process dissociation framework: separating automatic from intentional uses of memory. J. Memory Language , 30, 513–541. Jacoby, L.L., Marsh, E.J., and Dolan, P.O. (2001). Forms of bias: age-related changes in memory and cognition. In M.Naveh-Benjamin, Moscovitch, M., and Roediger, H.L. (Eds.), Perspectives on Human Memory and Cognitive Aging (240–252). New York: Psychology Press. Jenkins, L., Myerson, J., Joerding, J.A., and Hale, S. (2000). Converging evidence that visuospatial cognition is more age-sensitive than verbal cognition. Psychol. Aging , 15, 157–175. Johnson, C.A. (1986, February). Peripheral visual fields and driving in an aging population. Paper presented at the Invitational conference on work, aging, and vision, National Research Council, National Academy of Science, Committee on Vision, Washington, D.C. Johnson, M.M.S. (1990). Age differences in decision making: a process methodology for examining strategic information processing. J. Gerontol.: Psychol Sci. , 45, 75–78. Johnson, M.M.S. (1993). Thinking about strategies during, before, and after making a decision. Psychol. Aging , 8, 231–241. Johnson-Laird, P.N. (1983). Mental Models . Cambridge, MA: Harvard University Press. Johnson-Laird, P.N. (1989). Mental models. In M.I.Posner (Ed.), Foundations of Cognitive Science (pp. 469–499). Cambridge, MA: MIT Press. Karpel, M.E., Hoyer, W.J., and Toglia, M.P. (2001). Accuracy and qualities of real and suggested memories: nonspecific age differences. J. Gerontol.: Psychol. Sci. , 56B, 103–110. Kausler, D.H. and Puckett, J.M. (1981). Adult age differences in memory for modality attributes. Exp. Aging Res. , 7, 117–125. Kemper, S., Herman, R.E., and Lian, C.H.T. (2003). The costs of doing two things at once for young and older adults: talking while walking, finger tapping, and ignoring speech or noise. Psychol. Aging , 18, 181–192. Kemper, S. and Kemtes, K. (2000). Aging and message production and comprehension. In D.C.Park and N.Schwarz (Eds.), Cognitive Aging: a Primer (197–213). Philadelphia, PA: Psychology Press. Kite, M.E. and Johnson, B.T. (1988). Attitudes toward older and younger adults: a meta-analysis. Psychol. Aging , 3, 233–244.


269

Klaczynski, P.A. and Robinson, B. (2000). Personal theories, intellectual ability, and epistemological beliefs: Adult age differences in everyday reasoning biases. Psychol. Aging , 15, 400–416. Klavora, P. and Heslegrave, R.J. (2002). Senior drivers: an overview of problems and intervention strategies. J. Aging Phys. Activity , 10, 322–335. Kline, D., Kline, T.J.B., Fozard, J.L., Kosnik, W., Schieber, E, and Sekuler, R. (1992). Vision, aging, and driving: the problems of older drivers. J. Gerontol.: Psychological Sci. , 47, 27–34. Kline, D. and Scialfa, C.T. (1997). Sensory and perceptual functioning: basic research and human factors implications. In A.D.Fisk and W.A.Rogers (Eds.), Handbook of Human Factors and the Older Adult (296–333). San Diego, CA: Academic Press. Kline, D.W. and Scialfa, C.T. (1996). Visual and auditory aging. In J.E.Firren and K.W.Schaie (Eds.), Handbook of the Psychology of Aging (181–203). San Diego, CA: Academic Press. Kosnik, W., Winslow, L., Kline, D., Rasinski, K., and Sekuler, R. (1988). Visual changes in daily life throughout adulthood. J. Gerontol.: Psychol. Sci. , 43, 63–70. Koutstaal, W. and Schacter, D.L. (1997). Gist-based false recognition of pictures in older and younger adults. J. Memory Language , 37, 555–583. Koutstaal, W. and Schacter, D.L. (2001). Memory distortion and aging. In M.Naveh-Benjamin, Moscovitch, M., and Roediger, H.L. (Eds.), Perspectives on Human Memory and Cognitive Aging: Essays in Honour of Fergus Craik (362–383). New York: Psychology Press. Koutstaal, W., Schacter, D.L., and Brenner, C. (2001). Dual task demands and gist-based false recognition of pictures in younger and older adults. J. Memory Cognit. , 44, 399–426. Kramer, A.G. and Atchley, P. (2000). Age-related effects in the marking of old objects in visual search. Psychol. Aging , 15, 286–296. Kunda, Z. (1999). Social Cognition: Making Sense of People . Cambridge, MA: Bradford. Kwong See, S.T., Hoffman, H.G., and Wood, T.L. (2001). Perceptions of an old female eyewitness: is the older eyewitness believable? Psychol. Aging , 16, 346–350. Langer, E., Perlmuter, L, Chanowitz, B., and Rubin, R. (1988). Two new applications of mindlessness theory: alcoholism and aging. J. Aging Stud. , 2, 289–299. Larsson, M., Finkel, D., and Pedersen, N.L. (2000). Odor identification: Influences of age, gender, cognition, and personality. J. Gerontol.: Psychological Sci. , 55B, 304–310. Law, S., Hawkins, S.A., and Craik, F.I.M. (1998). Repetition-induced belief in the elderly: rehabilitating age-related memory deficits. J. Consumer Res. , 25, 91–107. Lawrence, B., Myerson, J., and Hale, S. (1998). Differential decline of verbal and visuospatial processing speed across the adult life span. Aging, Neuropsychology & Cognition , 5, 129–146. Lehrner, J.P., Glueck, J., and Laska, M. (1999). Odor identification, consistency of label use, olfactory threshold and their relationships to odor memory over the human lifespan. Chem. Senses , 24, 337–346. Levy, B. (1996). Improving memory in old age through implicit self-stereotyping. J. Personality Social Psychol. , 71, 1092–1107. Levy, B. (2000). Handwriting as a reflection of aging self-stereotypes. J. Geriatr. Psychiatry , 33, 81–94. Levy, B. (2003). Mind matters: cognitive and physical effects of aging self-stereotypes. J. Gerontol.: Psychol. Sci., 58B, 203–211. Levy, B., Ashman, O., and Dror, I. (1999–2000). To be or not to be: the effects of aging selfstereotypes on the will-to-live. Omega: J. Death Dying , 40, 409–420. Levy, B., Hausdorff, J., Hencke, R., and Wei, J.Y. (2000). Reducing cardiovascular stress with positive self-stereotypes of aging. J. Gerontol.: Psychol. Sci. , 55B, 205–213. Levy, B. and Langer, E. (1994). Aging free from negative stereotypes: successful memory in China and among the American deaf. J. Personality Social Psychol. , 66, 989–997. Levy, B., Slade, M.D., and Kasi, S.V. (2002a). Increased longevity by positive self-perceptions of aging. J. Personality Social Psychol. , 83, 261–270.


270

Levy, B., Slade, M.D., Kunkel, S., and Kasi, S.V. (2002b). Longitudinal benefit of positive selfperceptions of aging on functioning health. J. Gerontol.: Psychol. Sci. , 57B, 409–417. Light, L.L. (1991). Memory and aging: four hypotheses in search of data. Annu. Rev. Psychol. , 42, 333–376. Lindenberger, U. and Baltes, P.B. (1994). Sensory functioning and intelligence in old age: a strong connection. Psychol. Aging , 9, 339–355. Lindenberger, U. and Baltes, P.B. (1997). Intellectual functioning in old and very old age: crosssectional results from the Berlin Aging Study. Psychol. Aging , 12, 410–432. Lindsay, D.S. and Johnson, M.K. (1989). The eyewitness suggestibility effect and memory for source. Memory & Cognition , 17, 349–358. Lipman, P.D. and Caplan, L.J. (1992). Adult age differences in memory for routes: effects of instruction and spatial diagram. Psychol. Aging , 7, 435–442. List, J. (1986). Age and schematic differences in the reliability of eyewitness testimony. Dev. Psychol. , 22, 50–57. Loftus, E.F. (Ed.). (1979). Eyewitness Testimony . Cambridge, MA: Harvard University Press. Loftus, E.F. (1992). When a lie becomes memory’s truth: memory distortion after exposure to misinformation. Curr. Directions Psychol. Sci. , 1, 121–123. Loftus, E.F. and Hoffman, H.G. (1989). Misinformation and memory: the creation of new memories. J. Exp. Psychol.: Gen. , 118, 100–104. Loftus, E.F. and Ketcham, K. (1994). The Myth of Repressed Memory . New York: St. Martin’s Press. Loftus, E.F., Miller, D.G., and Burns, H.J. (1978). Semantic integration of verbal information into visual memory. J. Exp. Psychol.: Hum. Learning Memory , 4, 19–31. Loftus, E.F., Schooler, J.W., Boone, S.M., and Kline, D. (1987). Time went by so slowly: overestimation of event duration by males and females. Appl. Cognitive Psychol. , 1, 1–13. Lubinski, R. and Higginbotham, M. (1997). Communication Technologies for the Elderly: Hearing, Speech, and Vision . San Diego, CA: Singular. Lustig, C., Hasher, L, and Tonev, S.T. (2001). Inhibitory control over the present and the past. Eur. J. Cognitive Psychol. , 13, 107–122. Macrae, C.N., Bodenhausen, G.V., Schloerscheidt, A.M., and Milne, A.B. (1999). Tales of the unexpected: executive function and person perception. J. Personality Social Psychol , 76, 200– 213. Madden, D.J., Turkington, T.G., Provenzale, J.M., Denny, L.L., Langley, L.K., Hawk, T.C. et al. (2002). Aging and attentional guidance during visual search: functional neuroanatomy by positron emission tomography. Psychol Aging , 17, 24–43. Malhotra, N. (1982). Information load and consumer decision making. J. Consumer Res. , 8, 419– 430. Maltz, M. and Shinar, D. (1999). Eye movements of younger and older drivers. Hum. Factors , 41, 15–25. Mariske, M., Klumb, P., and Baltes, M.M. (1997). Everyday activity patterns and sensory functioning in old age. Psychol. Aging , 12, 444–457. Mason, S.E. (1986). Age and gender as factors in facial recognition and identification. Exp. Aging Res. , 12, 151–154. Mather, M. and Johnson, M.K. (2000). Choice-supportive source monitoring: do our decisions seem better to us as we age? Psychol Aging , 15, 596–606. Mather, M., Johnson, M.K., and De Leonardis, D.M. (1999). Stereotype reliance in source monitoring: Age differences and neuropsychological test correlates. Cognitive Neuropsychol , 16, 437–458. Mather, M., Shafir, E., and Johnson, M.K. (2000). Misremembrance of options past: source monitoring and choice. Psychol Sci. , 11, 132–138. May, C.P. (1999). Synchrony effects in cognition: the costs and a benefit. Psychonomic Bull Rev. , 6, 142–147.


271

May, C.P. and Hasher, L. (1998). Synchrony effects in inhibitory control over thought and action. J. Exp. Psychol: Hum. Perception Performance , 24, 363–379. May, C.P, Hasher, L., and Stolzfus, E.R. (1993). Optimal time of day and the magnitude of age differences in memory. Psychol. Sci. , 4, 326–330. Maylor, E.A. (1997). Proper name retrieval in old age: converging evidence against disproportionate impairment. Aging, Neuropsychol, Cognit. , 4, 211–226. Maylor, E.A., Joulson, J.M., Muncer, A.-M., and Taylor, L. (2002). Does performance on theory of mind tasks decline in old age? Br. J. Psychol , 93, 465–485. McCormack, T, Brown, G.D.A., Maylor, E.A., Richardson, L.B.N., and Darby, R.J. (2002). Effects of aging on absolute identification of duration. Psychol Aging , 17, 363–378. McDowd, J.M., Oseas-Kreger, D.M., and L., F.D. (1995). Inhibitory processes in cognition and aging. In E.N.Dempster and C.J.Brainerd (Eds.), Interference and Inhibition in Cognition (pp. 363–400). San Diego, CA: Academic Press. McDowd, J.M. and Shaw, R.J. (2000). Attention and aging: a functional perspective. In F.I.M.Craik and T.A.Salthouse (Eds.), The Handbook of Aging and Cognition (2nd ed., 221– 292). Mahwah, NJ: Erlbaum. McDowell, C.L., Harrison, D.W., and Demaree, H.A. (1994). Is right hemisphere decline in the perception of emotion a function of aging? Int. J. Neurosci. , 79, 1–11. McMahon, M. (2000). The effect of the enhanced cognitive interview on recall and confidence in elderly adults. Psychiatry, Psychol Law , 7, 9–32. Mello, E.W. and Fisher, R.P. (1996). Enhancing older adult eyewitness memory with the cognitive interview. Appl Cognitive Psychol , 10, 403–417. Memon, A. and Bartlett, J.C. (2002). The effects of verbalization on face recognition in young and older adults. Appl Cognitive Psychol , 16, 635–650. Memon, A. and Gabbert, R (2003a). Improving the identification accuracy of senior witnesses: do prelineup questions and sequential testing help? J. Appl Psychol , 88, 341–347. Memon, A. and Gabbert, R (2003b). Unravelling the effects of sequential presentation in culprit present lineups. Appl. Cognitive Psychol , 17, 703–714. Memon, A. Bartlett, J., Rose, R., and Gray, C. (2003). The aging eyewitness: effects of age on face, delay, and source-memory ability. J. Gerontol: Psychol. Sci. , 58B, 338–345. Memon, A., Hope, L., Bartlett, J., and Bull, R. (2002). Eyewitness recognition errors: the effects of mugshot viewing and choosing in young and old adults. Memory Cognit. , 30, 1219–1227. Memon, A., Hope, L., and Bull, R. (2003). Exposure duration: effects on eyewitness accuracy and confidence. Br. J. Psychol , 94, 339–354. Mergler, N.L., Faust, M., and Goldstein, M.D. (1984/1985). Storytelling as an age-dependent skill: oral recall of orally presented stories. Int. J. Aging Hum. Dev. , 20, 205–228. Messier, C., Tsiakas, M., Gagnon, M., Desrochers, A., and Awad, N. (2003). Effect of age and glucoregulation on cognitive performance. Neurobiol Aging , 24, 985–1003. Meyer, B.J.G., Russon, C., and Talbot, A. (1995). Discourse comprehension and problems solving: decisions about the treatment of breast cancer by women across the life span. Psychol. Aging , 10, 84–103. Michaels, D.D. (1993). Ocular disease in the elderly. In A.A.Rosenbloom and M.W.Morgan (Eds.), Vision and Aging (111–159). Boston, MA: Butterworth-Heinemann. Mingus, D., Reed, B.R., Jagust, W.J., DeCarli, C., Mack, W.J., Kramer, J.H. et al. (2002). Volumetric MRI predicts rate of cognitive decline related to AD and cerebrovascular disease. Neurology , 59, 867–873. Mitchell, K.J., Johnson, M.K., and Mather, M. (2003). Source monitoring and suggestibility to misinformation: adult age-related differences. Appl. Cognitive Psychol , 17, 107–119. Moffat, S.D. and Resnick, S.M. (2002). Effects of age on virtual environment place navigation and allocentric cognitive mapping. Behav. Neurosci. , 116, 851–859. Moffat, S.D., Zonderman, A.B., and Resnick, S.M. (2001). Age differences in spatial memory in a virtual environment navigation task. Neurobiol Aging , 22, 787–796.


272

Moore, B.C.J. (1989). An Introduction to the Psychology of Hearing . London: Academic Press. Morgan, R. and King, D. (1995). The older driver: a review. Postgrad. Med. J. , 71, 525–528. Mortimer, J.A., Snowdon, D.A., and Markesbery, W.R. (2003). Head circumference, education and risk of dementia: findings from the Nun study. J. Clin. Exp. Neuropsychol Special issue: cognitive reserve, 25, 671–679. Moscovitch, M. and Winocur, G. (1995). Frontal lobes, memory, and aging. Ann. NY Acad. Sci. , 769, 119–150. Multhaup, K.S. (1995). Aging, source, and decision criteria: when false fame errors do and do not occur. Psychol Aging , 10, 492–497. Multhaup, K.S., De Leonardis, D.M., and Johnson, M.K. (1999). Source memory and eyewitness suggestibility in older adults. J. Gen. Psychol , 126, 74–84. Myerson, J., Emery, L., White, D.A., and Hale, S. (2003a). Effects of age, domain, and processing demands on memory span: evidence for differential decline. Aging, Neuropsychology, and Cognition , 10, 20–27. Myerson, J., Hale, S., Zheng, Y., Jenkins, L., and Widaman, K.F. (2003b). The difference engine: a model of diversity in speeded cognition. Psychonomic Bull Rev. , 10, 262–288. National Institute on Aging. (1996). Crime and older people, from http://www.nia.nih.gov/. Nelson, T.D. (2002). Ageism: Stereotyping and Prejudice Against Older Persons . Cambridge, MA: MIT Press. Nielsen, K. and Peters, A. (2000). The effects of aging on the frequency of nerve fibers in rhesus monkey striate cortex. Neurobiol Aging , 21, 621–628. Nielson, K.A., Langenecker, S.A., and Garavan, H. (2002). Differences in the functional neuroanatomy of inhibitory control across the adult life span. Psychol Aging , 17, 56–71. Nilsson, L.-G. and Soderlund, H. (2001). Aging, cognition and health. In M.Naveh-Benjamin, Moscovitch, M., and Roediger III, H.L. (Eds.). Perspectives on Human Memory and Cognitive Aging (253–264). New York: Psychology Press. Norman, I. and Schacter, D.L. (1997). False recognition in younger and older adults: exploring the characteristics of illusory memories. Memory Cognit. , 25, 838–848. Nunez, N., McCoy, M.L., Clark, H.L., and Shaw, L.A. (1999). The testimony of elderly victim/witnesses and their impact on juror decisions: the importance of examining multiple stereotypes. Law Hum. Behav. , 23, 413–423. Nussbaum, J.F., Pecchioni, L.L., Robinson, J.D., and Thompson, T.L. (2000). Communication and Aging , 2nd ed. Mahwah, NJ: Erlbaum. Olsson, N. (2000). A comparison of correlation, calibration, and diagnosticity as measures of the confidence accuracy relationship in witness identification. J. Appl Psychol , 85, 504–511. O’Rourke, T.E., Penrod, S.D., Cutler, B.L., and Stuve, T.E. (1989). The external validity of eyewitness identification research: generalizing across subject populations. Law Hum. Behav. , 13, 385–395. Owsley, C., Ball, K., McGwin, G., Jr., Sloane, M.E., Roenker, D.L., White, M.E et al. (1998). Visual processing impairment and risk of motor vehicle crash among older adults. J. Am. Med. Assoc ., 279, 1083–1088. Owsley, C., Ball, K., Sloane, M.E., Roenker, D.L., and Bruni, J.R. (1991). Visual/cognitive correlates of vehicle accidents in older drivers. Psychol Aging , 6, 403–415. Panek, P.E., Barrett, G.V., and Sterns, H.L. (1977). A review of age-related changes in perceptual information processing ability with regard to driving. Exp. Aging Res. , 3, 387–399. Pantel, J., Kratz, B., Essig, M., and Schroeder, J. (2003). Parahippocampal volume deficits in subjects with aging-associated cognitive decline. Am. J. Psychiatry , 160, 379–382. Park, D. and Schwarz, N. (Eds.). (2000). Cognitive Aging: a Primer . Philadelphia, PA: Psychology Press. Park, D.C. (1999). Aging and the controlled and automatic processing of medical information and medical intentions. In D.Park, R.W.Morrell, and K.C.Chifren (Eds.), Processing of Medical


273

Information in Aging Patients: Cognitive and Human Factors Perspectives . Mahwah, NJ: Erlbaum. Park, D.C. (2000). The basic mechanisms accounting for age-related decline in cognitive function. In D.C. Park and N.Schwarz (Eds.), Cognitive Aging: a Primer (3–21). Philadelphia, PA: Psychology Press. Park, D.C. and Hedden, R. (2001). Working memory and aging. In M.Naveh-Benjamin, Moscovitch, M., and Roediger, H.L. (Eds.), Perspectives on Human Memory and Cognitive Aging: Essays in Honour of Fergus Craik (148–160). New York: Psychology Press. Park, D.C. and Gutchess, A.H. (2000). Cognitive aging in everyday life. In D.C.Park and N.Schwarz (Eds.), Cognitive Aging: a Primer (217–232). Philadelphia: Psychology Press. Park, D.C., Lautenschlager, G., Hadden, T, Davidson, N.S., Smith, A.D., and Smith, P.K. (2002). Models of visuospatial and verbal memory across the adult life span. Psychol. Aging , 17, 299– 320. Park, D.C. and Shaw, R.J. (1992). Effect of environmental support on implicit and explicit memory in younger and older adults. Psychol. Aging , 7, 632–642. Park, D.C., Smith, A.D., Morrell, R.W, Puglisi, J.T., and Dudley, W.N. (1990). Effects of contextual integration on recall of pictures by older adults. J. Gerontol: Psychol. Sci. , 45, 52– 57. Park, D.C., Willis, S.L., Morrow, D., Diehl, M., and Gaines, C.L. (1994). Cognitive function and medication usage in older adults. J. Appl. Gerontol. , 13, 39–57. Pasupathi, M. (1999). Age differences in response to conformity pressure for emotional and nonemotional material. Psychol. Aging , 14, 170–174. Pennington, N. and Hastie, R. (1993). The story model of juror decision making. In R.Hastie (Ed.), Inside the Juror (192–221). New York: Cambridge University Press. Perfect, T.J. (1994). What can Brinley plots tell us about cognitive aging?, J. Gerontology , 49, 60– 64. Persad, C.C., Aeles, N., Zacks, R.T., and Denburg, N.L. (2002). Inhibitory changes after age 60 and the relationship to measures of attention and memory. J. Gerontol: Series B: Psychol. Sci. Social Sci. , 57B, 223–232. Peters, A., Moss, M.B., and Sethares, C. (2000). Effects of aging on myelinated nerve fibers in monkey primary visual cortex. J. Comparative Neurol. , 419, 364–376. Peters, A., Moss, M.B., and Sethares, C. (2001). The effects of aging on layer 1 of primary visual cortex in the rhesus monkey. Cereb. Cortex , 11, 93–103. Peters, A., Rosene, D.L., Moss, M.B., Kemper, T.L., Abraham, C.R., Tigges, J. et al. (1996). Neurobiological bases of age-related cognitive decline in the rhesus monkey. J. Neuropathol. Exp. Neurol. , 55, 861–874. Pfutze, E.-M.S., Werner, and Schweinberger, S.R. (2002). Age-related slowing in face and name recognition: evidence from event-related brain potentials. Psychol. Aging , 17, 140–160. Pichora-Fuller, M.K. and Carson, A.J. (2001). Hearing health and the listening experiences of older communicators. In M.L.Hummert and J.F.Nussbaum (Eds.), Aging, Communication, and Health (43–74). Mahwah, NJ: Erlbaum. Pickles, J.O. (1988). An Introduction to the Physiology of Hearing , 2nd ed. San Diego, CA: Academic Press. Pratt, M.W., Boyes, C., Robins, S., and Manchester, J. (1989). Telling tales: aging, working memory, and the narrative cohesion of story retellings. Dev. Psychol. , 25, 628–635. Preusser, D.F., Williams, A.F., Ferguson, S.A., Ulmer, R.G., and Weinstein, H.B. (1998). Fatal crash risk for older drivers at intersections. Accident Anal. Prev. , 30, 151–159. Priester, J., Wegener, D., Petty, R., and Fabrigar, L. (1999). Examining the psychological process underlying the sleeper effect: the elaboration likelihood model explanation. Media Psychol. , 1, 27–48. Program, N.H. T.S. A.M. D.S. A.E. (1999). Safe mobility for older people: a notebook. Washington, D.C.: U.S. Department of Transportation.


274

Rabbitt, P., Lowe, C., and Shilling, V. (2001). Frontal tests and models for cognitive ageing. Eur. J. Cognitive Psychol. , 13, 5–28. Rabbitt, F.M. (1991). Mild hearing loss can cause apparent memory failures which increase with age and reduce with IQ. Acta Otolaryngologica , 476 (Suppl.), 167–176. Rahhal, T.A. and Hasher, L. (1998, April). It’s not all downhill: age differences in explicit memory performance can be eliminated by inteructions. Paper presented at the Cognitive Aging Conference, Atlanta, GA. Rankin, J.L. (2000). Age/cohort differences in judgments about criminal transgressions. J. Adult Dev. , 7, 127–135. Raz, N. (2000). Aging of the brain and its impact on cognitive performance: integration of structural and functional findings. In F.I.M.Craik and T.A.Salthouse (Eds.), The Handbook of Aging and Cognition (2nd ed., 1–90). Mahwah, NJ: Erlbaum. Reese, C.M., Cherry, K.E., and Norris, L.E. (1999). Practical memory concerns of older adults. J. Clin. Geropsychol. , 5, 231–244. Reger, M.A., Welsh, R.K., Watson, G.S., Cholerton, B., Baker, L.D., and Craft, S. (2004). The relationship between neuropsychological functioning and driving ability in dementia: a metaanalysis. Neuro-psychology , 18, 85–93. Reuter-Lorenz, P.A. (2000). Cognitive neuropsychology of the aging brain. In D.C.Park and N.Schwarz (Eds.), Cognitive Aging: a Primer (93–114). Philadelphia, PA: Psychology Press. Rodin, J. and Langer, E. (1980). Aging labels: the decline of control and the fall of self-esteem. J. Social Issues , 36, 12–29. Roediger, H.L. and McDermott, K.B. (1995). Creating false memories: remembering words not presented in lists. J. Exp. PsychoL.: Learning, Memory, Cognit ., 21, 803–814. Rogers, W.A. (2000). Attention and aging. In N.Schwarz and B.Knauper (Eds.), Cognition, Aging, and Self-Reports (57–73). Philadelphia, PA: Psychology Press. Rogers, W.A. and Fisk, A.D. (1996). Aging and Skilled Performance: Advances in Theory and Applications . Mahwah, NJ: Erlbaum. Rogers, W.A. and Fisk, A.D. (2000). Human factors, applied cognition, and aging. In F.I.M.Craik and T.A. Salthouse (Eds.), The Handbook of Aging and Cognition (2nd ed., 559–591). Mahwah, NJ: Erlbaum. Rubin, D.C. (2000). Autobiographical memory and aging. In D.C.Park and N.Schwarz (Eds.). Cognitive Aging: a Primer (131–149). Philadelphia, PA: Psychology Press. Rutledge, P.C., Hancock, R.A., and Walker, L. (1997). Effects of retention interval length on young and elderly adults’ memory for spatial information. Exp. Aging Res ., 23, 163–177. Ryan, E.B. (1992). Beliefs about memory across the life span. J. Gerontol.: Psychol. Sci. , 47, 41– 47. Salthouse, T.A. (1991). Theoretical Perspectives on Cognitive Aging . Hillsdale, NJ: Lawrence Erlbaum. Salthouse, T.A. (1993). Speed mediation of adult age differences in cognition. Dev. Psychol. , 29, 722–738. Salthouse, T.A. (1996). The processing-speed theory of adult age differences in cognition. Psychological Rev. , 103, 403–428. Salthouse, T.A. (2001). Attempted decomposition of age-related influences on two tests of reasoning. Psychol. Aging , 16, 251–263. Salthouse, T.A. (2004). What and when of cognitive aging. Curr. Directions in Psycholog. Sci. , 13, 140–144. Salthouse, T.A., Hancock, H.E., Meinz, E.J., and Hambrick, D.Z. (1996). Interrelations of age, visual acuity, and cognitive functioning. J. Gerontol: Psychological Sci. , 51B, 317–330. Sanfey, A.G. and Hastie, R. (2000). Judgment and decision making across the adult life span: a tutorial review of psychological research. In D.C.Park and N.Schwarz (Eds.), Cognitive Aging: a Primer (253–273). Philadelphia, PA: Psychology Press.


275

Santens, P., De Corte, T., Vingerhoets, G., Van Laere, K., Dierckx, R., and De Reuck, J. (2003). Regional cerebral blood flow and episodic memory in Parkinson’s disease: a single photon emission tomography study. Eur. Neurol. , 49, 238–242. Sataloff, J. and Vassallo, L. (1966). Hard-of-hearing senior citizens and their physician. Geriatrics , 21, 182–186. Scahill, R.I., Frost, C, Jenkins, R., Whitwell, J.L., Rossor, M.N., and Fox, N.C. (2003). A longitudinal study of brain volume changes in normal aging using serial registered magnetic resonance imaging. Arch. Neurol. , 60, 989–994. Schacter, D.L. (1996). Searching for Memory . New York: Basic Books. Schacter, D.L. (1999). The seven sins of memory: insights from psychology and cognitive neuroscience. Am. Psychologist , 54, 182–203. Schacter, D.L. (2001). The Seven Sins of Memory . New York: Houghton-Mifflin. Schacter, D.L., Kaszniak, A.W., Kihlstrom, J.F., and Valdiserri, M. (1991). The relation between source memory and aging. Psychol. Aging , 6, 559–568. Schacter, D.L., Norman, K.A., and Koutstall, W. (1998). The cognitive neuroscience of constructive memory. Annu. Rev. Psychol , 49, 289–318. Schaninger, C.S. and Sciglimpaglia, D. (1981). The influence of cognitive personality traits and demo-graphics on consumer information acquisition. J. Consumer Res. , 8, 208–216. Schieber, R and Goodspeed, C.H., IV. (1997). Nightime conspicuity of highway signs as a function of sign brightness, background somplexity and age of observer. In Proc. Hum. Factors Ergonomics Soc. 41st Annu. Meeting (1362–1366). Santa Monica, CA: Human Factors and Ergonomics Society. Schmader, T. and Johns, M. (2003). Converging evidence that stereotype threat reduces working memory capacity. J. Personality Social Psychol. , 85, 440–452. Schmitter-Edgecombe, M. and Simpson, A.L. (2001). Effects of age and intentionality on content memory and temporal memory for performed activities. Aging, Neuropsychol. Cognit. , 8, 81– 97. Schneider, A.L., Griffith, W.R., Sumi, D.H., and Burcart, J.M. (1987). Portland forward records check of crime victims. Washington, D.C.: U.S. Department of Justice. Schneider, B.A. (1997). Psychoacoustics and aging: implications for everyday listening. J. SpeechLanguage Pathol. Audiol , 21, 111–124. Schneider, B.A., Daneman, M., and Pichora-Fuller, M.K. (2002). Listening in aging adults: from discourse comprehension to psychoacoustics. Can. J. Exp. Psychol. , 56, 139–152. Schneider, B.A. and Pichor a-Fuller, M.K. (2000). Implications of perceptual deterioration for cognitive aging research. In F.I.M.Craik and T.A.Salthouse (Eds.), The Handbook of Aging and Cognition (155–219). Mahwah, NJ: Erlbaum. Schneider, J.A., Wilson, R.S., Cochran, E.J., Bienias, J.L., Arnold, S.E., Evans, D.A. et al. (2003). Relation of cerebral infarctions to dementia and cognitive function in older persons. Neurology , 60, 1082–1088. Schooler, J.W. and Engstler-Schooler, T.Y. (1990). Verbal overshadowing of visual memories: some things are better left unsaid. Cognitive Psychol , 22, 36–71. Schretlen, D.J., Pearlson, G.D., Anthony, J.C., and Yates, K.O. (2001). Determinants of Benton Facial Recognition Test performance in normal adults. Neuropsychology , 15, 405–410. Schwarz, N. and Knauper, B. (2000). Cognition, aging, and self-reports. In D.C.Park and N.Schwarz (Eds.), Cognitive Aging: a Primer (233–252). Philadelphia, PA: Psychology Press. Scialfa, C.T. (2002). The role of sensory factors in cognitive aging research. Can. J. Exp. Psychol , 56, 153–163. Scialfa, C.T., Garvey, F.M., Gish, K.W., Deering, L.M., Leibowitz, H.W., and Goebel, C.G. (1988). Relationships among measures of static and dynamic visual sensitivity. Hum. Factors , 30, 677–687.


276

Scialfa, C.T. and Joffe, K.M. (1997). Age differences in feature and conjunction search: implications for theories of visual search and generalized slowing. Aging, Neuropsychol., Cognit. , 4, 227–246. Searcy, J., Bartlett, J.S., and Memon, A. (1999). Age differences in accuracy and choosing in eyewitness identification and face recognition. Memory Cognit. , 27, 538–552. Searcy, J., Bartlett, J.S., and Memon, A. (2000). Influence of post-event narratives, line-up conditions and individual differences on false identification by young and older eyewitness. Legal Criminol. Psychol , 5, 219–235. Searcy, J.H., Bartlett, J.C., Memon, A., and Swanson, K. (2001). Aging and lineup performance at long retention intervals: effects of metamemory and context reinstatement. J. Appl Psychol , 86, 207–214. Sekuler, A.B., Bennett, P.J., and Mamelak, M. (2000). Effects of aging on useful field of view. Exp. Aging Res. , 26, 103–120. Sherman, J.W. and Bessenoff, G.R. (1999). Stereotypes as source-monitoring cues: on the interaction between episodic and semantic memory. Psychol. Sci. , 10, 106–110. Shimamura, A.P. and Jurica, P.J. (1994). Memory interference effects and aging: findings from a test of frontal lobe function. Neuropsychology , 8, 408–412. Shukitt-Hale, B., McEwen, J.J., Szprengiel, A., and Joseph, J.A. (2004). Effect of age on the radial arm water maze: a test of spatial learning and memory. Neurobiology of Aging , 25, 223–229. Simon, S.L., Walsh, D.A., Regnier, V.A., and Krauss, I.K. (1992). Spatial cognition and neighborhood use: the relationship in older adults. Psychol Aging , 7, 389–394. Simons, D.J. and Levin, D.T. (1998). Failure to detect changes to people during a real-world interaction. Psychonomic Bull. Rev. , 5, 644–649. Sivak, M., Olson, P.L., and Pastalan, L.A. (1981). Effects of driver’s age on nighttime legibility of highway signs. Hum. Factors , 23, 59–64. Skinner, B.F. (1983). Intellectual self-management in old age. Am. Psychologist , 38. Smith, A.D. and Winograd, E. (1978). Adult age differences in remembering faces. Dev. Psychol , 14, 443–444. Snowdon, D. (2001). Aging with Grace: What the Nun Study Teaches Us about Leading Longer, Healthier, and More Meaningful Lives . New York: Basic Books. Spaniol, J. and Bayen, U.J. (2002). When is schematic knowledge used in source monitoring? J. Exp. Psychol: Learning, Memory, Cognit. , 28, 631–651. Spear, P.D. (1993). Neural bases of visual deficits during aging. Vision Res. , 33, 2589–2609. Spencer, W.D. and Raz, N. (1995). Differential age effects on memory for content and context: a meta-analysis. Psychol Aging , 10, 527–539. Speranza, F., Moraglia, G., and Schneider, B.A. (2001). Binocular detections of masked patterns in young and old observers. Psychol Aging , 16, 281–292. Sporer, S.L. (2001). Recognizing faces of other ethnic groups. Psychol, Public Policy, Law , 7, 36– 97. Stamatiadis, N., Taylor, W.C., and McKelvey, F.S. (1991). Elderly drivers and intersection accidents. Transp. Q. , 45, 377–390. Staplin, L. and Lyles, R.W. (1991). Age differences in motion perception and specific traffic maneuver problems. Transp. Res. Rec. , 1325, 23–33. Steblay, N., Dysart, J., Fulero, S., and Lindsay, R.C.L. (2001). Eyewitness accuracy rates in sequential and simultaneous lineup presentations: a meta-analytic comparison. Law Hum. Behav. , 25, 459–473. Steele, C.M. (1997). A threat in the air: how stereotypes shape intellectual identity and performance. Am. Psychologist , 52, 613–629. Stein, R., Blanchard-Fields, E, and Hertzog, C. (2002). The effects of age-stereotype priming on the memory performance of older adults. Exp. Aging Res. , 28, 169–181. Stelmach, G.E. and Nahom, A. (1992). Cognitive-motor abilities of the elderly driver. Hum. Factors , 34, 53–65.


277

Stone, M.J. and Stone, L. (1997). Ageism: the quiet epidemic. Can. J. Public Health , 88, 293–294. Streufert, S., Pogash, R., Piasecki, M., and Post, G.M. (1990). Age and management team performance. Psychol. Aging , 5, 551–559. Strouse, A., Ashmead, D.H., Ohde, R.N., and Grantham, D.W. (1998). Temporal processing in the aging auditory system. J. Acoustical Soc. Am. , 104, 2385–2399. Sullivan, E.V., Rosenbloom, M., Serventi, K.L., and Pfefferbaum, A. (2004). Effects of age and sex on volumes of the thalamus, pons, and cortex. Neurobiol. Aging , 25, 185–192. Sullivan, S. and Ruffman, T. (2004). Social understanding: how does it fare with advancing age? Br. J. Psychol , 95, 1–18. Tisserand, D.J. and Jolles, J. (2003). On the involvement of prefrontal networks in cognitive aging. Cortex , 39, 1107–1128. Tran, D.B., Silverman, S.E., Zimmerman, K., and Feldon, S.E. (1998). Age-related deterioration of motion perception and detection. Graefes Arch. Clin. Exp. Ophthalmol , 236, 269–273. Tun, P.A., Wingfield, A., Rosen, M.J., and Blanchard, L. (1998). Response latencies for false memories: gist-based processes in normal aging. Psychol. Aging , 13, 230–241. U.S. Bureau of Justice Statistics. (2002). Victim characteristics: summary findings. Washington, D.C.: U.S. Government Printing Office. Underwood, J. and Pezdek, K. (1998). Memory suggestibility as an example of the sleeper effect. Psychonomic Bull. Rev. , 5, 449–453. Uttl, B. and Graf, P. (1993). Episodic spatial memory in adulthood. Psychol. Aging , 8, 257–273. Verhaeghen, P. (2002). Age differences in efficiency and effectiveness of encoding for visual search and memory search: a time-accuracy study. Aging, Neuropsychol Cognit ., 9, 114–126. Verhaeghen, P. and Meersman, L.D. (1998). Aging and the Stroop effect: a meta-analysis. Psychol Aging , 13, 120–126. Verhaeghen, P. and Salthouse, T.A. (1997). Meta-analyses of age-cognition relations in adulthood: estimates of linear and nonlinear age effects and structural models. Psychological Bull , 122, 231–249. Verhaeghen, P., Steitz, D.W., Sliwinski, M.J., and Cerella, J. (2003). Aging and dual-task performance: a meta-analysis. Psychol Aging , 18, 443–460. Villaume, W.A., Brown, M.H., and Darling, R. (1994). Presbycusis, cummunication, and older adults. In M.L.Hummert, J.M.Wiemann, and J.F.Nussbaum (Eds.), Interpersonal Communication in Older Adulthood: Interdisciplinary Theory and Research (83–106). Thousand Oaks, CA: Sage, von Hippel, W., Silver, L.A., and Lynch, M.E. (2000). Stereotyping against your will: the role of inhibitory ability in stereotyping and prejudice among the elderly. Personality Social Psychol Bull , 26, 523–532. Watson, D.G. and Humphreys, G.W. (1997). Visual marking: prioritizing selection for new objects by top-down attentional inhibition. Psychol Rev. , 104, 90–122. Watson, D.G. and Maylor, E.A. (2002). Aging and visual marking: selective deficits for moving stimuli. Psychol Aging , 17, 321–339. Wells, G.L. and Leippe, M.R. (1981). How do triers of fact infer the accuracy of eyewitness identifications? Using memory for peripheral detail can be misleading. J. Appl. Psychol , 66, 682–687. Wells, G.L., Small, M., Penrod, S., Malpass, R., Fulero, S.M., and Brimacombe, C.A.E. (1998). Eyewitness identification procedures: recommendations for lineups and photospreads. Law Hum. Behav. , 22, 603–647. Werner, J.S. (1998). Aging through the eyes of Monet. In W.G.K.Backhaus, R.Kliegl, and J.S.Werner (Eds.), Color Vision: Perspectives from Different Disciplines (3–41). New York: Walter de Gruyter. West, R. (2000). In defense of the frontal lobe hypothesis of cognitive aging. J. Int. Neuropsychol. Soc. , 6, 727–729.


278

West, R.L., Welch, D.C., and Knabb, P.D. (2002). Gender and aging: spatial self-efficacy and location recall. Basic Appl Social Psychol , 24, 71–80. Wheeler, S.C. and Petty, R.E. (2001). The effects of stereotype activation on behavior: a review of possible mechanisms. Psychol Bull , 127, 797–826. Willis, A. and Anderson, S.J. (2000). Effects of glaucoma and aging on photopic and scotopic motion perception. Invest. Ophthalmol. Visual Sci. , 41, 325–335. Willis, S.L. (1996). Everyday problem solving. In J.E.Birren and K.W.Schaie (Eds.), Handbook of the Psychology of Aging (4th ed., 287–307). San Diego, CA: Academic Press. Willott, J.E (1991). Aging and the Auditory System: Anatomy, Physiology, and Psychophysics . San Diego, CA: Singular. Wilson, B.A. (1995). Memory rehabilitation: compensating for memory problems. In R.A.Dixon and L. Backman (Eds.), Compensating for Psychological Deficits and Declines: Managing Losses and Promoting Gains (171–190). Mahwah, NJ: Erlbaum. Wilson, B.A. and Watson, P.C. (1996). A practical framework for understanding compensatory behavior in people with organic memory impairment. Memory , 4, 465–486. Wingfield, A. (2000). Speech perception and the comprehension of spoken language in adult aging. In D.C.Park and N.Schwarz (Eds.), Cognitive Aging: a Primer (175–195). Philadelphia, PA: Psychology Press. Wingfield, A. and Kahana, M.J. (2002). The dynamics of memory retrieval in older adulthood. Can. J. Exp. Psychol. , 56, 187–199. Wist, E.R., Schrauf, M., and Ehrenstein, W.H. (2000). Dynamic vision based on motion-contrast: changes with age in adults. Exp. Brain Res. , 134, 295–300. Wojciechowski, R., Trick, G.L., and Steinman, S.B. (1995). Topography of the age-related decline in motion sensitivity. Optometry Vision Sci. , 72, 67–74. Wright, D.B. and Stroud, J.N. (2002). Age differences in lineup identification accuracy: people are better with their own age. Law Hum. Behav. , 26, 641–654. Yarmey, A.D. (1984). Age as a factor in eyewitness memory. In G.L.Wells and E.F.Loftus (Eds.), Eyewitness Testimony: Psychological Perspectives (142–154). Cambridge, MA: Cambridge University Press. Yarmey, A.D. (1985). Accuracy and credibility of the elderly witness. Can. J. Aging , 3, 79–89. Yarmey, A.D. (1993). Adult age and gender differences in eyewitness recall in field settings. J. Appl. Social Psychol , 23, 1921–1932. Yarmey, A.D. (1995). Earwitness speaker identification. Psychol., Public Policy, Law , 1, 792–816. Yarmey, A.D. (1996). The elderly witness. In S.L.Sporer, R.S.Malpass, and G.Koehnken (Eds.), Psychological Issues in Eyewitness Identification (259–278). Mahwah, NJ: Erlbaum. Yarmey, A.D. and Jones, H. (1983). Is the psychology of eyewitness identification a matter of common sense? In S.Lloyd-Bostock and B.R.Clifford (Eds.), Evaluating Witness Evidence: Recent Psychological Research and New Perspectives (13–40). Chichester, England: John Wiley & Sons. Yarmey, A.D. and Kent, J. (1980). Eyewitness identification by elderly and young adults. Law Hum. Behav. , 4, 359–371. Yarmey, A.D., Jones, H.P., and Rashid, S. (1984). Eyewitness memory of elderly and young adults. In D. Mueller, D.Blackman, and A.Chapman (Eds.), Psychology and Law . New York: John Wiley & Sons. Yates, J.E and Patalano, A.L. (1999). Decision making and aging. In D.C.Park, R.W.Morrell, and K. Shifren (Eds.), Processing of Medical Information in Aging Patients (31–54). Mahwah, NJ: Erlbaum. Yonan, C.A. and Sommers, M.S. (2000). The effects of talker familiarity on spoken word identification in younger and older listeners. Psychol. Aging , 15, 88–99. Yoon, C., Hasher, L., Feinberg, E, Rahhal, T.A., and Winocur, G. (2000a). Cross-cultural differences in memory: the role of culture-based stereotypes about aging. Psychol. Aging , 15, 694–704.


279

Yoon, C., May, C.P., and Hasher, L. (2000b). Aging, circadian arousal patterns, and cognition. In D.C. Park and N.Schwarz (Eds.), Cognitive Aging: a Primer (151–171). Philadelphia, PA: Psychology Press. Zacks, R. and Hasher, L. (1997). Cognitive gerontology and attentional inhibition: a reply to Burke and McDowd. J. Gerontol: Psychol. Sci. , 52, 274–283. Zacks, R.T., Hasher, L., and Li, K.A.H. (2000). Human memory. In F.I.M.Craik and T.A.Salthouse (Eds.), Handbook of Aging and Cognition (2nd ed., 293–358). Mahwah, NJ: Erlbaum. Zwahr, M.D. (1999). Cognitive processes and medical decisions. In D.C.Park and R.W.Morrell (Eds.), Processing of Medical Information in Aging Patients: Cognitive and Human Factors Perspectives (pp. 55–68). Mahwah, NJ: Erlbaum. Zwahr, M.D., Park, D.C., and Shifren, K. (1999). Judgments about estrogen replacement therapy: the role of age, cognitive abilities, and beliefs. Psychology and Aging , 14, 179–191.

Further Information Birren, J.E. and Schaie, K.W. (1996). Handbook of the Psychology of Aging . San Diego: Academic Press. Craik, F.I.M. and Salthouse, T.A. (Eds.) (2000). The Handbook of Aging and Cognition . Mahwah, NJ: Erlbaum. (2nd ed.). Dixon, R.A. and Backman, L. (Eds.) (1995). Compensating for Psychological Deficits and Declines . Mahwah, NJ: Erlbaum. Fisk, A.D. and Rogers, W.A. (Eds.) (1997). Handbook of Human Factors and the Older Adult . San Diego, CA: Academic Press. Hess, T.M. and Blanchard-Fields, F. (Eds.) (1999). Social Cognition and Aging . San Diego, CA: Academic Press. Hummert, M.L. and Nussbaum, J.F. (Eds.) (2001). Aging, Communication and Health . Mahwah, NJ: Erlbaum. Mayr, U., Spieler, D.H., and Kliegl, R. (Eds.) (2001). Ageing and Executive Control . New York: Psychology Press. Morrell, R.W. (Ed.) (2002). Older Adults, Health Information, and the World Wide Web . Mahwah, NJ: Erlbaum. Naveh-Benjamin, M., Moscovitch, M., and Roediger, H.L., III (Eds.) (2001). Perspectives on Human Memory and Cognitive Aging . New York: Psychology Press. Nussbaum, J.E, Pecchioni, L.L., Robinson, J.D., and Thompson, T.L. (2000). Communication and Aging (2nd ed.). Mahwah, NJ: Erlbaum. Park, D.C., Morrell, R.W., and Shifren, K. (Eds.) (1999). Processing of Medical Information in Aging Patients . Mahwah, NJ: Erlbaum. Park, D.C. and Schwarz, N. (Eds.) (2000). Cognitive Aging: a Primer . New York: Psychology Press. Rogers, W.A. and Fisk, A.D. (Eds.) (2001). Human Factors Interventions for the Health Care of Older Adults . Mahwah, NJ: Erlbaum. Schacter, D.L. (2001). The Seven Sins of Memory . New York: Houghton-Mifflin. Schacter, D.L. (1996). Searching for Memory . New York: Basic Books. Snowdon, D. (2001). Aging with Grace: What the Nun Study Teaches Us about Leading Longer, Healthier, and More Meaningful Lives . New York: Basic Books.

12 Memory for Conversation on Trial Deborah Davis University of Nevada Markus Kemmelmeier University of Nevada William C.Follette University of Nevada 0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

12.1 Introduction Testimony regarding conversation is central to the vast majority of criminal and civil cases litigated in the courts. In some cases, testimony may concern relatively public group discussions such as those that might take place in a corporate boardroom. Witnesses to such group interactions must attempt to remember what was said, as well as which member of the group said it and, perhaps, to whom it was said. Some are relatively more private, such as the famous small group conversations among President Richard Nixon, John Dean, and others that became central to the Watergate hearings (Gold, 1974). Others are more private still, occurring between a rape victim and her attacker or between a doctor and patient. Across a variety of circumstances, testimony regarding the specific details of these conversations becomes central to the decisions of the jurors who decide the cases. When an employee sues for wrongful termination, discussions held in the corporate boardroom (or university committee meeting) can become central evidence of the reasons underlying the termination. Discussions between doctor and patient of issues such as the nature of the patient’s symptoms, possible diagnoses, risks and benefits of alternative treatments, or instructions for medication or other treatments become central to cases of malpractice—as do discussions between other professionals and their clients. Criminal defendants are convicted based in part on testimony from witnesses to threats, verbal fights, and other conversational indicators of hostile intentions or the conflictual status of the relationship between the defendant and victim. Such witnesses sometimes overhear the conversations of others and sometimes are central participants. However, in all cases, the quality of jury decisions based on this testimony depends upon the accuracy of witness memories. Like other witnesses, those reporting memories for conversation can voluntarily choose to lie. Perhaps more commonly, however, their memories are subject to the same


282

honest failures and distortions that plague witness memories for specific persons, locations, objects, and events. It is this latter source of inaccuracy that we will address in this chapter. Modern scientific theories of memory suggest that human memory systems operate in three general stages: (1) acquisition (or encoding), when information is first transferred into our memory system; (2) storage, when information is maintained in memory over a period of time; and (3) retrieval, when information is located and retrieved from storage. At each stage, memory may be compromised by failure or inaccuracy. Comprehensive reviews of principles of human memory have been written and we refer the reader to them for an introduction to the basics (Baddeley, 1999; Schacter, 2001; Spear and Riccio, 1994). Similarly, excellent treatments of eyewitness memory (Davis and Follette, 2001; Thompson et al., 1998) or the role of memory in the forensic interview (Eisen et al., 2002) are also available. Consequently, this chapter’s review focuses on aspects of memory that are particularly important to, or even peculiar to, conversational memory. Furthermore, we focus our discussion on memory for conversational features that are commonly the subject of testimony in trial. First and foremost of these are, of course, who said something and what was said. However, in addition to the identity of the speaker and other participants, and the specific content of a conversation, witnesses must often report on such contextual features as where and when the conversation took place, who else might have been present, in which of many conversations a particular exchange took place, and so on. We will identify the many issues witnesses to conversations are asked to address and review relevant research when it is available. 12.2 Failures of Encoding Information must first be successfully encoded in order to be remembered. Three primary factors may cause memory to fail at the stage of encoding: (1) failure to perceive (or in this instance to hear) and to see relevant face and body language, (2) failure to devote sufficient attention to the target information, and (3) failure to understand it correctly. Perception Accurate understanding of conversation requires accurate speech perception, as well as accurate perception of facial expressions and body language that may affect interpretation of what is said. A person may fail to encode the nature of a conversational contribution accurately due to personal sensory impairments in hearing or vision (see Davis and Loftus, Chapter 11, this volume, for review) or to environmental interference through such factors as background noise, distance, or blocked vision. Visual cues, such as gestures or lip movements, can compensate for an impaired auditory channel and make message transmission more resilient to the presence of background noise and other auditory disturbances (Argyle and Graham, 1975; Rogers, 1978). This is particularly important for the hearing impaired, who tend to compensate by lip reading (Davis and Loftus, this volume; Nussbaum and Coupland, 2004; Nussbaum et al., 2000; Villaume et al., 1994).

Memory for conversation on trial

283

However, even for unimpaired listeners, nonverbal cues such as gestures, facial expressions, or body movement and posture help the hearer to understand what the speaker is trying to convey and how it is to be understood. Thus, failure to see the speaker is likely to undermine the adequacy of encoding utterances such as those that, for example, use sarcasm, irony, and similar rhetorical means that often come with nonverbal signals indicating that the utterance is not to be interpreted literally. Finally, multichannel perception is important for evaluation of the emotional underpinnings of the message and its veracity (see DePaulo et al., 2003). Interpretations of conversational contributions based on the availability of, or attention to, a limited number of channels may be completely erroneous. Therefore, when an attempt is made to evaluate the accuracy of memory for conversation, it is of crucial importance to consider the sensory abilities of the witnesses, as well as the environmental challenges to vision and hearing present at the time of the conversation (see Davis and Loftus, Chapter 11, this volume; Davis and Friedman, in press, for reviews). Attention The success with which information is encoded is determined by the amount of attention devoted to it, as well as the depth with which it is processed. The more that one attends to something and the more that one thinks about it while attending to it, the more likely it is to be remembered later (see reviews in Davis and Follette, 2001; Schacter, 1999, 2001). Conversational involvement, for example, enhances overall memory for ideas mentioned in the interaction, presumably because it increases attention (Cegala, 1984). Thus, to understand the conditions under which a particular conversation, part of a conversation, or particular person participating in the conversation will be best remembered, one must first consider what determines the amount and direction of attention paid by the observer to the conversation and its participants. Davis and Follette (2001) reviewed in detail the way in which attention is determined by personal characteristics and states, and contextual features of a target event. First, attention is vulnerable, so it may be impaired by internal or external distractions. The person’s internal processing resources may be diminished by such factors as illness, fatigue, old age, intoxication, pressing personal concerns, anxiety, and so on, which would limit attention devoted to external events of most kinds, including conversation. External distractions such as noise, complexity in the environment, interruptions, etc. can also compromise attention to conversation partners and their utterances. Second, because attention is selective, greater attention is devoted to aspects of the environment or of a conversation that (see the review in Davis and Follette, 2001) • Are salient (i.e., stand out and draw attention) • Are threatening (including physical and emotional threats) • Are distinctive, unexpected, or unique • Are interesting • Are relevant to personal interests, goals, or current concerns • Are relevant to activated schemas, stereotypes, or expectations • Elicit powerful emotions


284

Furthermore, attention tends to go toward what is deemed the essential “core” of the event—or to what is deemed the central point of a conversation (see the review in Davis and Friedman, in press). Attention may be selectively devoted to topics and to people. Therefore, people deemed most important, interesting, distinctive, etc. tend to draw more attention than others. Among the interpersonal features most extensively investigated is status. Persons of high status are typically relatively novel and relatively important. They are novel because access to higher status persons is often restricted, and thus interaction occurs less frequently; they are more important because higher status persons are often in possession of desired resources or in control of one’s outcomes. Typically, lower status individuals tend to be highly vigilant with regard to the behavior and utterances of higher status individuals, in particular when their own outcomes are dependent on the higher status persons (Berscheid et al., 1976; Fiske and Depret, 1996; cf. Erber and Fiske, 1984). Attention is also affected by expectations associated with conversational norms. One such expectation, for example, derives from the conversational maxim “be relevant” (Grice, 1975), which dictates that conversational contributions should be relevant to the previous content of others’ remarks and to the general content and context of the interaction. Davis and Holtgraves (1984) demonstrated that such an expectation of relevance can result in poor memory for irrelevant (or tangential) contributions. The authors had participants read a supposed debate between two Nevada politicians. Each was asked four questions regarding the deployment of an “MX” missile system in Nevada. The replies of one politician were directly relevant to the questions asked, whereas those of the other were only tangentially relevant. Two sets of questions were used so that the first politician’s answers were directly relevant to the first set and tangentially relevant to the second, whereas the reverse was true for the second politician. Memory (recognition and recall) was clearly impaired for the tangentially relevant replies. In fact, when asked to recall each reply verbatim, subjects often simply wrote “did not answer the question” for tangential replies. Presumably, schematic processing directing attention to material relevant to the question led to superficial processing of the irrelevant content and thus poor memory. Generally, violations of social norms or strongly held expectations are attention grabbing, and people remember such instances better, primarily because they are trying to resolve the discrepancy between expectation and reality (see Graesser et al., 1979). For example, one of the implicit expectations in many conversations is that participants follow norms of politeness and modesty, i.e., that they do not engage in unfavorable comments about each other, inappropriate sexual innuendo, or excessive self-praise. To the extent that certain statements violate these norms, their content becomes more memorable than the remainder of the conversation (Pezdek and Prull, 1993; Wyer et al., 1994). Role-based expectations also guide social perception and memory. Status differences, often associated with social roles, give rise to expectations with regard to how high- and low-status individuals will behave in conversation. Typically, lower status individuals are more polite to higher status individuals, who often command the resources upon which the other is dependent (Lee, 1993). Consistent with the idea that social status induces politeness-related expectations, Kemper and Thissen (1981) found that impolite requests by low-status speakers (which do not fit our expectations) are more likely to be


285

remembered, whereas such requests by high-status speakers are more likely to be forgotten or distorted. Conversely, polite requests by high-status speakers were more memorable than the same requests by low-status speakers (but see Holtgraves, 1997, experiments 1 and 2, for more ambiguous results). Similarly, when speakers state a request in ways that are not appropriate for the specific context, the wording of the request is more likely to be remembered compared to when the request is made in an appropriate way (Gibbs, 1987). Whereas discrepancy between expectation and reality is generally attention grabbing, in other instances expectations may actually facilitate memory for information that is consistent with them. This is primarily true when a person possesses expectations concerning the organization or structure of a particular conversational event. For instance, Gamst (1982; study 2) tested the notion that people have expectations with regard to the structure of a dialogue and have better memory for conversations that conform to a generic script. When participants heard a dialogue that did not conform to the script, their memories were poorer compared to those for a dialogue that conformed to their expectations. The Role of Anxiety and Emotion Many of the circumstances giving rise to litigation involve intense emotion among participants and witnesses alike. Thus, it is of interest to understand how anxiety and emotion might affect memory for conversation. Christianson and his colleagues have suggested a two-stage process by which stress or arousal affects memory (Christianson and Safer, 1996). First, in the “preattentive” stage, emotion-eliciting stimuli, such as blood or personal threat, trigger an orienting response drawing attention to the emotion-eliciting stimuli. In the second stage, active attentional mechanisms engage elaborative encoding focused on the emotional material. This selective attention and elaboration limit processing capacity for peripheral information not central to the emotional aspects of the event. In cases of very strong emotion, the person may become preoccupied by intrusive thoughts regarding the threatening event, further narrowing the focus of attention/processing. Safer and colleagues (1993) refer to the outcome of the narrowed attention and heightened psychological focus on the source of the emotional arousal as “tunnel memory.” Events witnessed under this narrow processing mode will tend to promote better memory for central information—that is, the details of the emotion-provoking part of the event. In contrast, it will tend to inhibit processing of and memory for peripheral details, or details that are irrelevant or spatially peripheral to the core source of arousal (Easterbrook, 1959; see reviews in Brown, 2003; Christianson, 1992a, b; Christianson and Safer, 1996; Heuer and Reisberg, 1992). This dynamic of selective attention and selective recall is powerfully illustrated with regard to the so-called “weapon focus effect” (Kramer et al., 1990). To the extent that a weapon is present in the situation, individuals tend to focus their attention on this potentially threatening object, but are less attentive and have a poorer memory of other characteristics of the situation. Beyond the general instance of threat, emotional arousal has often been linked to decreased memory for conversation (Goss et al., 1985; Sillars et al., 1990; Singer, 1969; Stafford and Daly, 1984). However, it appears that in these studies emotion was


286

peripheral with regard to the conversation, so research participants did have a poor memory for aspects of the conversation that did not produce their emotional state. Arousal often facilitates the formation of detailed and personally meaningful event memories. That is, the person may be particularly likely to retain a memory (though not necessarily accurate) of the event and peripheral details associated with it. Indeed, MacWhinney and colleagues (1982) showed that episodes causing physiological arousal were more likely to be remembered than episodes that were less arousing (see also Bools et al., 2001). It appears that emotional arousal generally promotes long-term retention of the emotion-arousing event or message (Parkin et al., 1982). This may also explain why jokes and other emotion-eliciting utterances tend to be retained better than neutral utterances (Schmidt, 1994; cf. Keenan et al., 1977). However, as Brown and Kulik (1977) and others have shown with their research on “flashbulb memories,” these very vivid memories can be quite inaccurate. In particular, although the core features of the events can be well retained, memories for peripheral detail (such as where or when it happened) are often quite mistaken. Notwithstanding the potential of emotion to facilitate memory, it may also be expected to impair memory in some circumstances. Anxiety and other unpleasant emotions, for example, maybe responsible for the well-known failure of patients to remember the content of their conversations with physicians (DiMatteo, 1985; Ley, 1966, 1979, 1986; Ley and Spelman, 1967). During aspects of physical examinations, for example, patients may invoke distracting strategies to avoid thinking about the unpleasant aspects of the examination. Although these strategies may have the intended benefit, they are likely to prevent patients from paying attention to aspects of their environment. Thus, it would not be wise to present important information to patients during aversive components of examinations. That is not to say that the physician cannot do and say things to reduce anxiety such as explaining the purpose of the examination, what to expect in terms of physical sensations, and so forth. Engaging in conversation for the purpose of relaxing the patient is also warranted. However, the doctor should not assume that those are optimal times to try to impart important information to the patient, especially information upon which the patient needs to act at a later time. Furthermore, anxiety is known to interfere with successful communication and memory. Physicians and patients are under special stress during discussions of serious or life-threatening diseases. However, physicians appear to have anticipatory anxiety that peaks early in interactions with patients (though it may persist for some time afterward) and patients’ anxiety increases later in the consultation and even after an initial consultation about serious news has concluded (Ptacek and Eberhardt, 1996). Physician anxiety may interfere with clear expression, attention to content, and noticing cues of confusion that may be presented by the patient as the conversation proceeds and the patient becomes more anxious. Summary. The effects of emotion on memory for events have been widely studied within psychology; however, little research has addressed the relationship of emotion to memory for conversation. Research that has addressed these issues has been correlational in nature and is open to alternative explanations. This dearth of research is particularly unfortunate because emotions are high during incidents such as sexual harassment or rape and fights that degenerate into violence, and in a variety of stressful professional interactions—all situations commonly giving rise to litigation of some form.


287

Interpretation Difficulty of Perception The presence of noise, distraction, personal sensory impairments, or other factors that cause difficulty in simple perception of words has the additional effect of compromising comprehension. That is, even though individuals may understand all the words spoken, they are less likely to understand the overall meaning, draw appropriate inferences from utterances, and integrate the new information with their existing knowledge. The result of this is frequently diminished comprehension and retention (Schneider et al., 2000). Those who have tried to learn a foreign language may have experienced the analogous effect when attempting to process rapid speech in the new language: one must try so hard to understand the words that the meaning of the entire sentence is lost. This problem often plagues older listeners, partic ularly in the presence of background noise (see Davis and Loftus, Chapter 11, this volume). Meaning Depends upon Context Language comprehension inherently depends on context. Initial understanding of conversational utterances requires integration of verbal and nonverbal cues, as well as background, situational, and relationship contexts (Argyle and Graham, 1975; Cohen, 1977; Rogers, 1978). For example, the English language is full of words that carry different meanings dependent on the context in which they are used (“nut” as fruit vs. “nut” as counterpart of a screw vs. “nut” as deranged individual). Similarly, phrases and utterances need to be considered within the social and physical context in which they are made in order for a hearer to infer a speaker’s intention accurately (Sperber and Wilson, 1995). For adequate understanding to occur, hearers and speakers must share a common understanding of the exact context. Speakers routinely make assumptions about the kind of information shared between speaker and hearer (Clark and Haviland, 1977). These assumptions are often warranted, perhaps because interactants are in the same room, have had similar experiences, or have communicated before about a particular subject. From this perspective, a speaker’s utterances carry with them assumptions about what he thinks a hearer knows already. It is easy to see that in cases in which a speaker’s assumptions are incorrect, misunderstandings will occur. Frequently, these misunderstandings are detected and eliminated. However, in other cases, interactants may not notice that they are making different assumptions. To the extent that an utterance is interpretable to them, two people can agree on what was said, yet encode very different interpretations (cf. Fletcher, 1994; see Davis and Friedman, in press, for review). Features of Speech That Create Difficulty in Understanding As we are all aware, some speakers are much more difficult to understand than others. Of course, failure initially to understand an utterance correctly will cause memory to fail as well. Thus, it is important to consider the aspects of speech that affect initial understanding.


288

Generally, linguistically simple statements are understood more easily. Long sentences that include numerous clauses and subordinate clauses require more effort and put great processing demands on the individual. In order to understand such a sentence, a hearer must hold all information extracted from different clauses in working memory long enough to be able to integrate all parts and extract an overall meaning. Thus, listeners— especially older listeners, distracted listeners, and those with otherwise impaired processing capacity—are particularly likely to fail to comprehend complex utterances (see Davis and Loftus, Chapter 11, this volume). Hearers generally find it easy to understand statements that involve concepts, vocabulary, and information with which they are familiar but struggle if statements take an unfamiliar form or convey new content. This is a common reason why conversations between professionals and their clients often fail. A specific problem in the medical realm is the use of a sophisticated and highly specialized vocabulary that is inaccessible to the layperson. The physician has a special responsibility to communicate in a language that is comprehensible to the patient. If patients have a particularly acquiescent response style, it may be difficult for the doctor to discern that the patient does not understand what he has been told. Under these circumstances, it is useful for the doctor to ask the patient to repeat what he was just told. The physician then must ascertain that the patient actually understands the concepts and does not merely parrot what was said without really understand its meaning. In general, it is desirable if patients para-phrase the information using language meaningful to them rather than incorporating ambiguous jargon. High-stakes medical conversations are likely to be about complex topics about which the physician uses terms familiar to the medical community but less so to the lay community. For instance, one study indicated that 73% of women informed about breast cancer did not understand the meaning of the word “median” as in median survival time (Lobb et al., 1999). If the patient does not understand such a term during a high-stress interaction, it is easy to see how physicians could correctly claim they informed patients of the likely survival time of a particular treatment option but the patient could claim that no such discussion occurred. Ambiguity about the meaning of important words in medical decision making and communication is not limited to statistical terms. Considering that many documents intended for public consumption are written at the sixth-grade level, it is easy to see how terms like “prognosis” and “morbidity” can be misunderstood, not encoded, and therefore not recalled. Even terms that fall within everyday language become salient and misunderstood in medical interactions. Physicians and patients use terms relating to frequency (frequently, rarely, likely, probably, often, etc.) quite differently (Girgis and Sanson-Fisher, 1995). This means, of course, that a complication experienced following a conversation with a physician could take a patient very much by surprise if he and the doctor had different subjective understanding of what was meant by “occasionally.” Attempts have been made to quantify the probabilities implied by adjectives used in the medical literature (Kong et al., 1986), but ambiguities still remain. There is no single best way to be sure that patients and physicians mean or understand words or probability estimates of risk, treatment options, or side effects the same way (Mazur and Hickam, 1991; Nakao and Axelrod, 1983; Ohnishi et al., 2002; Robertson, 1983; Woloshin et al., 1994). What is clear is that professionals need to engage in extra effort to ensure understanding—as well as to help patients remember vital information and instructions.


289

In addition to knowing how to disclose high-stakes information, it is important for physicians to anticipate what patients want to know. Because of the high level of affect likely to be present for the patients, they may not be as good at communicating their desires for information. In a study of cancer patients’ experiences at receiving news of a cancer diagnosis, patients were less interested in the diagnosis than the prognosis. Although 14% of patients felt diagnosis was the most important part of cancer disclosure, more (18%) felt treatment options were most important and fully 52% wanted prognostic information. More than half the patients wanted to discuss life expectancy, though only 27% of physicians actually did so. Certainly, recall of these conversations would more likely be reported as unsatisfying, and the possibility of the patient inserting his understanding of what the prognosis is (even if it is incorrect) would be higher than if the physician supplied the information initially. Schematic Processing Schemas refer to units of knowledge that people have about recurring objects, situations, social categories, persons, or topics. Whether ordering a meal at a restaurant or going to the doctor, people have a clear conception of the kinds of people involved and the way in which these interactions typically proceed. This knowledge not only gives rise to specific expectations but also helps people to understand the events and people that they encounter. Ambiguous events are usually interpreted so that they are consistent with the schema for a particular situation (Bransford and Johnson, 1972). Similarly, schemas help to fill in the gaps (Bower et al., 1979) in what is directly perceived with schema-based assumptions about what is also likely to be true (see the review by Davis and Loftus, in press). People often compensate for lost words (missed or not clearly understood), for example, by filling in words they expect would typically fill such a spot. This is particularly likely to occur when the person has very strong expectations, when he is distracted and not processing carefully, or when his hearing is poor. This is particularly characteristic of elderly listeners, who rely on expectations for content as well as knowledge of syntax and semantics to fill in gaps (Nussbaum et al., 2000). Specific schemas for other persons or the circumstances of an interaction also affect expectations, which in turn determine which elements of an interaction are most likely to be noticed and how they will be interpreted when ambiguity exists. Consider a new mother whose 10-month-old child is just beginning to walk and is heading toward a toy he clearly does not see. The mother, noting the child is about to trip, grabs the child by the hand just as the child falls. The child cries out in pain and is unable to move his arm. Upon taking the child to the pediatrician, the new mother worries that the physician will think she abused the child, although all she did was stop him from falling. During the conversation with the physician, she is tense, nervous, and defensive when the doctor asks her if the child is meeting his developmental milestones, a routine question. The doctor also tells the mother that one must be very careful when placing pressure on the arm of children under the age of six because they can become easily dislocated. Though the physician never overtly questions about child abuse, the new mother feels a dislike for the doctor and ultimately tells her friends that the doctor is suspicious and condescending during interactions with her. From the doctor’s point of


290

view, a mild dislocation (radial head subluxation) of the elbow is common. He was never suspicious of child abuse and simply provided information to a new mother about an action and condition commonly referred to as “nursemaid elbow.” The mother’s expectation created a confirmatory bias in her assessment of the doctor’s suspicion. In recalling the conversation to others, she then highlights the “evidence” that the doctor was suspicious. In this case of nursemaid elbow, one can just as easily see how a physician with an expectation of abuse based on a stereotype or pre-existing suspicion could read the nervous new mother’s body language and verbal defensiveness as evidence for the abuse hypothesis. Research supports the possibility that the physician could selectively recall evidence from the conversation to support his hypothesis that there was reason for concern about possible abuse (Sande et al., 1986). Personal Determinants of Understanding Personal difficulties may exist that compromise understanding in conversation, many of which have been studied in the context of medical interactions. For example, beyond patients suffering age-related cognitive impairments or dementias, people who are acutely or subclinically cognitively impaired tend to encode or retrieve information improperly. Acute impairment may be due to such factors as mild disorientation secondary to illness or even medication toxicity. Medication-induced acute impairments may come from drug interactions or may have developed because the patient is taking drugs improperly. The patient may also inadequately metabolize the drugs, resulting in toxicities due to higher than anticipated blood levels of medications. In both instances, patients often try to cover up early signs of cognitive decline by being overly acquiescent. It is easy to understand how a doctor could report informing such a patient about the side effects of a drug while the patient denies being told of any such effects. 12.3 Failures of Retention Memory of the Fact of the Conversation Perhaps the most fundamental aspect of memory for conversation is memory for whether the conversation took place at all. Like memory for all information, memory for the very fact of a conversation can fail. Unfortunately, this form of failure of memory for conversation plays an integral role in a common crime against the elderly known as the “where’s the check?” scam (Schacter, 2001). Essentially, perpetrators of this scam convince elderly targets that they have already agreed in some way to give the perpetrator a check. In some cases, the perpetrator may simply claim that the elderly target promised to send the check in an earlier conversation, say it has not arrived as agreed, and remind the person to send it. In others, the perpetrator may say he has received the agreed-upon check, but that it was written for more than necessary. He then asks the elderly target to send a replacement check for the lesser amount. Perpetrators of such scams often pretest their elderly targets to assess their memory by calling in advance to collect personal information. When the perpetrator later calls back


291

to initiate the scam, he first determines whether the older targets remember the previous conversation and thus whether they would be likely to remember other events. If not, the perpetrator proceeds to make his false claim. Many older people expect their memories to be poor and would not be surprised to hear that they had forgotten a phone call or conversation or the details of past financial agreements or transactions. Thus, they are easily convinced that they had previously written (and then forgotten writing) a check or that they had agreed to donate to a particular cause. Memory of What Is Said How Much Is Remembered? Most of what is said in conversation is lost. Stafford and colleagues (1987) found that immediately after a mundane conversation, participants could recall only about 10%. After 1 month, this number dropped to 4%. What we retain from a conversation is primarily the gist of what was perceived as its core. Such poor retention of what is said contributes to poor outcomes of professional interactions that are later litigated in the courts—prominently, those between doctors and patients. It has long been noted, for example, that patients are unable to recall about half of what they have been told by their physicians (DiMatteo, 1985; Ley, 1966, 1979, 1986; Ley and Spelman, 1967). Gender effects have also been noted to affect recall of medical information. One should be cautious about how changing cultural influences could alter these findings over time; however, it appears that, especially for women, medical information presented by a female physician with high degrees of expressiveness produces the most recall of information (Bush, 1985). Similar effects were present for males, but to a lesser degree. High degrees of nonverbal expressiveness in opposite sex dyads may actually interfere with recall for both sexes. Memory of Exact Content vs. Inferred Meaning Often the exact wording of a statement is crucially important in forensic contexts. For example, the statement “I’m going to kill the son of a bitch!” may constitute a clear threat, whereas “The son of a bitch doesn’t deserve to live!” might be taken as a statement of opinion. The two, however, would often be understood as equivalent and consequently remembered equivalently later—particularly if the person in question ended up dead. Memory for exact wording is often poor. That is, only minutes after an utterance is made, a conversation partner may be unable to recall its exact wording. Hearers routinely extract the basic idea underlying an utterance and forget the linguistic surface form (Bock and Brewer, 1974; Brewer and Hay, 1984; Graesser and Mandler, 1975; see Fletcher, 1994, for a review). They process “between the lines,” remembering not precisely what was said but, rather, what was pragmatically implied or what appears to make most sense in a given situation (Harris and Monaco, 1978; Hilton, 1995). For practical purposes, specific wording is often irrelevant because the same prepositional content (or meaning) can be expressed in a variety of different ways and because hearers primarily care about the message that is conveyed. Language


292

comprehension and, ultimately, memory for utterances are geared toward inferences hearers draw from speakers’ utterances, not the utterances alone. As is readily apparent, this may cause problems in contexts in which the specific wording used has great legal relevance or when the hearer has inferred and remembered the wrong meaning (see Davis and Friedman, in press, for review). This meaning-driven or gist-based memory applies not only to individual utterances, but also to the entire episode of a conversation. People may not recall any details of what was said or even who said what, but rather may remember the gist of what the conversation was about. Specifically, individuals may recall the general kinds of opinions or beliefs that the participants in a conversation expressed but no longer remember any specific statements. This type of memory is much more abstract and primarily a reflection of how a person interpreted and experienced a situation. For example, a witness may report hearing a “fight” between two persons, one of whom is later found dead, but the witness may be unable to report any of the content. This interpretation could have resulted from other cues such as tone of voice, volume, body language, and so on, perhaps in the absence of ever having heard the actual content. More dangerously, the label “fight” may be the result of “reconstructive” memory processes (or “hindsight” biases) resulting from the witness’s post hoc discovery of the death. Unfortunately, listeners may later confidently report that they remember exactly what the speaker said when in fact they remember only the gist of their interpretation of what was said (for example, see Neisser, 1981). This situation arose for a local instructor, whose students accused her of calling them stupid in class. As is characteristic of any qualified college instructor, she was very careful to avoid any insults or inappropriate language in her interactions with students. One day, she dismissed the class early because students had quite obviously failed to do the readings and were unprepared for the scheduled in-class discussion. The instructor, who had expressed her dissatisfaction with this situation before dismissing the class, later was astonished to find out that students reported she had called them “stupid.” In hindsight, however, one can understand this disagreement in terms of interpretation-based or gist-based processing on the part of the students. They interpreted the instructor’s behavior to mean she thought they were stupid and, later, confidently “remembered” that stupid was exactly what she said. This example illustrates the principle that subsequent memory for an utterance can only be as accurate as the inference initially drawn by the hearer. It is easy to see how the combination of loss of linguistic form combined with inference drawing or interpretation can create a number of problems so that the hearer does not interpret an utterance in line with what the speaker intended to say or remembers primarily what the speaker only implied but did not say. The prevalence and importance of such inferences was illustrated by Fulero and Finkel (1991), who found that juror inferences from the actual content of expert testimony strongly affected verdicts. The authors presented three versions of a trial in which expert testimony regarding insanity was presented. In one version, the expert presented only diagnostic information that the patient suffered from delusional paranoid disorder. In the second version, the expert testified regarding the diagnosis and, in addition, testified that because of the delusions associated with the disorder, the patient would be unable to consider the consequences of his behavior or to appreciate the wrongfulness of his acts. Finally, for the third version, the expert testified to the preceding facts and further stated


293

the conclusion that the defendant was insane at the time of the crime. No differences in verdict were obtained among the three levels of testimony, in part because jurors tended to infer the final determination of insanity in all conditions and/ or to remember falsely that the expert had stated that conclusion even when he had not. The importance of conversation-based inferences has also been extensively discussed in the literature on false confessions. Kassin (1997) noted that judges typically exclude confessions elicited through explicit threats and promises, but they commonly admit confessions elicited through implied threats and promises. Kassin and McNall (1991) directly tested whether certain interrogation practices were interpreted as threats or promises of leniency. The interrogation techniques they focused on were “maximization” (in which the interrogator overstates the seriousness of the offense and the magnitude of charges, and makes false or exaggerated claims regarding the strength of evidence against the defendant) and “minimization” (in which the interrogator offers sympathy or excuses for the accused, perhaps blaming the victim or someone else, and thus minimizes the seriousness of the crime). Participants read one of five versions of an interrogation. The interrogator (1) explicitly promised leniency if the defendant confessed, (2) explicitly threatened harshness if he did not confess, (3) used minimization techniques, (4) used maximization techniques, or (5) did none of these. Participants then estimated the likely sentence to be imposed on the suspect. Results indicated that (1) minimization and actual promises of leniency led to equivalently lenient expectations of punishment and (2) maximization and explicit threats of harshness led to equivalently harsh expectations of punishment. That is, observers inferred the promises or threats that were never explicitly made. Thus, Kassin (1997) concluded that common interrogation tactics essentially violate the intent of the law by using pragmatic implication to accomplish what they are forbidden to do explicitly. Which Contributions Are Remembered? Which conversational contributions will be remembered depends on a complex web of factors. To illustrate this point, Hirst and Gluck (1999) conducted a study comparing the testimony of John Dean before the Senate subcommittee on the Watergate affair in 1973 with transcripts of White House recordings of 1972 conversations among Richard Nixon, H.R.Haldeman, and John Dean. The tapes of the conversations were an unpublished transcription prepared by the impeachment inquiry staff for the House Judiciary Committee at the time of the Watergate Hearings (National Archives, 1996), which contained a more complete transcription than the previously published White House Transcripts (Gold, 1974). From these, the authors selected the transcription of the September 15 meeting of Nixon, Haldeman, and Dean involving a review of events related to Watergate. Each person’s contributions to the original conversation were coded as narration, facilitating remarks (mentoring), or monitoring (evaluating and correcting). The authors found a number of interesting relationships between conversational role and the content of Dean’s statement to the Watergate subcommittee. Although there are many potential explanations of the results, the authors suggest that the observed patterns may reflect the influences of schematic-processing and conversational roles (Hirst and Gluck, 1999).


294

Perhaps the largest observed difference was in memory for Dean’s contributions versus those of others. Whereas Dean offered 58% of the narrations in the meeting, 71% of the narratives offered in his statement to the subcommittee were his own. The authors attributed this difference to the greater relevance of Dean’s contributions to his preexisting knowledge and beliefs regarding the Watergate events, along with personal motivations such as a need for self-importance. The authors also predicted that because Dean’s schematic organization would dictate what he chose to tell the others and what he attempted to elicit from them, Dean would be more likely later to recall and report these schema-consistent narratives than those reflected in answers to others’ less schema-relevant questions and narratives. Indeed, Dean’s memory for conversational contributions due to his own initiatives was superior to that resulting from the initiatives of others. The narratives elicited from Dean by Haldeman and Nixon were more likely to remain unstated than to appear in Dean’s statement. On the other hand, the narratives elicited by Dean from Haldeman and Nixon were more likely to make it into the statement than remain unstated. Hirst and Gluck (1999) concluded that narrative responses to others’ questions make it into subsequent recollections only if they can be assimilated into existing schemata. Presumably, Dean thought his responses to questions from Nixon and Haldeman were relatively unimportant and thus failed to assimilate them into his preexisting schema. In contrast, their answers to Dean’s chosen questions, which he undoubtedly considered relevant, were easily assimilated: he probably knew the details about the events following Watergate extremely well. Thus, nothing Nixon or Haldeman could do—narrating the events or asking Dean questions—could alter Dean’s perspective on the events. One might have expected that power relationships would have an effect on memory such that Dean would be more likely to remember something said by a powerful person than a less powerful individual. In contrast, Dean’s memory appeared to be affected by a combination of perceived expertise and authority. That is, he correctly believed he possessed the greatest expertise about the events that had transpired since the Watergate break-in. However, Dean realized that Nixon had the authority to decide future actions concerning Watergate and that Nixon’s beliefs concerning how the Watergate situation would unfold in the future would affect those decisions. From this perspective, it is not surprising that Dean remembered his own narratives better than those offered by Nixon, but that he remembered Nixon’s prognostications and plans for the future better than his own. Thus, a complex interplay between one’s goals, existing knowledge structures, schematic perceptions of the issues and persons involved, the relationship between parties, and much more determines the focus of attention within a given conversational interaction and therefore determines what is later remembered. Beyond the factors described here, conversational memory for specific utterances is highly affected by the function that these utterances play in a conversation. Studies show that utterances high in interactional content are highly memorable in the sense that individuals are highly successful in recognizing the exact surface form of the utterances (Keenan et al., 1977; MacWhinney et al., 1982). In this research, interactional content was defined as pragmatic information that conveyed a great deal about the communicative situation as well as the speaker and his “intentions, beliefs, and his relations to the listener” (Keenan et al., 1977, p. 550). Concretely, this category includes,


295

but is not limited to, statements of personal opinions, evaluations, jokes, profanity, and sexually suggestive language (see also Bates et al., 1978; Pezdek and Prull, 1993). Consistent with our review of the role of attention for conversational memory, Kintsch and Bates (1977) hold that the specific wording of utterances is encoded and remembered to the extent that it draws attention to itself. Source Monitoring in Memory for Conversations: Who Said What to Whom, under What Circumstances? Memory researchers have examined errors in source memory or source monitoring across a number of domains. Source memory refers to memory for the context in which a target object or item of information was encountered, whereas source misattribution (i.e., failure of source monitoring) occurs when a person has some form of memory that is misattributed to an incorrect time, place, person, or other context. Such errors are considered to result from failure to bind the contextual and target information together successfully during encoding, as well as from lack of careful attempts to monitor the source of information accurately at retrieval (see reviews in Davis and Follette, 2001; Johnson et al, 1993; Koutstaal and Schacter, 2001; Meiser and Broder, 2002; Schacter, 2001). In the context of conversation, several kinds of source memory are of interest, including: • Who said what? • To whom was something said? • Did one actually say what one had considered, imagined, or planned to say? • In which conversation (of a number of possible conversations) did a particular exchange take place? • When or where did a particular exchange take place? • In what order within a conversation or interaction did a particular exchange take place? • What other participants or witnesses were present, if any? • What other features of the context or previous utterances would alter the meaning of the target utterance? • Was information acquired from a particular conversational source or from some other medium? • When planning a particular conversational contribution, has one said these things to this person before? The last of these is simply a problem of source monitoring that can create negative reactions to a speaker who repeats information all too often. However, the remaining nine are relevant in legal settings. Of these, some have been investigated empirically, whereas others remain unexplored. We will discuss each, however, pointing to its relevance in forensic contexts and reviewing empirical research when it is available. Memory of Who Said What On June 8th, 2002, a drug dealer named Jack Costos was shot and killed by a masked intruder in his home. As part of the investigation, police interviewed a group of Costos’


296

customers who had recently spent an evening together playing cards and using drugs and alcohol. The group included three Caucasians, two Hispanics, and four African Americans. Tyler Jackson—one of the African Americans—was identified as a suspect by police. Generally, the group members disliked Costos and were prone to criticizing him in their discussions. Upon questioning the group members, police focused their questions on Tyler. Two Caucasian members of the group told police that they remembered hearing Tyler say that he wanted to “waste” Costos on the night of the recent party, which was only 2 days before the murder. Did Tyler indeed make such a threat, or could this testimony have been an honest mistake of memory? Several aspects of the situation suggest that it might have been. We explore them as we consider some of the documented influences on memory for who said what in conversations. Social Categorization and Source Confusion One factor of considerable potential importance for Tyler Jackson’s case is the fact that he was one of a group of four African Americans present at the party, whereas his two accusers were Caucasian. A large body of literature has shown that when we try to remember which specific person in a group conversation has said something, we are particularly likely to attribute one person’s statement mistakenly to another member of the same social category. In this case, the two Caucasian witnesses could have confused Tyler with one of the other three African American participants. According to the source-monitoring framework proposed by Johnson and her colleagues (Johnson et al., 1993), source monitoring is more likely to be accurate when the original memory is highly elaborated and differentiated (or distinct) from other memories (see also Koutstaal and Schacter, 2001). Thus, source differentiation is encouraged by encoding conditions that encourage deep and elaborative processing and impaired by those making it more difficult to distinguish information associated with one source from that associated with another (such as two sources that are similar rather than different). Regarding conversational partners, one factor that can impair this ability to distinguish source-information connections is similarity in speaker characteristics and, in particular, similarity in the social category membership of the speaker—and, in Tyler Jackson’s case, similarity of race. The popular “who said what?” paradigm in social psychology has been widely used to demonstrate this phenomenon (Taylor et al., 1978). This line of research has shown that when asked to remember which of several speakers contributed a particular statement to a discussion, participants are more likely to misattribute the statements of one member of a social category (race, gender, age, physical attractiveness, hometown, clothing color, educational status, and many more) to another person from the same category than to someone from a different category (see the review in Klauer and Wegener, 1998). Justice Ruth Bader Ginsburg recognized this tendency to confuse members of the same social category in her complaint that attorneys arguing before the Supreme Court commonly confuse her with the other female justice, Sandra Day O’Connor (The New York Times, October 5, 1997). Although misattributions of statements between members of the same social category are common, the incidence of such confusions is affected by personal and situational factors that influence (1) the depth with which the characteristics of the speakers and their


297

statements are originally processed, (2) the salience of social categories at encoding, or (3) use of heuristic versus controlled decision criteria at retrieval. Thus, for example, category-based source confusions are more common when the person is faced with distraction or heavy processing demands at encoding than when he is encouraged to use elaborative encoding strategies. Source confusions are more common among persons for whom the social category in question is salient and who tend to think in stereotypical terms regarding the category, as well as in situations that encourage category-based thinking (such as when one’s own category is in competition with the other). Finally, source confusions are encouraged by heuristic strategies at retrieval (such as guesses based on which statement would be most likely to fit a member of a specific category, rather than careful, controlled attempts to retrieve associations between source and content; see reviews in Brewer et al., 1995; Gawronski et al., 2003; Klauer and Wegener, 1998; Klauer et al., 2002). With respect to the Costos case, source confusion between members of Tyler Jackson’s (the defendant’s) race would have been made more likely by several factors, including (1) widespread use of drugs and alcohol, which would have rendered deep and elaborate encoding more difficult, and (2) the fact that Costos was discussed (and “dissed”) by other members of the group, which would have rendered any threat made against him similar to other negative statements and complaints. The impact of age. Older persons are generally more susceptible to failures of source memory (see reviews in Davis and Loftus, Chapter 11, this volume; Glisky, 2001; Koutstaal and Schacter, 2001). Thus, it is not surprising that they are similarly more susceptible to source misattributions in conversation. This has been demonstrated in straight learning paradigms in which the person is tested on which of two speakers presented each word or item of information (Bayen and Murnane, 1996; Ferguson et al., 1992; Hashtroudi et al., 1994; McIntyre and Craik, 1987; Schacter et al., 1991), as well as in memory for who says what in conversation (Brown et al., 1995; Mather et al., 1999), and for which witness offers which testimony in a videotaped mock trial (Fitzgerald, 2000). Furthermore, older adults have more difficulty remembering whether an item was merely thought about or spoken out loud (Hashtroudi et al., 1989), although this has not yet been demonstrated in a conversational context. One thing that appears to enhance source-monitoring difficulties for older adults is similarity between sources. Thus, older persons are more susceptible to category-based source confusion than younger adults. For example, Ferguson and colleagues (1992) showed that, although older adults had poorer source memory than younger adults for information offered by two female speakers, they did not differ from young adults in source memory for those offered by a male and a female speaker. The Influence of Schematic Processing Conversational source misattributions are often the result of our expectations regarding what a person is, or is not, likely to say. Just as our schemas for particular persons, social categories, or social situations lead us to expect specific characteristics, appearance, behaviors, sequences of events, and so on, they also lead us to expect certain kinds of conversational contributions from particular sources under particular circumstances. Mather and colleagues (1999) examined the extent to which older and younger people rely on schematic knowledge to attribute statements to sources. Participants watched a


298

videotaped discussion and later attempted to remember which persons made which statements. These researchers also investigated the extent to which focusing upon oneself rather than on the speaker during encoding (thereby inducing shallow processing of speaker content) led to more schema-based source errors. Indeed, older as well as younger participants made source errors and both were particularly likely to attribute statements mistakenly to a source who, based on his category membership, would be expected to make such a statement. For example, the statement “I work out almost every day” was disproportionately misattributed to a person described as an athlete when it was actually made by a different speaker (i.e., a speakerinconsistent statement was reattributed to render it speaker consistent). Older participants made more stereotype-consistent errors in misattributing speaker-inconsistent statements than younger adults (but not more errors attributing speaker-consistent statements). The extent to which they did so was related to scores on neurological tests assessing frontal and medial-temporal area functioning of the brain (see Davis and Loftus, this volume, for further discussion of the relationship of brain functioning to source monitoring failures). Furthermore, self-focused participants made more schema-consistent source misattributions—an effect that was stronger for older participants. Similar schematic influences on speaker source memory were obtained by Spaniol and Bayen (2002), who showed that schematic effects were most likely among those with poorer overall memory, and by Gawronski et al. (2003), who showed that persons with stronger stereotyped associations (as measured by the implicit attitudes test; Greenwald et al., 1998) made more stereotype-consistent source errors. Generally, schematic influences on memory tend to be greater under conditions that encourage shallow or heuristic processing, including when veridical memory is poorer, when processing demands are high in comparison to processing resources, or when processing capacity is restricted in some way—all conditions more likely to occur for older adults, who are more likely to engage in schematic processing of a wide variety of stimuli (Bayen et al. 2000; Davis and Loftus, Chapter 11, this volume; Davis and Loftus, in press; Hess, 1999; Hess and Slaughter, 1990; Sherman and Bessenoff, 1999; Spaniol and Bayen, 2002). Tyler Jackson’s accusers may well have been affected by schematic processing when trying to recall whether, and by whom, any threats against Costos were made. Tyler was a large and muscular man who was known to anger easily and start fights. Thus, he would have seemed like the type of person who would make a threat (and even carry it out). In addition, because the police seemed to be targeting Jackson in their interviews, he would readily come to mind as a potential source of threat—thus creating the perfect conditions for misattribution to him of any threats that may have occurred. Differential Source Monitoring for Varying Conversational Roles Several investigators have asked the question of whether a person is better able to distinguish between (1) what he has said versus what another has said or (2) things two other sources have said. Raye and Johnson (1980), for example, had individuals play the role of speaker (one of two speakers alternating contributions), recorder (a person matched with each speaker who was to write down that speaker’s contributions), or listener. When later asked to identify which contributions came from which speaker, speakers did better than recorders or listeners (who did not differ). In a second


299

experiment, the authors added the role of director, which involved requesting information from each speaker. Again, speakers did better than recorders or listeners—and directors did worst of all! Apparently, the additional processing demands for the director impaired encoding. In an elaboration of the Ray and Johnson (1980) design, Brown and colleagues (1995) attempted to examine the full range of source judgments relevant to a more naturalistic conversation, including which speaker asked a question and which answered the question (oneself or one of three other participants). The exchange was structured so that each person played each role (inquirer, respondent, listener) an equal number of times. Generally, source memory was best for the respondent role—primarily due to greater success at self-identifications by responders. Strangely, for the other active role of inquirer, source memory was poor and was no better among the inquirers than among the respondents or listeners. Similarly, responders were no better than listeners at identifying who the inquirer had been. After 1 week, memory for the inquirer role had dropped to chance. Apparently, people are not good at remembering who asked questions but do better than chance at memory for who answered them. Finally, as in other arenas, source memory among older participants was poorer than among younger participants. Unconscious Plagiarism (Cryptomnesia) One of the most frequently studied conversational source misattributions has been the phenomenon dubbed unconscious plagiarism or cryptomnesia, whereby another’s ideas are mistakenly remembered as one’s own. Although deliberate instances of plagiarism are common (particularly in academic settings, for example), far more common are instances in which people in everyday life encounter ideas that later occur to them in a different context, but with no memory of the original source. The issue of plagiarism is often adjudicated in formal academic contexts—as well as in the press—when scientists or authors are accused of plagiarizing the work of others. Witness the recent flood of discussion in the press and on the Internet regarding historian Doris Kearns Goodwin’s book, The Fitzgeralds and the Kennedys, as a plagiarism of Lynne McTaggart’s previous work, Kathleen Kennedy: Her Life and Times. (A brief Google search as of 2/23/2003 produced 239 entries.) Although Goodwin claimed her plagiarism was “inadvertent” in that she simply forgot to footnote correctly, other wellknown authors have responded to accusations of plagiarism with claims of lack of awareness of the source of the material—in other words, with claims of unconscious plagiarism. For example, George H.Daniels, in his published explanation of plagiarized passages in his book, Science in American Society, explained that he had an unusually good memory and must have unconsciously memorized passages from other sources, which he later included in his book with (ironically) no memory of where they had come from (Daniels, 1972). Although this phenomenon initially appeared difficult to study in the laboratory, Brown and Murphy (1989) developed just such a procedure for studying cryptomnesia in group discussions. Groups of four people were asked to produce examples of a particular concept or category, taking turns announcing their contributions. Even while working in this initial group context, many subjects offered contributions already stated by another group member. Then, on a delayed test, subjects were asked to offer new examples of the original categories, but were very explicitly told not to use any previously offered by any


300

member of the group. Despite such instructions, many subjects nevertheless offered “plagiarized” responses from the group session, even though they truly believed them to be new (as reflected in ratings of confidence that the new contribution was a novel contribution of their own invention). Plagiarism was greater under conditions that would tend to reduce overall memory (including successful binding of source and content), including situations in which the original generation sequences were more complex, the items came from the presumably more difficult orthographic (e.g., “name a word beginning with BE”) rather than semantic (e.g., “name a fruit”) categories, and for responses generated by the person responding immediately prior to the subject (presumably self-focused concern regarding one’s own turn would be greatest at this point and reduce “binding” of the source and content of the contribution). Subsequent work has further documented this tendency toward unconscious plagiarism and has also shown that such cryptomnesia tends to become greater (Bink et al., 1999; Brown and Halliday, 1991; Brown et al., 1995; Macrae et al., 1999; Marsh and Bower, 1993; Marsh and Landau, 1995): • With increasing delay between the original group interaction and the subsequent attempt to generate novel contributions • When the original information comes from a high- rather than low-credibility source • For contributions from a member of one’s own sex (presumably a more similar and therefore more easily confused source) • When participants are distracted during the original generation of ideas • When retrieval occurs in a context different from that of the original task • For older participants Cryptomnesia is less under conditions encouraging deeper processing. Essentially, with the addition of source credibility as a factor, unconscious plagiarism tends to be most likely under the same conditions that promote other source memory errors. Foley and her colleagues (Foley and Ratner, 1998; Foley et al, 2002; Ratner et al, 2002) studied children’s source-monitoring errors for collaborative activities. Children commonly suffered from source memory errors, in that they often remembered actions performed by others as if they had performed them themselves. These researchers offered evidence that the processes of imagining or planning activities that are actually performed by others simultaneously induce false memories of performing the action oneself and facilitate learning of the activity. Memory for Medium: Did You Tell Me This or Did I Read It Somewhere? Closely related to the issue of who told us something is that of whether information came from a conversational source or from another medium such as television, radio, or printed material. This was the central issue in a malpractice case against Dr. Randolf Hicks. Mr. James Ball alleged that Dr. Hicks prescribed an aspirin daily to reduce Mr. Ball’s high risk of cardiovascular incidents such as stroke or heart attack. Mr. Ball was obese and diabetic, with an unfavorable lipid profile and a history of blood clots in his leg. However, Mr. Ball also had a history of bleeding stomach ulcers. Mr. Ball began taking a


301

full 325-mg aspirin daily and, within 2 months, suffered a near-fatal stomach bleed requiring resectioning of his stomach. Dr. Hicks denied that he had recommended aspirin to Mr. Ball. His office notes for the visit in question indicated that he had discussed options for controlling risk with Mr. Ball, and that he had prescribed Lipitor™ (a statin)—not aspirin. Mr. Ball agreed that he had been given the prescription for Lipitor™, but insisted that he had also been told to take an aspirin daily, as well as to make lifestyle and diet changes. Thus, clearly in this situation, the potential for failures of memory exist on each side. Recommendations for use of aspirin are offered in multiple sources in our culture, ranging from television ads to printed media such as health magazines to conversations with friends and professionals. Therefore, a primary concern for this case was the potential for Mr. Ball to have confused the source of the aspirin recommendation. He could have heard such a recommendation from friends or family aware of his risk profile or read it in any number of printed sources. Mr. Ball also accurately remembered that he and the doctor had discussed a number of measures for reducing his risks. Would it not seem reasonable—and even likely—that they would have discussed such a widely known measure as taking aspirin? Dr. Hicks, on the other hand, would clearly believe that he would never prescribe aspirin to a patient with a history of bleeding stomach ulcers. That is, his knowledge of what he should not do and what he routinely does not do would make it difficult for him to believe that he had done it in a specific instance. However, was the history of bleeding ulcers salient for him at the time of his discussion with Mr. Ball? Dr. Hicks is a cardiologist and was not the treating physician for the ulcers. Thus, although the information was in Mr. Ball’s records, these records were extensive, and the ulcer was not specifically the focus of discussion on the day in question. Therefore, an issue arose concerning whether Dr. Hicks in fact remembered the ulcer at the time at which he discussed measures Mr. Ball should take. If not, he might well have mentioned the possibility of aspirin (although most likely a smaller dose than Mr. Ball actually took). Memory for Target: To Whom Was It Said? Failure to remember to whom one has said something is commonly the source of interpersonal as well as legal difficulties. We may get in trouble with our spouses, parents, or close friends for failure to convey important gossip—even as we clearly remember having told them. Often, such difficulties are generated by failure to remember whom we actually did tell. Perhaps remembering that we have already told several people, we mistakenly remember that our spouse was one of them. This problem is compounded by the fact that we have often planned to tell the persons in question, perhaps even planning how and when to tell them. As we will discuss in the next subsection, it is common to remember falsely that one said something that one had actually only planned to say. In legal settings, the issue of to whom one has said something most commonly becomes relevant with respect to delivery of instructions, warnings, or information crucial to an impending decision. As noted earlier, accurate source monitoring is much more difficult when attempting to distinguish between very similar incidents than when attempting to do so for those that are dissimilar. Unfortunately, across a wide range of


302

circumstances, professionals are faced with a series of interactions that are very similar in nature. For example, doctors, lawyers, therapists, financial consultants, real estate agents, professors, military officers, salespersons and countless others interact with a large number of clients and co-workers each day. The kinds of information, instructions, orders, and so on exchanged with others are often quite similar for many (if not most) of their conversational partners across the days, weeks, months, or even years of their professional lives. This sometimes extreme similarity across interactions with many parties presents a formidable challenge for memory of to whom one has given what information (as well as, when, where, or under what circumstances it may have been given). Nonetheless, the content of a specific professional interaction with a specific person is often crucial for cases such as those involving malpractice, fraud, wrongful termination, and others. The case of Larry Wilkes illustrated just such a difficulty. Larry Wilkes was a stockbroker employed by his brokerage for 5 years before he was fired for failure to execute a sell order allegedly given by his boss, Carlton Stackerton. Less than 1 week before the incident in question, Stackerton had been transferred to Larry’s office; he supervised 20 employees, including Larry and 15 other male employees within roughly 5 years of Larry’s age. Larry was of average height, with brown hair cut in a short professional style, and thus looked similar to a number of his co-workers. Stackerton claimed to have come to Larry’s desk and given him a verbal order to sell 10,000 shares of Microsoft stock for one of the firm’s clients. This sell order was never carried out, and within the weeks that followed Microsoft fell sharply in value, leaving the client with a considerable loss. When the failure to sell was discovered by the client about 3 months later, Larry was fired for failure to carry out the sell order, which his boss confidently remembered giving to him. Larry, of course, denied having received such an order and filed suit for wrongful termination. This situation highlights several important common issues of memory. Most relevant for our current point is the issue of to whom Stackerton actually gave the sell order (assuming that he did give it to someone). Stackerton had no need to remember the target of the specific sell order in question until approximately 3 months later. Thus, in the period between when he came to the brokerage and when he needed to remember the specific Microsoft sell order for the specific client, countless orders had been given to all of the 20 brokers whom he supervised. This similarity between incidents, combined with the passage of considerable time, would have seriously compromised his ability to remember the one in question accurately. However, another issue was of considerable importance in considering the accuracy of Stackerton’s memory for the target of his sell order. That is, he had been in that office for less than 1 week and was only beginning to become acquainted with the individual brokers. Larry Wilkes was similar in appearance, age, style of dress, and hairstyle to several other brokers. Therefore, a question arose concerning whether Stackerton could have confused which of the young men to whom he had actually given the order (and, of course, no others were likely to admit voluntarily to the mistake and be fired instead). Indeed, research on eyewitness identification would suggest that similarity between innocent foils and the target suspect will tend to enhance the likelihood of a false identification.


303

Finally, the complexity of the daily activities at the brokerage presented a further barrier to accurate memory for a particular order. The brokerage was a highly active and successful business, involving a high rate of client interactions and processing of orders. The records for the day in question indicated that over 100 orders were processed on that day. Memory research would suggest that the greater the complexity of the environment in which the alleged specific order took place, the more difficult it would have been for Stackerton to bind the memory of that order successfully with memory of to whom it had been directed. The overall similarity of many instances of sell orders combined with the fast rate at which they occurred, in the context of generally complex circumstances, would present a particularly powerful mix of forces known to impair source memory. Similar issues of similarity between many instances of interactions with clients or coworkers and similarities in appearance or other features of the clients or co-workers commonly arise in memory-based litigation across many contexts. For example, a traffic cop might be asked such questions as “How many traffic tickets have you given in the 2 months since you issued this one to my client?” “So how is it that you specifically remember the exact words my client said to you that you think meant he was resisting arrest?” “Are you sure that it wasn’t Mr.___, whom you arrested that same day, who said this?” Similarly, specific instructions or warnings to a particular patient on a particular occasion become crucial to medical malpractice litigation; whether or not specific positive and negative information regarding potential investments was conveyed to a particular client becomes crucial to litigation against financial advisors; or whether or not all pertinent information was conveyed to a specific investor, buyer, or other becomes pertinent to allegations of fraud. In all such cases, if notes or other records are not available, the professional must rely on the specific memory for a specific instance of a type of conversation that occurs on a regular basis—a situation in which accurate source monitoring is notoriously difficult. Faced with such a task, the professional often falls back on script-based knowledge of what he typically does in such a situation. The doctor, for example, may commonly issue standardized instructions for specific ailments, drugs, or postprocedure care, and thus assume (and perhaps falsely remember) that he did so for the particular patient in question. In most instances, the “memory” that one performed a highly routine behavior in a specific instance will be accurate simply because the established routine makes it likely that the behavior occurred in the specific instance. However, behavior is rarely fully uniform. The doctor can plan to give an instruction or warning but forget to do so. He can be interrupted by a nurse or the patient and fail to return to the interrupted goal. In some instances, he can simply forget to provide the information to the specific patient. By the time the omission causes injury and the issue is raised with the doctor, he has seen countless other similar patients with similar problems. Without notes specifying the content of the interaction in question, it is highly unlikely that the doctor can accurately remember the conversation in question and whether he actually said the particular thing to the particular patient.


304

Reality Monitoring: Did I Say It or Just Think about Saying It? In many contexts, ranging from household conversations to professional interactions with clients, trouble arises when one party claims to have said something the other swears never to have heard. Larry Wilkes’ case of wrongful termination against his brokerage is one instance of such trouble. However, in addition to the issues of memory raised for his case, yet another issue is relevant. That is, did Stackerton actually issue the sell order, or did he simply plan to, only to become distracted by other events and forget? A similar issue arose for a local psychologist, Dr. Barkley, recently sued by the wife of a client, Mr. Bremmer, who became violent and almost killed her. Unlike many such cases in which the therapist agrees that no warning was given but disputes the contention that evidence from therapy supported the prediction of dangerousness, the therapist in this case claimed that he did warn the wife. Mrs. Bremmer, on the other hand, adamantly denied that she was ever given any warning. Did he actually warn her? Did he intend to warn her and later falsely remember that he actually did? Although we can never know what actually happened, several interesting aspects of the case parallel factors known to promote false memories of having said the (in reality) unsaid. Parks (1997) conducted a series of very clever studies designed to study false memories of having “said the unsaid.” He began with a study in which participants were shown a series of phrases. Each phrase was followed by a command to say it out loud or not. For the second study, participants were asked a series of questions, which they were asked to answer out loud or not. For the third study, subjects were asked a series of questions in a public polling situation, but for some questions the participant was interrupted before having a chance to answer. Finally, for the fourth study, subjects participated in a debate. They were led to plan to use a particular point, but ultimately were prevented from doing so. Thus, in each experiment subjects were led to think of a particular phrase, answer, or debate point; however, they were prevented from saying some, but allowed to say others. The question of interest was whether subjects would later “remember” having said the things that they had actually only thought of saying. Indeed, participants often reported with great confidence that they had actually said things which they actually had only thought. Just as in these research examples, participants in real-life conversations routinely think of things to say that may never actually be said. Perhaps most often, failure to carry out conversational intentions is the result of some form of distraction or diversion. The phone rings, another person appears and interrupts, or one’s conversation partner changes the flow of the conversation away from the topic that one had planned. For one reason or another, the statements that one has imagined never come about. Each of the participants in the preceding two example cases was faced with exactly the kinds of interruptions that could have prevented them from carrying out their intentions. The stockbroker, Mr. Stackerton, lived in the epitome of a multitasking environment in which he was fielding questions from clients and subordinates, keeping track of the markets and the office transactions, giving instructions to subordinates, and receiving their reports. In that context, he could well lose track of which things he had said or done and which he had planned but had not yet carried out. The therapist, Dr. Barkley, reported that Mrs. Bremmer (his client’s wife) was extremely agitated during his discussion with her and that he had trouble getting his points across due to her constant interruptions and topic shifts. In such a case, either


305

party’s memory could easily fail. Mrs. Bremmer, who was obviously very upset and full of anxiety, could well have failed to pay sufficient attention to what she was told to remember accurately later. On the other hand, Dr. Barkley was continually foiled in his attempts to tell Mrs. Bremmer the information he intended to convey. Just like the debate participants in Parks’ (1997) study who were prevented from making the points they planned, Dr. Barkley may well have falsely remembered giving his intended warning to Mrs. Bremmer, but was never allowed to deliver. Interruptions pose a substantial problem for professional interactions such as those between doctors and their patients. Indeed, Beckman and Frankel (1984) showed that doctors interrupted their patients within the first 18 seconds of the start of an interview. More recent findings indicate that interruptions are still a problem but that physicians in the newer study waited until 23 seconds to interrupt (Marvel et al., 1999). One result of interruptions is that, like Mr. Stackerton or Dr. Barkley, the patients failed to tell the doctor the things that they had planned. They did not get to explain their complaints fully and when the patients were able to raise them, these topics emerged late in visits with less time available to assess the concerns. If patients present with an explicit agenda but are interrupted before completing their story, it is easy to understand why there might be confusion about whether an important clinical topic was or was not actually covered. Later, the patient may complain of a misdiagnosis, fully believing that he had informed the doctor of all relevant symptoms, when in reality his plans had been disrupted by the doctor’s interruptions or by time pressures that prevented much of the discussion the patient had planned (Liang et al., 2002). Memory of Time and Space: Where, When, and in Which Conversation Did This Exchange Take Place? Studies of autobiographical memory have shown that memory for when something happened is far less accurate than memory for what happened (Larsen et al., 1996; Wagenaar, 1986, 1996). Nevertheless, memory for when something (in this case a conversation) happened is often crucial—as illustrated by the case of John Shinn. John Shinn was accused of embezzling over $200,000 from the manufacturer for whom he had worked for the last 6 years. John was identified as a likely suspect because he had access to company computers and financial transactions and was among the employees suffering the most obvious financial problems. The theft remained undiscovered for 6 months. When it was discovered, the investigation quickly focused upon six employees who had the means to commit the crime, and in short order narrowed to John. In large part, the investigation turned to focus on John due to the testimony of a single witness who reported a conversation with John that was alleged to have taken place prior to the theft. John’s co-worker, Lilly Baker, testified that John had discussed his financial problems with her on several occasions and that he had told her within a few weeks prior to the theft that he had a plan that would solve all of his financial problems. Lilly alleged that when she asked what it was, John had refused to tell her, saying that he did not want to talk about it until it actually worked out. John agreed that he had had such a conversation with Lilly Baker, but disagreed about the timing. He testified that the conversation took place approximately 3 months after the theft and that it referred to a real estate transaction he was attempting to arrange with a


306

friend. This transaction was expected to result in substantial profit for both but failed to materialize. John’s friend verified their joint attempts to arrange the transaction and the time frame in which they were working on it. Eventually, the charges against John were dropped because the police were unable to prove he had taken the money. Meanwhile, however, John lost his job, and his reputation in the community suffered permanent damage as a result of media coverage and widespread knowledge of the accusation. Lilly Baker sincerely believed she remembered the timing of her conversation with John accurately. She was a highly credible witness, one who led the police to devote considerable energy to their investigation of John as their primary suspect. What could have caused Lilly Baker’s memory for the timing of their conversation to fail? We suggest that in this case, poor general memory for timing, combined with schema-based reconstructive memory processes, might account for her failure. Generally, memory for exactly when events take place is poor—particularly when they are in the relatively distant past (in this case several months past), and when there is nothing distinctive about the timing that associates the target event with others that can clearly be placed in time (such as an event taking place on one’s wedding anniversary). Furthermore, memory for past events can be distorted by knowledge acquired after the fact. In other words, memory is reconstructed so that it seems consistent with what is currently known to be true (see the review in Davis and Follette, 2001). Studies concerned with the retrospective bias have shown that reports of past attitudes or behaviors are biased by current attitudes or recently acquired information (Dawes, 1991; Levine, 1997). In this case, Lilly’s discovery in hindsight that the crime had been committed could have led her to reconstruct the timing of her conversation with John to be consistent with her crime-relevant stereotypical knowledge of motives. Commonly held crime schemas dictate that among the most likely motives for embezzlement is being in debt (Vanous and Davis, 2001). By the time that Lilly learned of the crime, she knew that John was severely in debt and that he clearly fit the prototypical model of an embezzler in her mind. When she was questioned by the police about who among those who could have committed the crime might have a motive, John immediately leapt to mind. For her story to fit, however, the conversation in which John told her he had a plan for solving his financial problems would have had to take place before the crime, not several months after. Both parties agreed that the conversations they had regarding John’s financial problems occurred in the context of casual on-the-job coffee break chats. Thus, no distinctive events associated with the conversations existed to help locate them clearly in time. Given an unclear memory for the real timing of the conversation, Lilly may well have unwittingly reconstructed her memory for the timing of the conversation to fit her current knowledge of the crime and her hypothesis regarding who might have committed it. Memory of Order: Which of These Things Was Said When? What Was the Last Thing We Decided Upon? Memory for the sequence of events is often poor (see, for example, SchmitterEdgecombe and Simpson, 2001). This problem is found in memory for single events, in


307

which the sequence of specific actions within the event may be an issue (who hit whom first, for example). However, the problem can become particularly acute for sequences of events ranging over longer time spans, such as those typically involved in complex civil litigation. A particular problem of memory for order arises when one fails to remember which of a number of alternatives discussed in a conversation was the final choice. For example, two friends may each end up in a different restaurant (or in the same restaurant but at a different time) because several alternatives were discussed, and one person remembered the final choice differently than the other. The first author experienced an unfortunate incident in a musical performance in which musicians’ memories for the final outcome of a discussion of whether to take a particular repeat in the music differed. This problem involves two failures of source monitoring: one for the sequence of the discussion (in which minds may be repeatedly changed and changed back again) and one for the association of a particular alternative with the final decision to choose it. Both create the potential for serious misunderstanding and, perhaps, unfortunate outcomes—such as failure to meet for dinner or the cacophony created by musicians’ divergent memories for the final performance decision. This problem arose in a more serious manner in another case involving a stockbroker, Jesse Chandris, who bought 150,000 shares of a stock that his client claimed he was not authorized to buy. The stock later plummeted, resulting in a loss of over half a million dollars for the client. Chandris and his client, William Stokes, agreed that they had discussed the potential purchase, but their memories for the final decision of whether to purchase diverged. Stokes insisted his final instruction was not to buy, but Chandris insisted it was to buy. Why would their memories diverge so completely? Both agreed that they had discussed the purchase extensively, and both agreed that the pros and cons were closely balanced, making the decision difficult. Both also agreed that Stokes had changed his mind several times during the course of the conversation, instructing Chandris to buy and not to buy at various points. The sole disagreement was over the direction of the final decision. Clearly, each person’s memory was colored by motivation and self-interest, as many memories are. Furthermore, at least Stokes’ memory may have been affected by the hindsight bias (Arkes et al., 1981). Once an outcome is known, victims of this hindsight bias tend to overestimate, in retrospect, how likely that particular outcome was to occur. Blissfully ignorant of their own hindsight biases, they also fail accurately to remember their own judgments or behavior before the outcome was known. Instead, they recall that they were wiser (“I knew it would fall in value all along!”) and more confident (“…and I was sure of it!”) and that their behavior was more consistent with that knowledge (“I told you not to buy that stock.”) before the event than was actually the case (see reviews by Davis, 1991; Hawkins and Hastie, 1990). Thus, Stokes may well have fallen victim to the hindsight bias in his memory for his final instructions to Chandris. Memory of Other Participants: Who Else Was There? The case of Lydia Barnes and her sexual harassment claim against her boss, Tony Simpson, illustrates the importance of memory for other witnesses and/or participants in conversation. Ms. Barnes claimed that her boss harassed her over a period of many


308

months, through continual sexual innuendo, requests for dates and sexual favors, and implied linkage of her compliance with job advancement. Prominent among the evidence she claimed to support her allegation was a statement Mr. Barnes allegedly made in front of witnesses to the effect that she should remember “who buttered [her] bread” as she walked away from him after rebuffing a whispered advance. This alleged witnessed incident was crucial because the other reported incidents were not witnessed and boiled down to “he said” versus “she said.” Ms. Barnes claimed that two co-workers had been working nearby and turned their heads as Mr. Simpson yelled his bread-buttering threat after her as she walked away from him. Although neither had reason to fear for their jobs at the time of their testimony, each co-worker claimed to have no memory of this incident. How, in this instance, can we assume that memory failed? Other people who witness or participate in a conversation are, in source-monitoring terms, part of the context in which the focal event takes place. In order for such contextual features to be successfully “bound” to memory of the focal event, they must receive adequate encoding and processing. Ms. Barnes may well have been so focused upon Mr. Simpson and her emotional reactions to his advances and threat that she failed to adequately attend to the identities of the coworkers who witnessed the exchange. As noted earlier, highly emotional events may lead to tunnel memory in which memory for the central core of the event is enhanced, but at the expense of more peripheral aspects of the context in which it occurs (cf. Safer et al., 1993). Unfortunately for Ms. Barnes, the conspicuous failure of her two co-workers to verify her report of this crucial threatening exchange undermined her credibility with the jury. Memory of Context: What Else Happened that Would Define What Was Meant? A final sense in which memory for context can be crucial is memory for contextual features of an interaction that in essence determine the actual meaning of an utterance or conversational exchange. As noted earlier, the full meaning of a conversational contribution is revealed not only in the words, but in the tone of voice, body language, and facial expression of the speaker, as well as the immediate and historical context of each person and the dyad, and the external circumstances of the encounter. Therefore, full and accurate memory for the meaning of a verbal conversational exchange must reflect awareness of the complete personal and situational context in which it occurred. This problem was exemplified in testimony surrounding an accusation of sexual harassment against Professor Janice Hill. A female student accused Dr. Hill of sexual harassment via the content of her course in that Dr. Hill covered sexual topics in her seminar, which was required of students in the graduate program. The student further testified that she had witnessed Dr. Hill making “rude and demeaning” comments regarding a male student’s body. When asked for the details of this commentary, the student testified that she had seen the professor touch the hair on the student’s chest and sideburns and make insulting statements regarding the rapidity with which he was aging. The student was unable to offer any testimony regarding the general context of these alleged remarks other than that they occurred at a department party at a faculty home.


309

Further investigation revealed that the student in question had graduated 3 years earlier and was 5 years older than the professor. The two had become friends and worked together on consulting jobs outside academia. The student initiated the exchange of agerelated insults, whereupon the professor responded by asking whether he should be talking, given the amount of gray in his sideburns and the hair protruding from his shirt and touching the gray hair to illustrate. The entire exchange was part of a general teasing and bantering session among the student in question, the professor, and other students. Given this relational and situational context, which was verified by other student participants, the professor’s behavior took on quite a different meaning from that originally alleged by the complaining student. Clearly, responses must be interpreted in light of what they are responses to, as illustrated in the previous example. However, this issue can be relevant in a variety of ways in forensic contexts. For example, Bruck and colleagues (1999) recently explored the way in which a target utterance may be shaped by previous conversational contributions of others, who fail to remember the manner in which they triggered their partner’s responses. Specifically, they explored mother’s memories for interviews they conducted with their own preschool children. As expected, mothers’ memories for meaning were superior to those for exact wording or structure. However, they had difficulty with two forms of source monitoring: (1) whether the child’s statements had been spontaneous or prompted and (2) whether they or their child had offered particular utterances. Both of these issues are crucial to understanding testimony regarding such issues as child sexual abuse, when mothers can be expected to question their children repeatedly about the events in question. Failure to remember the way in which their own questions and suggestions may have prompted the child’s reports can contribute to their own certainty of the existence of foul play as well as increase belief in the crime by police or jurors. 12.4 Summary and Overall Conclusions As we have illustrated with a variety of case examples, testimony regarding the content of conversations is central to a vast array of criminal and civil cases litigated in our judicial system. Notwithstanding this centrality of memory for conversation, memory researchers have largely neglected basic research in this area and memory experts have rarely been asked to testify regarding the determinants of accuracy in memory for conversation. Although basic memory research offers a rich source of hypotheses regarding the determinants of memory for conversation, it remains for future research to explore the ways in which the principles governing memory for conversation converge (or not) with those governing memory for other events. Meanwhile, it is clear that memory for conversation can and does fail for most, if not all, of the reasons that memory for other events fails. Thus, memory experts can be helpful to trial attorneys faced with potentially inaccurate witness testimony regarding the contents or context of conversation.


310

Checklist of Sources of Failure of Memory for Conversation Primary factor Considerations Encoding of conversation Perception

Attention

Interpretation

Retention

Source monitoring

Impairments of hearing or vision Complex or distracting circumstances Insufficient loudness Blocked vision Distracting environment Internal states impairing attention Salience, distinctiveness; interest value of conversation Perceived threat Goals; current concerns of hearer Presence of emotion Relevance of utterance Expectations • Expectancy-consistent events • Expectancy-inconsistent events Situational circumstances interfering with processing Biasing context Difficulty of speech or language use Reliance on schemas in interpretation and filling in gaps Person characteristics interfering with understanding Use of multiple channels (facial expression, body language, words, tone of voice, etc.) Memory for meaning better than wording Recall of inferred rather than stated meaning Passage of time Misattributing utterances to speakers • Misattribution of utterances to individuals similar to the original speaker • Stereotypical attribution of utterances to speakers • Inattention to sources of utterances • Forgetting sources of utterances Misattributing utterances to other sources Misattributing the intended recipient of a message Misremembering imagined events as actual events Misremembering for time and space of conversations Misremembering for sequence and order in conversations Misremembering one’s audience Misremembering context

References Anderson, R.C. and Pichert, J.W. (1978). Recall of previously unrecallable information following a shift in perspective. J. Verbal Learning Verbal Behav. , 17, 1–12. Argyle, M. and Graham, J. (1975). A cross-cultural study of the communication of extra-verbal meanings in gestures. Int. J. Psychol. , 1, 21–28. Arkes, H.R., Wortmann, R.L., Saville, P.D., and Harkness, A.R. (1981). Hindsight bias among physicians weighing the likelihood of diagnoses. J. Appl Psychol. , 66, 252–254. Baddeley, A.D. (1999). Essentials of Human Memory . Hove, UK: Psychology Press.


311

Bates, E., Masling, X., and Kintsch, W. (1978). The role of pronominalization and ellipsis in texts: some memory experiments. J. Exp. Psychol.: Learning, Memory Cognit. , 6, 676–691. Bayen, U.J. and Murnane, K. (1996). Aging and the use of perceptual and temporal information in source memory tasks. Psychol. Aging , 11, 293–303. Bayen, U.J., Nakamura, G.V., Dupuis, S.E., and Yang, C.L. (2000). The use of schematic knowledge about sources in source monitoring. Memory Cognit. , 28, 480–500. Beckman, H.B. and Frankel, R.M. (1984). The effect of physician behaviour on the collection of data. Ann. Intern. Med. , 101, 692–696. Benoit, W.L., Benoit, P.J., and Wilkie, J. (1996). Participants’ and observers’ memory for conversational behavior. S. Commun. J. , 61, 139–155. Berscheid, E., Graziano, W., Monson, T, and Dermer, M. (1976). Outcome dependency: attention, attribution, and attraction. J. Personality Soc. Psychol. , 34, 978–989. Bink, M.L., Marsh, R.L., Hicks, J.L., and Howard, J.D. (1999). The credibility of a source influences the rate of unconscious plagiarism. Memory , 7, 293–308. Bock, K. and Brewer, W. (1974). Reconstructive recall in sentences with alternative surface structures. J. Exp. Psychol. , 105, 837–843. Bools, P.D., Lang. A., and Potter, R.E (2001). The effects of message valence and listener arousal on attention, memory and facial muscular responses to radio advertisements. Commun. Res. , 28, 627–651. Bower, G.H., Black, J.B., and Turner, T.J. (1979). Scripts in the memory for text. Cognit. Psychol. , 11, 177–220. Bransford, J.D. and Johnson, M.K. (1972). Contextual prerequisites for understanding: some investigations of comprehension and recall. J. Verbal Learning Verbal Behav. , 11, 717–726. Brewer, M.B., Weber, J.G., and Carini, B. (1995). Person memory in intergroup contexts: categorization versus individuation. J. Personality Soc. Psychol. , 69, 29–40. Brewer, W.F. and Hay, A.E. (1984). Reconstructive recall of linguistic style. J. Verbal Learning Verbal Behav. , 23, 237–249. Brown, A.S. and Halliday, H.E. (1991). Cryptomnesia and source memory difficulties. Am. J. Psychol. , 104, 475–490. Brown, A.S., Jones, E.M., and Davis, T.L. (1995). Age differences in conversational memory. Psychol. Aging , 10, 111–122. Brown, A.S. and Murphy, D.R. (1989). Cryptomnesia: delineating inadvertent plagiarism. J. Exp. Psychol.: Learning, Memory, Cognit ., 15, 432–442. Brown, J.M. (2003). Eyewitness memory for arousing events: putting things into context. Appl. Cognit. Psychol. , 17, 93–106. Brown, R. and Kulik, J. (1977). Flashbulb memories. Cognition , 5, 73–99. Bruck, M., Ceci, S.J., and Francoeur, E. (1999). The accuracy of mothers’ memories of conversations with their preschool children. J. Exp. Psychol.: Appl. , 5, 89–106. Bush, D.F. (1985). Gender and nonverbal expressiveness in patient recall of health information. J. Appl. Commun. Res. , 13, 103–117. Cegala, D.J. (1984). Affective and cognitive manifestations of interaction involvement during unstructured and competitive interactions. Commun. Monogr. , 51, 320–338. Christianson, S.-A. (1992a). Emotional stress and eyewitness memory: a critical review. Psychological Bull. , 112, 284–309. Christianson, S.-A. (1992b). Remembering emotional events: potential mechanism. In S.A.Christianson (Ed.), The Handbook of Emotion and Memory: Research and Theory (307–340). Hillsdale, NJ: Erlbaum. Christianson, S.-A. and Safer, M.A. (1996). Emotional events and emotions in autobiographical memories. In D.C.Rubin (Ed.), Remembering Our Past: Studies in Autobiographical Memory (218–243). Cambridge: Cambridge University Press. Clark, H.H. and Haviland, S.E. (1977). Comprehension and the given-new contract. In R.O.Freedle (Ed.), Discourse Production and Comprehension (1–40). Hillsdale, NJ: Erlbaum.


312

Cohen, A.A. (1977). The communicative function of hand illustrators. J. Commun. , 27, 54–63. Daniels, G.H. (1972). Acknowledgment. Science , 175, 124–125. Davis, D. (1991). What would I think if I didn’t know the patient died? From the Mind’s Eye , 1, 1– 6. Davis, D. and Follette, W.C. (2001). Foibles of witness memory for traumatic/high profile events. J. Air Law Commerce , 66, 1421–1549. Davis, D. and Friedman, R.D. (in press). Memory for conversation: the orphan child of witness memory researchers. In M.P.Toglia, J.D.Read, D.R.Ross, and R.C.L.Lindsay (Eds.), Handbook of Eyewitness Psychology ( Vol. 1 ): Memory for Events . Mahwah, NJ: Erlbaum. Davis, D. and Holtgraves, T. (1984). Perceptions of unresponsive others: attributions, attraction, understandability, and memory of their utterances. J. Exp. Soc. Psychol. , 20, 383–408. Davis, D. and Loftus, E.F. (in press). Internal and external sources of distortion in adult witness memory. In M.P.Toglia, J.D.Read, D.R.Ross, and R.C.L.Lindsay (Eds.), Handbook of Eyewitness Memory ( Vol. 1 ): Memory for Events . Mahwah, NJ: Erlbaum. Dawes, R.M. (1991). Biases of retrospection. Issues in Child Abuse Accusations , 1, 25–29. DePaulo, B.M., Lindsay, J.J., Malone, B.E., Muhlenbruck, L, Charlton, K., and Cooper, H. (2003). Cues to deception. Psychological Bull , 129, 74–112. DiMatteo, M.R. (1985). Physician-patient communication: promoting a positive health care setting. In J.C.Rosen and L.J.Solomon (Eds.), Prevention in Health Psychology (328–365). Hanover, NH: University Press of New England. Easterbrook, J.A. (1959). The effect of emotion on the utilization and the organization of behaviour. Psychol. Rev. , 66, 183–201. Eisen, M., Quas, J.A., and Goodman, G.S. (Eds.). (2002). Memory and Suggestibility in the Forensic Interview . Mahwah, NJ: Erlbaum. Erber, R. and Fiske, S.T. (1984). Outcome dependency and attention to inconsistent information. J. Personality Soc. Psychol. , 47, 709–726. Ferguson, S.A., Hashtroudi, S., and Johnson, M.K. (1992). Age differences in using source-relevant cues. Psychol. Aging , 7, 443–452. Fiske, A.P., Haslam, N., and Fiske, S.T. (1991). Confusing one person with another: what errors reveal about the elementary forms of social relations. J. Personality Soc. Psychol. , 60, 656– 674. Fiske, S.T. and Depret, E. (1996). Control, interdependence, and power: understanding social cognition in its context. In W.S.M.Hewstone (Ed.), European Review of Social Psychology (Vol. 7, 31–61). Chichester, U.K.: Wiley. Fitzgerald, J.M. (2000). Younger and older jurors: the influence of environmental supports on memory performance and decision making in complex trials. J. Gerontology: Psycholog. Sci. , 55B, 323–331. Fletcher, C.R. (1994). Levels of representation in memory for discourse. In M.A.Gernsbacher (Ed.), Handbook of Psycholinguistics (589–607). San Diego: Academic Press. Foley, M.A. and Ratner, H.H. (1998). Children’s receding in memory for collaboration: a way of learning from others. Cognit. Dev. , 13, 91–108. Foley, M.A., Ratner, H.H., and House, A.T. (2002). Anticipation and source-monitoring errors: children’s memory for collaborative activities. J. Cognit. Dev. , 3, 385–414. Fulero, S.M. and Finkel, NJ. (1991). Barring ultimate issue testimony: an “insane” rule? Law Hum. Behav. , 15, 495–508. Gamst, G. (1982). Memory for conversation: toward a grammar of dyadic conversation. Discourse Processes , 5, 33–51. Gawronski, B., Ehrenberg, K., Banse, R., Zukova, J., and Klauer, K.C. (2003). It’s in the mind of the beholder: the impact of stereotypic associations on category-based and individuating impression formation. J. Exp. Soc. Psychol. , 39, 16–30. Gibbs, R.W. (1986). Comprehension and memory for nonliteral utterances: the problem of sarcastic indirect requests. Acta Psychol. , 62, 41–57.


313

Gibbs, R.W. (1987). Memory for requests in conversation revisited. Am. J. Psychol. , 100, 179– 191. Girgis, A. and Sanson-Fisher, R.W. (1995). Breaking bad news: consensus guidelines for medical practitioners. J. Clin. Oncol. , 13, 2449–2456. Glisky, E.L. (2001). Source memory, aging, and the frontal lobes. In M.Naveh-Benjamin, Moscovitch, M., and Roediger III, H.L. (Eds.), Perspectives on Human Memory and Aging: Essays in Honour of Fergus Craik (pp. 265–276). New York: Psychology Press. Gold, G. (Ed.). (1974). The White House Transcripts . New York: Bantam. Goss, B., Neuliep, J.W., and O’Hair, D. (1985). Reduced conversational memory as a function of negative arousal: a preliminary analysis. Commun. Res. Rep. , 2, 202–205. Graesser, A. and Mandler, G. (1975). Recognition memory for the meaning and surface structure of sentences. J. Exp. Psychol. , 5, 837–843. Graesser, A.C., Gordon, S.E., and Sawyer, J.D. (1979). Recognition memory for typical and atypical actions in scripted activities: tests of a script pointer+tag hypothesis. J. Verbal Learning Verbal Behav. , 18, 319–332. Greenwald, A.G., McGhee, D.E., and Schwartz, J.L.K. (1998). Measuring individual differences in implicit cognition: the implicit association test. J. Personality Soc. Psychol. , 74, 1464–1480. Grice, H.P. (1975). Logic and conversation. In P.C.J.Morgan (Ed.), Syntax and Semantics : Vol. 3 . Speech Acts (68–134). San Diego, CA: Academic Press. Harris, R.J. and Monaco, G.E. (1978). Psychology of pragmatic implication: information processing between the lines. J. Exp. Psychol.: Gen. , 107, 1–22. Hashtroudi, S., Johnson, M.K., and Crosniak, L.D. (1989). Aging and source monitoring. Psychol. Aging , 4, 106–112. Hashtroudi, S., Johnson, M.K., Vnek, N., and Ferguson, S.A. (1994). Aging and the effects of affective and factual focus on source monitoring and recall. Psychol. Aging , 9, 160–170. Hawkins, S.A. and Hastie, R. (1990). Hindsight: biased judgments of past events after the outcomes are known. Psychological Bull. , 107, 311–327. Hess, T.M. (1999). Cognitive and knowledge-based influences on social representations. In T.M.Hess and F.Blanchard-Fields (Eds.), Social Cognition and Aging (239–263). San Diego, CA: Academic Press. Hess, T.M. and Slaughter, S.J. (1990). Schematic knowledge influences on memory for scene information in young and older adults. Dev. Psychol. , 26, 855–865. Heuer, F. and Reisberg, D. (1992). Emotion, arousal, and memory for detail. In S.-A.Christianson (Ed.), The Handbook of Emotion and Memory: Research and Theory (151–180). Hillsboro, NJ: Erlbaum. Hilton, D.J. (1995). The social context of reasoning: conversational inference and rational judgment. Psychological Bull. , 118, 248–271. Hirst, W. and Gluck, D. (1999). Revisiting John Dean’s memory. In E.Winograd and R.Fivush (Eds.), Ecological Approaches to Cognition: Essays in Honor of Ulric Neisser. Emory Symposia in Cognition (253–281). Mahwah, NJ: Erlbaum. Holtgraves, T. (1997). Politeness and memory for the wording of remarks. Memory Cognit. , 25, 106–116. Johnson, M.K., Hashtroudi, S., and Lindsay, D.S. (1993). Source monitoring. Psychological Bull. , 114, 3–28. Kassin, S. (1997). The psychology of confession evidence. Am. Psychologist , 52, 221–233. Kassin, S. and McNall, K. (1991). Police interrogations and confessions: communicating promises and threats by pragmatic implication. Law Hum. Behav. , 15, 233–251. Kausler, D.H. and Hakami, M.K. (1983). Memory for topics of conversation: adult age differences and intentionality. Exp. Aging Res. , 9, 153–157. Keenan, J.M., MacWhinney, B., and Mayhew, D. (1977). Pragmatics in memory: a study of natural conversation. J. Verbal Learning Verbal Behav. , 16, 549–560.


314

Kemper, S. and Thissen, D. (1981). Memory for the dimensions of requests. J. Verbal Learning Verbal Behav. , 20, 552–563. Kintsch, W. and Bates, E. (1977). Recognition memory for statements from a classroom lecture. J. Exp. Psychol.: Learning, Memory Cognit. , 3, 150–159. Klauer, K.C. and Wegener, I. (1998). Unraveling social categorization in the “who said what?” paradigm. J. Personality Soc. Psychol. , 73, 1155–1178. Klauer, K.C., Wegener, I., and Ehrenberg, K. (2002). Perceiving minority members as individuals: the effects of relative group size in social categorization. Eur. J. Soc. Psychol. , 32, 223–245. Kong, A., Barnett, G.O., Mosteller, E, and Youtz, C. (1986). How medical professionals evaluate expressions of probability. New Eng. J. Med. , 315, 740–744. Koutstaal, W. and Schacter, D.L. (2001). Memory distortion and aging. In M.Naveh-Benjamin, Moscovitch, M., and Roediger III, H.L. (Eds.), Perspectives on Human Memory and Cognitive Aging: Essays in Honour of Fergus Craik (362–383). New York: Psychology Press. Kramer, T.H., Buckhout, R., and Eugenio, P. (1990). Weapon focus, arousal, and eyewitness memory. Law Hum. Behav. , 14, 167–184. Larsen, S.E, Thompson, C.P., and Hansen, T. (1996). Time in autobiographical memory. In D.C.Rubin (Ed.), Remembering Our Past: Studies in Autobiographical Memory (129–156). Cambridge: Cambridge University Press. Lee, E (1993). Being polite and keeping mum: how bad news is communicated in organizational hierarchies. J. Appl. Soc. Psychol. , 23 , 1124–1149. Levine, L.J. (1997). Reconstructing memory for emotions. J. Exp. Psychol.: Gen. , 126, 165–177. Ley, P. (1966). What the patient forgets. Med. Opinion Review , 1, 71–73. Ley, P. (1979). Memory for medical information. Br. J. Soc. Clin. Psychol. , 18, 245–256. Ley, P. (1986). Cognitive variables and noncompliance. J. Compliance Health Care , 1, 171–188. Ley, P. and Spelman, M.S. (1967). Communicating with the Patient . London: Staples. Liang, W, Burnett, C.B., Rowland, J.H., Merapol, N.J., Eggert, L., Yi-Ting, H. et al. (2002). Communication between physicians and older women with localized breast cancer: implications for treatment and patient satisfaction. J. Clin. Oncol. , 20, 1008–1016. Lobb, E.A., Buttow, P.N., Kenny, D.T., and Tattersall, M.H. (1999). Communicating prognosis in early breast cancer: do women understand the language used? Med. J. Aust , 171, 290–294. Macrae, C.N., Bodenhausen, G.V., and Calvini, G. (1999). Contexts of cryptomnesia: may the source be with you. Soc. Cognit. , 17, 273–297. MacWhinney, B., Keenan, J.M., and Reinke, P. (1982). The role of arousal in memory for conversation. Memory Cognit. , 10, 308–317. Marsh, R.L. and Bower, G.H. (1993). Eliciting cryptomnesia: unconscious plagiarism in a puzzle task. J. Exp. Psychol.: Learning, Memory, Cognit. , 19, 673–688. Marsh, R.L. and Hicks, J.L. (2002). Comparisons of target output monitoring and source input monitoring. Appl. Cognit. Psychol. , 16, 845–862. Marsh, R.L. and Landau, J.D. (1995). Item availability in cryptomnesia: assessing its role in two paradigms of unconscious plagiarism. J. Exp. Psychol.: Learning, Memory, Cognit. , 21, 1568– 1582. Marsh, R.L., Landau, J.D., and Hicks, J.L. (1997). Contributions of inadequate source monitoring to unconscious plagiarism during idea generation. J. Exp. Psychol.; Learning, Memory, Cognit. , 23, 886–897. Marvel, M.K., Epstein, R.M., Flowers, K., and Beckman, H.B. (1999). Soliciting patients’ agendas: have we improved? JAMA , 281, 283–287. Mather, M., Johnson, M.K., and De Leonardis, D.M. (1999). Stereotype reliance in source monitoring: age differences and neuropsychological test correlates. Cognit. Neuropsychol. , 16, 437–458. Mazur, D.J. and Hickam, D.H. (1991). Patients’ interpretations of probability terms. J. Gen. Intern. Med. , 6, 237–240.


315

McIntyre, J.S. and Craik, F.I.M. (1987). Age differences in memory for item and source information. Can. J. Psychol. , 41, 175–192. Meiser, T. and Broder, A. (2002). Memory for multidimensional source information. J. Exp. Psychol., Learning, Memory, Cognit. , 28, 116–137. Nakao, M.A. and Axelrod, S. (1983). Numbers are better than words. Verbal specifications of frequency have no place in medicine. Am. J. Med. , 74, 1061–1065. National Archives (1996). White House tapes (Watergate Prosecution Force File Segment). Transcript of conversation, Sept. 15, 1972, 5:27–6:17 PM, Conversation No.: 779–002. (Available from National Archives). Neisser, U. (1981). John Dean’s memory: a case study. Cognition , 9, 1–22. Nussbaum, J.E and Coupland, J. (Eds.). (2004). Handbook of Communication and Aging Research . Mahwah, NJ: Erlbaum. Nussbaum, J.R, Pecchioni, L.L., Robinson, J.D., and Thompson, T.L. (2000). Communication and Aging . Mahwah, NJ: Erlbaum. O’Brien, B.J. (1989). Words or numbers? The evaluation of probability expressions in general practice. J. R. Coll. Gen. Pract. , 39, 98–100. O’Connor, A.M., Stacey, D., Rovner, D. et al. (2001). Decision aids for people facing health treatment screening decisions. Cochrane Syst. Rev. , 3. Ohnishi, M., Fukui, T, Matsui, K., Hira, K., Shinozuka, M., Ezaki, H. et al. (2002). Interpretation of and preference for probability expressions among Japanese patients and physicians. Fam. Pract. , 19, 7–11. Parkin, A.J., Lewinsohn, J., and Folkard, S. (1982). The influence of emotion on immediate and delayed retention: Levinger and Clark reconsidered. Br. J. Psychol. , 73, 389–393. Parks, T.E. (1997). False memories of having said the unsaid: some new demonstrations. Applied Cognitive Psychology , 11, 485–494. Pezdek, K. and Prull, M. (1993). Fallacies in memory for conversations: reflections on Clarence Thomas, Anita Hill, and the like. Appl Cognit. Psychol. , 7, 299–310. Ptacek, J.T. and Eberhardt, T.L. (1996). Breaking bad news: a review of the literature. JAMA , 276, 496–502. Ratner, H.H., Foley, M.A., and Gimpert, N. (2002). The role of collaborative planning in children’s source-monitoring errors and learning. J. Exp. Child Psychol. , 81, 44–73. Raye, C.L. and Johnson, M.K. (1980). Reality monitoring vs. discriminating between external sources of memories. Bull. Psychonomic Soc. , 15, 405–408. Robertson, W.O. (1983). Quantifying the meanings of words. JAMA , 249, 1061–1065. Rogers, W.T. (1978). The contribution of kinesic illustrators toward the comprehension of verbal behavior within utterances. Hum. Commun. Res. , 5, 54–62. Safer, M.A., Christianson S.-A., Autry, M.W., and Oesterlund, K. (1993). Tunnel memory for traumatic events. Appl. Cognit. Psychol. , 12, 99–117. Sande, G.N., Ellard, J.H., and Ross, M. (1986). Effect of arbitrarily assigned status labels on selfperceptions and social perceptions: the mere position effect. J. Personality Soc. Psychol. , 50, 684–689. Schacter, D.L. (1999). The seven sins of memory: insights from psychology and neuroscience. Am. Psychologist , 54, 182–203. Schacter, D.L. (2001). The Seven Sins of Memory . New York, NY: Houghton Mifflin. Schacter, D.L., Kaszniak, A.W., Kihlstrom, J.F., and Valdiserri, M. (1991). The relation between source memory and aging. Psychol. Aging , 6, 559–568. Schmidt, S.R. (1994). Effects of humor on sentence memory. J. Exp. Psychol.: Learning, Memory, Cognit. , 20, 953–967. Schmitter-Edgecombe, M. and Simpson, A.L. (2001). Effects of age and intentionality on content memory and temporal memory for performed activities. Aging, Neuropsychol. Cognit. , 8, 81– 97.


316

Schneider, B.A., Daneman, M., Murphy, D.R., and See, S.K. (2000). Listening to discourse in distracting settings: the effects of aging. Psychol. Aging , 15, 110–125. Sherman, J.W. and Bessenoff, G.R. (1999). Stereotypes as source-monitoring cues: on the interaction between episodic and semantic memory. Psychological Sci. , 10, 106–110. Sillars, A.L., Weisberg, J., Burggraf, C.S., and Zeitlow, P.H. (1990). Communication and understanding revisited: married couples’ understanding and recall of conversations. Commun. Res. , 17, 500–522. Singer, S. (1969). Factors related to participants’ memory of a conversation. J. Personality , 37, 93–100. Spaniol, J. and Bayen, U.J. (2002). When is schematic knowledge used in source monitoring? J. Exp. Psychol.: Learning, Memory, Cognit. , 28, 631–651. Spear, N.E. and Riccio, D.C. (1994). Memory: Phenomena and Principles . Boston: Allyn and Bacon. Sperber, D. and Wilson, D. (1995). Relevance: Communication and Cognition . Cambridge, MA: Blackwell. Stafford, L., Burggraf, C.S., and Sharkey, W.F. (1987). Conversational memory: the effects of time, recall, mode and memory expectancies on remembrances of natural conversations. Hum. Commun. Res. , 14, 203–229. Stafford, L. and Daly, J.A. (1984). Conversational memory: the effects of mode and memory expectancies on remembrances of natural conversations. Hum. Commun. Res. , 10, 379–402. Taylor, S.E., Fiske, S.T., Etcoff, N.L., and Ruderman, A.J. (1978). Categorical and contextual bases of person memory and stereotyping. J. Personality Soc. Psychol. , 36, 778–793. Thompson, C.P., Herrmann, D.J., Read, J.D., Bruce, D., Payne, D.G., and Toglia, M.P. (Eds.). (1998). Eyewitness Memory: Theoretical and Applied Perspectives . Mahwah, NJ: Erlbaum. Vanous, S. and Davis, D. (2001). Motive evidence: probative or just prejudicial? Paper presented at the Rocky Mountain Psychological Association, Reno, Nevada. Villaume, W.A., Brown, M.H., and Darling, R. (1994). Presbycusis, communication and older adults. In M.L.Hummert, J.M.Wiemann, and J.F.Nussbaum (Eds.), Interpersonal Communication in Older Adulthood: Interdisciplinary Theory and Research (pp. 83–106). Thousand Oaks, CA: Sage. Wagenaar, W. (1996). Autobiographical memory in court. In D.C.Rubin (Ed.), Remembering Our Past: Studies in Autobiographical Memory (180–196). Cambridge: Cambridge University Press. Wagenaar, W.A. (1986). My memory: a study of autobiographical memory over six years. Cognit. Psychol , 18, 225–252. Woloshin, K.K., Ruffin IV, M.T., and Gorenflo, D.W. (1994). Patients’ interpretation of qualitative probability statements. Arch. Fam. Med. , 3, 961–966. Wyer, R.S.J., Budesheim, T.L., Lambert, A.J., and Swan, S. (1994). Person memory and judgment: pragmatic influences on impressions formed in a social context. J. Personality Soc. Psychol , 66, 254–267.

III Driving Environments

13 Human Factors in Traffic Crashes Rudolf G.Mortimer Human Factors Engineering Richard D.Blomberg Dunlap and Associates, Inc. Gerson J.Alexander Positive Guidance Applications Evelyn Vingilis University of Western Ontario 0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

13.1 Introduction to Human Factors Principles and Standards of Care Traffic crashes are the most frequent cause of unintended fatal injuries in the U.S. when compared with accidents occurring in public places, in homes, or at work places (National Safety Council, 2001). They accounted for about 41,821 deaths and 3.2 million injuries that resulted from about 6.4 million crashes in 2000 (USDOT, 2001). Though these numbers are staggering and involve huge personal, emotional, and financial losses, the fatality rate per distance traveled has been steadily declining. If the fatality rate per vehicle kilometer in 2000 had remained the same as it was in 1970 at 2.9 fatalities per 100 million km, there would have been about 131,000 fatalities in 2000. Therefore, it is evident that highway safety has improved significantly over the years. This trend has also been found in most other countries. The rates/distance traveled in Finland, Sweden, and the Netherlands are somewhat lower than in the U.S., while those in France and Israel, for example, are about 60% greater. Some countries, such as Morocco and Turkey, have rates about 20 and 10 times greater, respectively, than many European countries (National Safety Council, 2002). It has been estimated that road traffic crashes result in about 22 deaths per 100,000 population worldwide and are about 40% of unintended deaths. Among the factors that have contributed to the reduction in fatalities are improvements in the protection afforded to occupants of cars and trucks, such as air bags, seat belt systems, child restraints and interior packaging; stepped-up enforcement and educational campaigns to reduce impaired driving, especially targeted against drinking drivers; restricted licenses for young drivers, etc. At the same time, other events have increased the toll of injuries and fatalities, such as the allowable duty time of truck

Human factors in traffic crashes

321

drivers (which can lead to fatigue); elimination of helmet laws for motorcyclists in many states or requiring helmets to be worn only by the young; inadequate lighting and marking of bicycles and cyclists at night; proliferation of jogging as a health measure with its attendant hazards when joggers use the roadway or cross it (especially at night); the historical difficulty of seeing pedestrians in darkness; and continued problems of visibility and glare in night driving, among many others. The vehicle, the environment, and the human all contribute variables that need to be considered in any analysis of traffic collisions. The emphasis in this chapter is on the human as a participant in the traffic system as a pedestrian, cyclist, or an operator of a vehicle, a perceiver of the environment, a decision maker, and a responder. The environment includes that within the vehicle and outside it. The vehicle and its relevant characteristics are those assimilated by the human so that the operator’s input and the vehicle’s output relationships have been learned. However, the vehicle also includes input-output relationships that are not well learned or understood by the operator, such as car drivers braking on a slippery road surface or maximum braking of motorcycles. The highway-traffic-environment-human system is dynamic and complex. There are no simple solutions to increasing safety and no simple reasons for crashes. As Haddon (1970) pointed out, the system is multidimensional and requires an epidemiological approach to unearth the underlying causes of crashes. The human as driver, pedestrian, motorcyclist, cyclist, and other participant in the traffic system is at the center of it and has been assigned a large proportion of the causes of crashes, as the principal cause or as one cause in association with other parts of the system. In almost all cases, it would be inappropriate to allocate cause only to the human element, independently of the other components. It is important that those involved in forensic work have a systems orientation so that the performance of the human is considered in the context of the overall situation. Forensic Data Gathering Traffic collisions result from a failure in one or more parts of the traffic system. The forensic analysis starts with the collection of the basic factual data that may include a police report; an accident reconstruction; statements of witnesses and the actors directly involved; photographs or video pictures of the scene and vehicles; and a survey of the area drawn to scale. These data are usually supplemented by depositions of persons having some knowledge that may be of relevance. An examination of the location of the scene is useful or imperative to allow observations and measurements of distances, lighting, or sound features, assuming that the scene has not been significantly altered in the interim. Based on such background information, the human factors investigator applies fundamental principles of psychology, physics, and other sciences to discern and evaluate the underlying failures that were causally related to the event and the means by which those failures may have been avoided. In evaluating the standards of care that may have been breached and that contributed to the event, it is not unusual for the human factors person to work with an accident reconstructionist or specialist in another discipline; thus, the combined effort results in a realistic appraisal of the contributing factors to the failure


322

and the standards of care that should have been in place that could realistically, practically, and economically have been expected to be employed. 13.2 Objective and Scope The subject matter of this chapter is extensive and books have treated various aspects of it. The chapter must be incomplete in some respects but will attempt to deal with some of the major areas of forensic human factors in the most common or lethal traffic collisions. The major topics of this chapter include human factors aspects of: • Vehicles, such as marking and signaling systems and their role in rear-end crashes; headlighting for night-driving visibility and its effects on drivers’ visibility and night crashes; motorcycle braking systems and rider performance in crash avoidance; and railroad crossing safety factors in drivers’ information processing (Section 13.3, prepared by R.Mortimer) • Pedestrian and bicyclist safety (Section 13.4, prepared by R.Blomberg) • Highway signing and driver information systems (Section 13.5, prepared by G.Alexander) • Effects of alcohol and drugs on traffic safety and human performance (Section 13.6, prepared by E.Vingilis)

13.3 Human Factors Engineering Affecting Visibility and Perception, and Performance of Drivers, Motorcyclists, and Other Road Users Visibility in Night Driving The rate of crashes at night is about four times that in daytime (National Safety Council, 2002). Although other factors are at play, such as a greater incidence of alcohol use in hours of darkness, reduced visibility must be a major factor in the elevated risk in night driving. One way this is demonstrated is by the expected 30% reduction in crashes that occurs when street lighting is installed where none was before (Rowan and Walton, 1976). Drivers must rely on the illumination provided by headlights when other lighting is not provided. Motor vehicles are normally equipped with headlamps that provide a high beam and a low beam. The requirements for the performance of headlamps in the U.S. is set forth in Federal Motor Vehicle Safety Standard 108, “Lamps, Reflective Devices, and Associated Equipment,” which first became effective in 1968. The headlamp performance standard was essentially based on the Society of Automotive Engineers Standard J579a. The federal standard was amended in 1978 to adopt the revised SAE J579c Standard, which allowed a doubling of the total intensity of the high beam to 150,000 cd and an increase of 5000 cd (to 20,000 cd) in the maximum permitted intensity at the 0.5° down to 1.5° right seeing point. The change in the maximum intensity of the U.S. high beam brought it closer to the intensities permitted in Europe. The change in the low beam has meant an increase in direct glaring intensities from oncoming vehicles and


323

indirectly by headlights of following vehicles reflected in rearview mirrors with some increase in visibility. The High Beam The high beam can only be appropriately used when there is no oncoming traffic and no traffic close ahead so as to avoid excessive glare to other drivers; it provides a wide swath of light with its greatest intensities aimed about parallel to the road surface. The intent is to provide visibility of the road at substantial distances ahead of the vehicle and to the sides of the traveled lane. Those goals can be readily achieved with high-intensity light sources. The Low Beam The low beam is designed to be used when meeting other traffic or when another vehicle is ahead and nearby. Low beams are now also used in urban areas in most countries with or without street lighting, although they have not been designed specifically to be used in that environment. The design of the low beam has been undergoing slow evolution over many years in order to try to compromise the need to maximize visibility while minimizing discomfort and disability glare. In a situation in which two cars drive toward each other on a straight, unlighted, level road at night, with both using the low beam, the visibility of the drivers will gradually decrease as the cars come closer due to the glare from the opposing car’s headlamps. Recovery from glare in this case will occur before the cars meet because the glare effect is an inverse exponential function of the angle between the glare source and the driver’s direction of gaze. At large separation distances between the vehicles, that angle is small and the glare effect is potentially large because it is then mainly a function of the glare illumination at the driver’s eyes from the oncoming car. When the cars are approximately 100 m apart and drawing closer, the glare angle becomes much larger, reducing the glare effect so that the visibility can increase. The eye’s sensitivity is rapidly affected by glare or any bright source; when it is adapted to the usual mesopic levels found in night driving, the readaptation of the eye to lower levels of glare and illumination is relatively slow and can take a number of seconds. Up to 10 seconds may be needed for approximately full recovery from headlight glare (Mortimer and Becker, 1973). Figure 13.1 shows an example of the visibility distance to a 12% reflectance object at the right edge of the road for car drivers at 30 mph approaching each other with low- and high-beam headlamps on a level, two-lane road. In this example, the visibility is a minimum at about 400 ft for the high beam meeting and at about 300 ft for the low beam before the cars meet and pass each other. Then recovery from the glare allows the visibility to increase. Figure 13.1 also shows that the curves cross when the cars are about 2400 ft (731 m) apart, indicating the separation distance when dimming from the high beam to the low beam should occur to best maintain visibility. However, many variables affect this general process, such as the geometry of the highway; reflectivity of the pavement; lateral separation between the vehicles; location of the object to be detected (pedestrian, parked vehicle, bicyclist, sign, delineation stripe, etc.); reflective properties of the object; approaching vehicles; following vehicles;


324

properties of interior and exterior mirrors; mounting location of headlamps and the properties and aim of their beams; street lighting; environmental conditions such as fog, rain, and snow; and driver variables such as age, workload, and attention.

FIGURE 13.1 Visibility of car drivers meeting on two-lane road with low and high beams. Because of the many variables to consider that can affect the ability of a driver to detect and recognize an object, it is evident that no simple methodology can be reasonably applied to predict a driver’s visibility. In a human factors recreation of a collision, it is essential to know as many of the important variables as possible. Some of those will be detailed in the police report, photographs, and scaled diagrams of the scene. The paths traveled by vehicles, cyclists, or pedestrians may be determined from skid marks, postimpact trajectories, marks on the pavement, and other features normally provided in the accident reconstructionist’s report. The statements of witnesses may also provide useful information, although they are sometimes inconsistent with physical evidence and need to be treated with due respect. Night Photography Photographs taken at night will show limited information. Also, those photos will normally be taken with a flash device in order to show as much of the scene as the camera and film are capable of capturing. Retroreflective materials will be highlighted in such night photography. It is important to remember that the scene as shown in those photos will be different from what was visible to a driver who was relying on the illumination provided by headlamps. One important reason is that flash strobes have a much wider field of illumination and also are usually at greater intensities than headlamps. Photographs are sometimes made in a recreation of the scene at night in an attempt to simulate the view available to a driver. Attempts to reproduce what was seen by a driver


325

at night using photos, film, or digital media most often cannot do this fairly, even when an effort is made to use a control object with which to compare the scene, such as varying shades of gray that are part of the picture. The photos are then made at various exposures so that the gray that is just at threshold in the actual scene and in a photo is taken as an estimate of the apparent visibility of other objects of interest in the photo to replicate a driver’s view (Holohan et al., 1989). That process may have some limited value in a very bare night scene, but if other lights than the headlights of the driver’s vehicle are present—and particularly if there is an approaching vehicle whose headlights will produce glare at the eyes of the observer—the photographic method is fraught with error and cannot be used as a representation of the visibility. The recording medium, film or tape, only has about 1/10,000 of the range of sensitivity of the eye, so true gradations of light and darkness are not reproducible as seen by the eye. In a case in which glare sources are in the scene, such as from the headlights of vehicles or from street lighting or extraneous sources such as store fronts or parking lot lighting, photographic or video methods are incapable of providing a fair rendition of the visibility. In addition, still photographs can present a false impression on those who view them, such as the jury, because they provide an unrealistic amount of time for the scene to be viewed and because the jury has already been alerted to the location and nature of the object of relevance (Hyzer, 1993). On the other hand, the driver had a limited time to view the scene and did not have advance information of the nature or location of the object. Because crashes occur in a dynamic context, still photos can create a false presentation. Another reason that night photography, whether still or motion, is not able to reproduce the scene viewed by the eye is that the driver’s eyes are subject to changing levels of adaptation, which directly affect the sensitivity of the eyes and the thresholds at which objects can be detected. Cameras have no such characteristic. Visibility Reconstruction In order to try to estimate what was discernable to a driver before a night crash, a type of visibility reconstruction may be undertaken. This is not the same as an accident reconstruction whose objective is to determine the paths taken by the vehicles and pedestrians, as well as their speeds and trajectories, shortly before, during, and after the crash. Based on the accident reconstruction, the vehicles and other relevant objects can be inserted into the scene, assuming that the environmental factors are still the same. In a simple case, such as one involving a vehicle on a two-lane road and a pedestrian who was struck while walking on the pavement, the recreation can be done reasonably easily. This assumes that the road has not been repaved; lighting has not been inserted or removed; other traffic can be controlled; weather conditions are similar; and the moon phase is the same, particularly if it was a clear night without clouds. The pedestrian must be simulated with a person of similar stature, wearing the same types of clothing. The original clothing from the victim, including shoes, socks, lower and upper body outer clothing, and headwear, must be measured for reflectance. The clothing to be used in the reconstruction must be of similar reflectance. If the original clothing is no longer available, the best estimate of the reflectance must be used. For example, in one such case, the pedestrian was known to have worn blue jeans but none of


326

the actual material was available. Therefore, a survey was made of 20 randomly selected persons wearing blue jeans and the reflectance of their pants was measured. The minimum reflectance was 2% and the maximum 23% with a median of 9.5%. The range of reflectances is quite large and shows that knowledge of the actual reflectance is important and affects the visibility of the blue jeans in darkness. Knowing the distribution of the reflectances can allow the effects to be taken into consideration. The vehicle should be the same type as or similar to the one in the crash and be equipped with the same types of headlamps and auxiliary lamps if such were used. The headlamps should be correctly aimed, unless there is some reason to assume otherwise. Headlamp alignment can be affected by the loading of the vehicle, so that needs to be considered. The driver/observer in the vehicle should have normal eyesight unless the driver in the crash had known deficiencies, such as color blindness or monocular vision, in which case those would need to be taken into account then or later. The approximate location of the collision usually will be known from the police report or accident reconstruction. That point of impact acts as a reference point from which known distances can be measured and marked with paint or by small traffic cones. For example, in making the visibility reconstruction, the distance at which an object, such as a pedestrian, becomes just detectable may be sought. By having markers at the side of the road that show the distance from the point of impact, the distance at which the pedestrian becomes detectable, for example, can be readily obtained by noting the position of the car with respect to the nearest marker when detection occurs. To allow the markers or cones to be easily seen at night by a car’s headlights, it is useful to apply reflective tape to each and to number them. The forensic expert can make actual measurements of visibility by driving the car toward the pedestrian at a slow or moderate speed, not necessarily at the speed of the vehicle in the collision, so that the reaction time of the driver is not a factor. The position of the car when the pedestrian becomes just detectable is noted and is the detection distance. A number of such trials should be made to ensure that the data are reliable. Alternatively, another person of similar characteristics as the driver in the crash can be used to make the visibility observations. The disadvantage is that the person may be requested to appear at a deposition to describe what was done. It is important that good notes are kept of the procedure and the data. Photos should be made in daylight to show the vehicle and the test location and setup, which may become useful exhibits for later use in court to demonstrate to the jury. Although the use of night photography has its limitations, such still or motion photography can be used to illustrate the general scene to the jury. In the present example, a video made during the approach to the pedestrian could have a narrative simultaneously describing what can be seen by the observer and when the pedestrian becomes detectable by the eye. This would provide a foundation for use of the video recording as a support of the actual observations. It is not unusual that, very correctly, opposing counsel makes strong objections to the introduction of video or other photography. The underlying purpose of the photography or video needs to be made clear: namely, it represents an approximation of the visual scene and not a rendition of it as would be seen by the eye. In these kinds of tests, it will usually be noted that the pedestrian will be detectable at a certain distance and then will become much more clearly visible a short distance later. The detection criterion adopted by the observer can introduce variability, so it needs to be


327

specified clearly. The detection phase is usually the one with the greatest variability, but the objective of the visibility reconstruction is to determine the distance at which the pedestrian would be at threshold of detection, given an observer who is looking in the appropriate location. This set will reduce the variability in the measurements. It is also helpful to obtain the distance at which the pedestrian is clearly detectable, which is often also when he is recognizable. This will provide an endpoint to the detection distance, when almost all observers looking in the direction of the pedestrian would detect it. The example just presented of a nighttime collision with a pedestrian on an unlighted road with no other traffic is relatively simple to examine in the general fashion described. However, if street lighting is present, the scene is more complex because of the relationship of the location of the pedestrian with respect to the luminaire. This can result in direct visibility of the pedestrian by illumination from the street lamp and the headlights of the car. Alternatively, if the pedestrian is between the street lamp and the driver of the vehicle, the pedestrian may be visible to the driver initially in silhouette as a dark object against the road. As the car gets closer, the headlamps increasingly illuminate the pedestrian, who will be more difficult to detect because of the reduction of the contrast of the pedestrian against the lighted road behind the pedestrian. At a certain point, the pedestrian or a part of the pedestrian may be momentarily lost to view when the luminance of the pedestrian is equal to that of the pavement that is the background of the pedestrian as viewed by the driver of the car. Then, as the car gets closer and the headlights provide more illumination, the pedestrian can be seen as a brighter object against the pavement. Therefore, the location of the pedestrian over time can be very important in affecting the illumination from street lighting and car headlighting; the effects on the resulting contrast of the pedestrian against the pavement are important determiners of visibility. For an example of the procedures in an analysis of the night visibility aspects of a pedestrian collision, see Mortimer (2001). Glare Another complicating factor often involved in night crashes is the glare from oncoming vehicles and from vehicles to the rear. The latter glare levels can often be greater than those from oncoming vehicles (Miller et al., 1974). Glare from headlights and other sources can have the effect of reducing the ability to see and causing discomfort, thus leading to fatigue and other responses. Although drivers are told not to look directly at the headlights of oncoming vehicles to reduce the deleterious effects of glare, eye fixation data (Mortimer, 1975) show that drivers do look at oncoming vehicles at night, at least on two-lane highways. Probably the reason is that drivers have a concern for the lateral position of those vehicles to ensure they are not encroaching into their lane of travel. Glare reduces visibility because the light entering the eyes of the driver from the source, such as headlights of an oncoming vehicle, effectively reduces the contrast of the object’s image on the driver’s retina. The glare effect is a veiling brightness added to the background brightness, thereby effectively raising the denominator in the contrast equation and reducing the contrast between the object and its background. The result is that the object is less detectable than if the glare was not present.


328

The veiling brightness is a function of the illumination produced at the eyes by the glare source and the inverse of the glare angle (the angle of the glare source from the driver’s line of gaze) raised to a power. For example, using the equation of Fry (1954), the veiling glare is shown as: VG=KI/D 2θ (θ+1.5), where I is the glare intensity in candelas at the eye; D is the distance, in feet, from the glare source; θ is the glare angle in degrees; and K is a constant determined by the age of the observer. It should be evident that the effect of the glare drops off sharply as the glare angle increases. When two vehicles approach each other at night, the angle between the forward line of sight of the drivers and the oncoming vehicle will increase rapidly as the separation distance between the vehicles becomes small. This is a reason that the visibility will usually begin to increase before the vehicles pass each other, as shown in Figure 13.1. Another major factor is the illumination (I/D 2 ) at the eyes of the driver from the oncoming headlamps, which will be a function of the distribution of light from the headlamps. Age is also related to disability glare (Wolf, 1960), and older drivers can attest to the impairing and discomforting effects of glare, which become increasingly pronounced after about 40 years of age. In one series of experiments (Olson and Sivak, 1983), drivers aged 18 to 30 and 65 or more recognized low-reflectance typical pedestrians standing at the right and left side of the road when there was no oncoming car (no glare) and with an oncoming car at 300 ft (91 m) behind the location of the pedestrian. The older drivers recognized the pedestrian on the right side at about 38% of the distance compared with the drivers aged 18 to 30 when using conventional low-beam headlamps without glare from an oncoming vehicle. When there was an oncoming vehicle, the visibility of the young drivers was about 78% of the no-glare case. The oncoming vehicle reduced the visibility distance of the pedestrian for the older drivers to about 75% of their no-glare case.

TABLE 13.1 Recognition Distancesa in Darkness of Dark Pedestrian at Right and Left Sides of Road by Young (18 to 30) and Older (65+) Drivers Pedestrian location

No glare Young Older

Low-beam glare Young Older

Right side 50 19 39 14 Left side 27 15 a Meters. Source: Adapted from Olson, P.L. and Sivak, M., 1983. University of Michigan, Report UMTRI 83–9, Ann Arbor, MI.

Table 13.1 also shows that the location of the object has a substantial effect on its visibility by low-beam headlamps. The pedestrian on the left side of the road was recognized by the young and older drivers at about 53 and 79%, respectively, of the distances he was seen when on the right side of the road. The data shown in the table can be used as a general indication of the trends in recognition distances of a pedestrian wearing dark clothing with low-beam headlamps to show the effect of the location of the object, an opposing vehicle, and age of the driver. The data also demonstrate that older drivers need more illumination than younger drivers to see objects at night as well as young drivers do. However, the mean visibility


329

distance of the younger drivers in this test was also far from what is needed even when there was no glare. At a speed of 88 km/h (55 mph), a vehicle will cover 50 m in 2.04 sec, which will not allow a driver not previously alerted to the pedestrian enough time to react to stop within that distance. Because of the design of the low beam to limit light in the locations of the eyes of approaching drivers, minimal light is distributed to the left side of the road in the lanes used by oncoming traffic. This means that a pedestrian crossing from the left to the right at night will be particularly difficult to see until he enters the lane used by the driver and moves into the part of the beam that includes higher intensities of light. Drivers will have less warning of the pedestrian crossing from the left than of one crossing their path from the right. Computational Analysis It should be evident that it is a complex task for a driver to estimate the detectability of an object, such as a vehicle, retroreflective device, or pedestrian, at night due to the many variables involved. One approach is the site examination and visibility reconstruction already described. Another approach is to use calculations or computer simulations. A few simulations to estimate detectability of objects in night driving have been developed, but the accessibility of most appears to be limited. Among the earliest is the one developed at the University of Michigan (Mortimer and Becker, 1973; Becker and Mortimer, 1974) that extended the work of Jehu (1955) of the Road Research Laboratory in England. That model simulates the approach of two vehicles on straight or curved roads at night and calculates the distance at which a simple characteristic of a specific test object is recognized as the vehicles approach and after the vehicles pass each other, as visibility is regained. The output of the program can be generalized in a limited way to objects other than the kind used in the validation studies. Mortimer (1974) offers an example of the extensive way in which the program can be used. The Ford Motor Company developed a program (Bhise et al., 1976; Farber and Matle, 1989) based on the visual target detection data of Blackwell (1952). This program is more versatile with respect to the kinds of objects that are to be detected and also incorporates age-related functions. Although probably not enough research has been done to validate the model, it is a useful program. Like all programs, it needs to be used judiciously and requires a fundamental understanding of factors affecting visibility. Although other simulation programs have been reported in the literature, their details are unclear and do not appear to be publicly available. Hand calculations can also be made, relying on the contrast threshold data of Blackwell, some of which are summarized by the Illuminating Engineering Society (1972). In any computational method, the characteristics of the headlamps and street lights will need to be known, usually in the form of iso-candela diagrams. When the locations of the target object (e.g., pedestrian) and the vehicle at a given point in time or distance from the collision are known, the illumination from the light source can be calculated. Using the reflectance of the clothing and the illumination on it, its luminance (B t ) is calculated (B t = illumination×reflectance).


330

FIGURE 13.2 Photo looking south. The tractor-trailer pulled out from the right leg of the intersection and turned to go north. For an object seen against the road, the luminance of the road at the object can be used as the background luminance. However, it may be more appropriate to use the driver’s adaptation level, which is determined by the general illumination in the driver’s vicinity; this is often a function of the foreground illumination provided by the headlamps and the reflectance of the pavement or perhaps by street lighting. The available contrast can then be calculated and compared with Blackwell’s data of the contrast needed for a given probability of detection. The veiling glare caused by one or more approaching vehicles and by street lights can also be incorporated into the contrast equation. Example of a Forensic Examination of a Nighttime Car-Truck Collision The photo in Figure 13.2 shows a view to the south of a rural unsignalized intersection. At about 11 p.m. in December, 1992, a tractor and flat-bed trailer pulled out from the right (eastbound) leg of the intersection. The tractor crossed the southbound lane and turned left to head north. A few seconds later a southbound car struck the truck’s trailer just ahead of the rear tandem axles. The speed limit was 55 mph. As the car approached the intersection, it was descending around a left and then a right curve that led to the intersection. Prior to the crash, a truck traveling north had passed the southbound car about 122 m from the intersection. It will be noted that a street light was on the northeast corner and some lamps were also on the porch of the white building (a restaurant) and in some of the windows. In addition, on the east side of the road was a gas station that had just closed. A dynamic analysis of the truck’s acceleration into the intersection and its path of travel was made to enable the position of the truck and the car to be fixed by time to crash. Analyses were made of the visibility of the trailer’s side rail, retroreflectors in the


331

center and rear side of the trailer, and the side-marker lamps that were standard, required equipment at the time and with which the trailer was equipped. Those analyses were made at intervals of 30 m starting at 122 m before the crash. Further analyses were made in which the trailer was assumed to have a continuous strip of retroreflective red/white alternating striped tape on the side such as that now required on all large trailers. The analyses included the effects of the street lamp in illuminating the trailer and also the effect of the glare of the low-beam headlamps of the tractor. Photometric measurements were made of the retroreflective performance of the standard yellow and red retroreflectors, the retroreflective tape, and the side-marker lamps as a function of their angle to the light source (“entrance angle”). This was done because the amount of light returned from retroreflective devices decreases rapidly at entrance angles greater than about 40° and the specifications for the performance of side-marker lamps do not go beyond 45° (FMVSS-108) (SAE J-592). The basic results of the analyses showed that the side-marker lights were detectable at 50% probability or better between 122 to 30 m from the trailer; the yellow retroreflector in the center of the side of the trailer was detectable at 122 to 91 m and the red at 91 m; the side of the trailer body was detectable at 15m; and the retroreflective tape was detectable at 122 to 61 m. (The analysis of the visibility of the tape was not done at 30 m.) The basic scene was recreated at the site of the crash and night observations were made of the visibility of the trailer at positions of the car and trailer equivalent to 1-sec intervals starting at 10 sec before the crash. With the trailer in the same condition as it was in at the time of the crash, the center yellow side marker was detectable at 78 m, but the trailer only became faintly visible at about 1 sec to crash. With tape on the side of the trailer, it became quite visible at about 205 m and clearly visible at 130 m; at 78 m, it was clearly noticeable that the tape went across the southbound lane. When additional sidemarker lamps were added to the side of the trailer with a separation of 1.5 m between them, they could be seen at more than 244 m. It was clear that an object was across the road at 104 m or 4 sec to crash, although the trailer was not visible. In the last condition, the reflective tape and additional side-marker lamps were used. At more than 244 m, the side-marker lamps were visible and, at 206 m, it was noticeable that they all moved as one unit, providing a good cue that the object was moving across the driver’s lane. The tape also became clearly visible at 130 m, and at 104 m it was clear that an object was across the road lane, although the body of the trailer could not be seen. In all cases, the tractor’s headlights and other front lighting gave the appearance of a truck traveling in the northbound lane toward the car. It was concluded that the truck that had passed the car before the crash exposed the driver to glare and would have attracted the attention of the driver away from the activity at the intersection as the truck involved in the crash began to pull into the intersection. After the first truck had passed, the driver would be in a right curve and should have noticed the headlights of another truck approaching. Although the center yellow, and later the red rearmost, side-marker lamps would be visible to the car driver, they would be faint and a long distance from the tractor: about 6 and 12 m, respectively. They may not have been readily associated as a part of the trailer, but may have been seen as ambiguous lights in the distance or driveway reflectors.


332

The addition of retroreflective tape, additional side-marker lamps spaced about 1.5 m apart, and, especially, the combination of these methods provided greatly improved visibility and recognition of the trailer across the lane. The driver had little expectation of a trailer across the lane and was unable to see or recognize it in time to avoid it when it was provided only with the reflectors and side-marker lamps that were then required. The retroreflective tape along the side would have aided the driver substantially. The use of additional side-marker lamps, with or without the retroreflective tape, would have provided the driver with even better information that a continuous object was across the lane of travel (Figure 13.3). It would be expected that most drivers would infer that this was the truck’s trailer even though they could not see the body of the trailer. The use of the extra lighting and marking would have allowed the driver to avoid the crash or at least to reduce the speed and severity of the impact. Such aids to the visibility and recognition of trailers were readily available at the time of this crash and could have been implemented by the manufacturer or operator. Rear-End Collisions Rear-end collisions are about 25% of all crashes, approximately 27% of crashes causing injuries, and 5% of the fatalities (USDOT, 1998). It has been found that, in more than half of rear-end crashes, the struck vehicle was stopped or moving very slowly (Mortimer, 1981), which suggests that drivers do not perceive the speed of the lead car or perhaps the relative speed.

FIGURE 13.3 Driver’s view of turning truck at 76 m: (top left) standard lighting/marking; (top right) retroreflective tape; (bottom left)


333

added side-marker lamps; (bottom right) tape and added side-marker lamps on the trailer. Drivers are reasonably sensitive to changes in headway between their vehicle and a vehicle ahead of them. The basis for this performance is the change in the visual angle subtended by the leading vehicle at the eyes of the following driver. For small angles, this can be represented also as the change in the headway. The driver’s ability can then be described most intuitively by a simple Weber function, ∆H/H =0.12, in which H is the initial headway distance (Hoffman, 1968). This means that alert drivers can be expected to notice a change in headway of about 12% and infer that the gap between their vehicle and one in front has changed. Unless the headway is very small, the detection of change in headway is normally adequate to alert drivers to a potential need to pay attention to the vehicle ahead of them and to be ready to respond. If the initial headway is small and closure is detected, the driver may need to react immediately. However, in most rear-end crashes, the vehicles are not in a close-coupled state (Perchonok, 1972) so there is usually time after the change in headway has been detected. The question then is: why do the crashes occur? Having detected closure by a change in headway, the driver still must determine how much time remains to the crash to determine the appropriate response. The time remaining is a function of the headway and the relative speed. Although headway can be estimated moderately well, drivers are decidedly limited in their ability to detect relative speed. The major cue to spacing changes is the change in the visual angle of the rear of the vehicle. At night, in particular, this will be the angle at the driver’s eyes subtended by the rear lighting. The rate of change of this angle provides the cue to relative velocity. Research has shown that the threshold for detection of relative velocity is about 0.003 rad/sec (Hoffmann and Mortimer, 1996). Table 13.2 shows the relative speed, time to crash, and distance between the vehicles at the relative velocity threshold for a lead vehicle 5 ft wide or for a vehicle whose rear lamps are set 5 ft apart, if viewed in darkness.

TABLE 13.2 Relative Speed, Time, and Distance between Vehicles at Threshold of Relative Velocitya Relative Time to crash Distance to crash speed (mph) (km/h) (sec) (ft) (m) 70 113 60 97 50 80 40 64 30 48 20 32 a 0.003 rad/sec.

4.0 4.3 4.8 5.3 6.1 7.5

414 380 350 312 270 220

126 116 107 95 82 67


334

The amount of time remaining to crash after the threshold of detection of relative velocity is attained is inversely related to the relative speed but directly related to the distance between the vehicles (Table 13.2). It also needs to be remembered that the braking distance is directly related to the square of the initial speed. Consider the case of a car driver traveling at 113 km/h approaching a stopped car. If the driver has not slowed based on the cues provided by the change of visual angle or headway, the cue to relative velocity will be at threshold at about 126 m from the stopped car. Assume that the driver has a lag time of 1 sec in detecting the relative velocity. The driver will be 95 m from the stopped car. The driver will still need to process the information, decide on a response, and implement it, so assume that in another 1.5 sec the brakes are applied to achieve a deceleration of 0.8 g, assuming the pavement is dry. The distance remaining is 48 m. However, the braking distance to stop is about 63 m, so a crash results. Consider the same situation with a relative speed of 48 km/h. Braking again is assumed to start at 2.5 sec after the threshold of relative velocity detection is achieved. The distance then remaining to crash is 49 m. To stop from 48 km/h at 0.8 g requires 11 m, so the crash is averted. The driver’s ability to scale the perceived relative velocity is limited to a few categorical values such as low, medium, and high, which are nevertheless adequate to allow the driver to determine the urgency of the response that is needed. The relatively poor ability of drivers in estimating relative speed results in a short time remaining to crash, especially at the higher relative speeds. When combined with the other tasks involved in driving, complex driving situations, glare, and other visual degrading conditions, drivers’ relative velocity detection thresholds can often be further degraded and increase the probability of rear-end crashes. Overtaking and Passing It is worth noting that the same process just described for detection of relative velocity in the approach to the rear of a vehicle applies to determining the time available to pass another vehicle when there is an oncoming vehicle. However, because the distance when a passing maneuver is initiated is necessarily much larger than that involved in approaching a vehicle from the rear, the threshold for relative velocity perception will normally not have been exceeded when the decision to pass must be made. This means that drivers have no way of estimating the speed of the approaching vehicle. As Rumar and Berggrund (1973) stated, “Drivers cannot estimate the speed of the oncoming car. Being in that vague position they assume that the speed of the oncoming car is the same as their own speed and consequently base their decision mainly on the estimated distance to the oncoming car.” Farber and Silver (1967), in their studies of overtaking behavior, reported that “the threshold passing distance adopted by drivers tends to remain constant regardless of oncoming car speed”—again confirming that the drivers were not able to assess the relative speed due to the small visual angle subtended by the oncoming car and its small, subthreshold rate of change at the large distances involved.


335

Expectancy Elsewhere in this chapter there is a discussion of driver positive guidance which has as a central tenet that systems should be designed to ensure that the expectations of drivers are met as much as possible. This will reduce ambiguities and reduce the time drivers need to make decisions. There are expectancies that are not easy to design in or out. An example would be a stalled vehicle in the travel lane of an interstate highway or a pedestrian on an unlighted rural road at night. Drivers do not expect either of these kinds of events. A driver approaching the stalled vehicle will initially assume that it is moving at normal highway speed. Even as the driver receives information about closure with the vehicle, the lack of expectancy that the vehicle is stationary will reduce the salience of the information as well as the driver’s attention to the vehicle, thus increasing the driver’s response time to the hazard. At night, if the stalled vehicle has an inoperative electrical system and the vehicle is unlighted, the hazard is extreme and almost inevitably will result in a crash. If the vehicle has operating tail lamps, the approaching driver will again assume that the stalled vehicle is moving and a crash is likely. Even the use of hazard flashers on a stopped vehicle, while aiding in detecting the vehicle and suggesting that it is stopped or moving at a speed less than the normal highway speed, is still an unexpected condition that may override the value of the hazard signal and delay the oncoming driver’s response. Few studies have tried to quantify the effect of a lack of expectancy on drivers’ reactions. One reason is that such studies are not easy to conduct because of the difficulty of introducing the unexpected condition without alerting the driver and the necessity of not creating an unsafe situation. In one early study (Roper and Howard, 1938), the observer drove the test car at night and was told that the task was to evaluate some aspect of the car such as the headlights. On the way to the test area, the driver encountered a pedestrian object in his lane. The distance driven from release of the accelerator to coming adjacent to the pedestrian was measured and taken as the detection-response distance in the “unexpected” condition. The task was repeated, but this time the observer knew that a pedestrian object would be encountered. These detection-response distances were taken as the observer’s response in the “expected” condition. It was found that the pedestrian was detected at about half the distance in the unexpected as in the expected condition. Some have assumed that this implies that drivers require twice the distance to detect an unexpected object as for one that is expected. That assumption is not appropriate, even in a situation similar to the one in which the data were obtained, for numerous reasons (Mortimer, 1996). However, there is no question that an unexpected event will require a longer time for the observer to respond, even in simple situations (Johansson and Rumar, 1971; Summala, 1981). In practice, the effect of a lack of expectancy will be difficult to quantify because of the paucity of current data. However, it is clear that drivers who are confronted with an unexpected situation will respond with great variability. For example, Summala (1981) found that some drivers needed up to 4 sec to respond by steering their car away from the side of the road when someone opened the driver’s door of a car on the road’s shoulder; the 50th percentile was about 2.5 sec. The extent of the variability will likely be greater in more complex tasks such as detecting an unexpected object in darkness. In one study of the detection of an unexpected stop sign at night (Olson, 1988), the ratios of the unexpected to expected response distances of the 10 subjects were between a low of


336

approximately 1:08 to a high of about 1:2.6; thus, the intersubject variation was so great that it would be inappropriate to take any fixed value to represent the effect of expectancy. Even though these subjects knew they were in a test and can be assumed to have been attentive to the task, the variation among the subjects was large. Familiarity with an environment can reduce the negative effects of expectancy if prior exposure has created a heightened awareness of an event occurring. A driver who often drives through a residential area and has noted children playing near the road has an expectation of seeing children in that area and should be more ready to respond than if a child appears near the road in an area where children were not seen before. However, familiarity with an area can also have a negative transfer effect. A driver who comes upon a truck-trailer that is being backed across the road where this had not been seen before will have a response lag due to lack of expecting such an event. As Deese (1959) noted, the observer’s expectancy is a function of prior experience with the task, and the level of expectancy of the event determines the observer’s level of vigilance and the probability of detection. Stress When a small amount of time is available for a response to try to avoid a crash, the individual will experience stress. Although a low level of stress can lead to boredom and poor performance, a moderate

TABLE 13.3 Change in Probability of Response Error Associated with Stress and Operator Skill Level Stress level

Change in probability of error Skilled Novice

Very low ×2 ×2 Optimum ×1 ×1 Moderately high ×2 ×4 Extremely high ×5 ×10 Source: Adapted from Swain and Guttman, 1983.

level is usually beneficial to arousal and to maintain performance on a task. A high level of stress degrades performance. The familiar U-shaped relationship between stress and performance defines the Yerkes-Dodson law (Welford, 1973). For example, as a driver nears an unguarded railroad crossing where a train was not previously seen and suddenly detects a train close and at high speed, the driver will almost certainly experience high stress. In such a situation, the driver will tend to sample the most salient cues, such as the approaching locomotive, and neglect other cues that may be important in reaching a decision on how to avoid a collision. Under high stress, fewer cues are sampled and attention narrows (Broadbent, 1978). That behavior can lead to errors in selecting a response and also to a lag time in responding. In the extreme situation, drivers may “freeze” on the controls and fail to respond at all or fail to modify a response. In many precrash situations, drivers will use


337

only the brakes to slow the vehicle, even when the stopping distance is insufficient and it may have been possible to steer around the hazard. Swain and Guttman (1983) developed a model to act as a guide of the effects of stress and operator level of experience on the increase in the probability of errors in routine tasks, which can include driving (Table 13.3). According to Table 13.3, novices are more affected by moderate or high levels of stress, as would be expected, because of lower spare information-processing capacity before the added workload imposed by stress. Although a good deal of research on the general topic of stress has been conducted, like expectancy, it is a difficult area to study and to generalize the results. Stress has an impact on working memory, attention, perceptual-motor skills, and communication skills (Stokes and Kite, 1999), but the studies provide little guidance on the extent of the effects in a given situation. In spite of this lack of guidance and quantification, the effects of stress on response-time lag and response errors should be given consideration. Response Time There is no such thing as a fixed response time for people. Nevertheless, statements that a driver’s response time is a fixed value, such as 0.75 or 2.5 sec, have often been made. Traffic and highway engineers, for example, do need a value of response time that they can incorporate into their everyday design decisions, such as determining stopping distance sight lines that drivers will need for various driving maneuvers. In the approach to a railroad grade crossing, adequate sight lines must be provided to enable the driver to detect and respond to a train in time to avoid a collision or to determine the distance that a stop sign needs to be visible to allow drivers sufficient time to come to a stop safely. Based on research, a value of 2.5 sec is often used for some of these purposes, but the complexity of deriving a useful figure of merit is exemplified by Olson et al. (1984), who found that other factors need to be considered. The detection and subsequent information-processing tasks involved have a great impact on the time needed to reach a decision and make a response. Detection of a relevant stimulus in traffic can very often be done quickly, such as in less than 1 sec; however, in some situations the time needed can be much longer due to weather, lighting, workload, and salience of the stimulus. The state of the driver or pedestrian is also a factor. Once the stimulus has been detected, the information processing, in terms of recognition of the stimulus and the decision-making task, can also vary greatly in time due to the complexity of the situation, ambiguity of the stimulus, expectancy, and stress. Response times to simple, easily detectable situations, such as a brake signal on a car ahead, can be about 0.8 to 1.2 sec. What if the driver of the following car is considering passing the car ahead and is looking in the outside rearview mirror when the brake signal appears? Then the time to detect the brake signal will be increased, perhaps by up to 2 sec. Detection time is the time from a stimulus being at threshold of detectability and its detection. A pedestrian on a straight, level road in daytime is detectable at a much greater distance than at night if illumination is only provided by headlamps. However, the pedestrian in the daytime may be detected after a longer time has elapsed from the threshold than the pedestrian at night would be. This may be because the pedestrian will present as a stimulus at a large distance, but not be detected as a relevant stimulus that gets the attention of the driver until he is much closer. If the threshold is at


338

300 m from the pedestrian but he is detected at 200 m, the elapsed time for detection at 97 km/h is 3.4 sec. If, at night, the pedestrian is just detectable at 91 m and is detected at 61 m at 97 km/h, the elapsed time is 1.1 sec. The time remaining after detection is for recognizing the object, deciding on a response, and executing the response. Time to move the foot to the brake or to move the steering wheel in an evasive maneuver is about 0.5 sec in a quick response. There is also a short lag time for the brakes to take hold or the car to yaw. Although brake system lags are less than 0.25 sec in automobiles, they are about 0.5 sec in large trucks with air brakes. In measuring the time for drivers to swerve out of the way of the driver’s door being opened of a car stopped on a road’s shoulder (Summala, 1981), the mean time to initiate a steering response was about 1.5 sec while the longest time to execute the swerve was about 4 sec. This means that, even in a clear-cut situation in which some expectation exists that a driver may open the door of the car stopped on the shoulder, as much as 4 sec may be needed to accomplish a lateral movement of only about 25 cm. Another study of unobtrusive measurements of the response times of drivers (Triggs and Harris, 1982) found 85th-percentile times of 1.2 to 3.6 sec in a variety of situations. Driver response times of up to 4 sec are not necessarily unusual and need not indicate that the driver’s performance is impaired or shows a lack of alertness. Each situation must be evaluated in consideration of the variables that may affect performance. Motorcycles In 2000, about 2862 occupants of motorcycles were fatally injured and about 58,000 were injured in crashes in the U.S. (USDOT, 2001). The fatality rate of motorcyclists is estimated at about 27 per 100 million miles or about 21 times the rate for occupants of passenger cars. More than any other group of drivers, motorcyclists involved in 38% of fatal crashes had consumed some alcohol; in 27% of their crashes, the operator’s bloodalcohol concentration was at or more than 0.10%. Because of the nature of the vehicle, many motorcycle crashes are not reported to the police and do not enter the official records. This will be most true of property damage and minor injury crashes when no other person or vehicle is involved. In spite of that, almost half of motorcycle crashes involving injuries were reported to be single-vehicle crashes. This is much greater than the 15% of automobiles in single-vehicle crashes and illustrates the problems of motorcycle riders in maintaining safe control of the machines. In crashes involving another vehicle, most of which occur at intersections, the driver of the other vehicle was reported to have turned or crossed in front of the motorcycle in about 70% of the cases (Hurt et al, 1981). Most often the motorcyclist was proceeding straight and the other vehicle driver, approaching from the opposite direction, was turning left across the motorcyclist’s path. The drivers of the other vehicles often stated that they did not see the motorcycle in time. It is apparent that at least three basic factors contribute to motorcycle crashes: motorcyclists’ use of alcohol, lack of ability to control the motorcycle; and violation of the right of way by other drivers. The effects of alcohol and other drugs are described elsewhere in this chapter. Lateral (steering) control of motorcycles is much more complex than that of four-wheeled vehicles. Braking control is also much more complex and most motorcyclists are unable


339

to achieve maximum braking performance (Ecker et al., 2000), especially in emergency situations (Hurt et al., 1981). The reason for this poor performance is the design of most motorcycles’ brake systems, which have separate brake controls for the front and rear

FIGURE 13.4 Visual angle of car and motorcycle by distance. wheels. The right foot can operate a brake lever for the rear wheel and the right hand can operate a lever on the handlebars that operates the brake on the front wheel. To exert maximum braking, the rider must activate both these controls and modulate them to achieve braking forces that do not quite lock either wheel. Because of the convenience of operation of the foot brake, riders favor it in everyday braking, and in emergencies they often fail to also use the front brake. Use of the foot-operated rear brake permits a maximum of only about 40% of the braking force of which the motorcycle is capable to be obtained. A brake system in which a single control operates the rear and front wheel brakes would lead to greatly improved braking performance (Mortimer, 1998), especially if combined with antilocking capability. Another factor in motorcycle crashes concerns the detectability of motorcycles and the ability of other motorists to judge their distance and speed. In many countries as well as states in the U.S., motorcycles are required to have the headlight on at all times, which should help to allow the motorcycle to be more readily detected by others in daytime. However, the narrow aspect of the motorcycle and its operator makes it less easy to estimate its distance, to notice a change in distance, and to estimate its relative speed. Figure 13.4 shows how the visual angle, based on the widths of a car and a motorcycle, changes with distance from the observer. A car at 1000 ft (304 m) subtends a similar angle as a motorcycle at 300 ft (91 m), which would lead to overestimations of the distance of a motorcycle. Other factors, such as size constancy and the surroundings of the motorcycle, help to correct partly for this error. The sensitivity of other drivers to changes in distance of the motorcycle will also be less than to the car, and their ability to estimate relative velocity will reach threshold when the motorcycle is closer to them than


340

the car for the same actual relative speed. At night, compared with daytime, those estimations will be even more affected if the motorcycle is depicted only by its headlight(s). The performance of other drivers in turning and crossing across the path of a motorcycle can be made safer if the motorcycle has auxiliary lamps spaced as far apart as practicable, such as at least the width of the handlebars, on each side of the headlamp (Mortimer and Schuldt, 1980). Recently, some manufacturers of automobiles and light trucks have begun incorporating headlights that are lighted in daytime. As a result, it seems likely that the headlights of motorcycles will become a less prominent signal in daytime so that they will be more difficult to detect among those other lights, as is already the case in darkness. This will make the use of auxiliary lamps important also in daytime. Railroad Crossings Collisions between motor vehicles and trains have been significantly reduced in the past 20 years in the U.S. In 1980, about 739 fatalities in collisions with trains occurred and in 2000 there were about 400 fatalities. Motor vehicle collisions with trains are very severe in that fatalities occur in about 10% of them compared with about 0.3% in other crashes. Some caution needs to be exercised in interpreting crashes reported to have occurred at grade crossings because only about 25% involve a train. The others are mainly rear-end crashes involving vehicles that are stopping or stopped at the crossing. The two basic human factors considerations in evaluating vehicle-train collisions or those involving pedestrians or cyclists and trains are (1) those that affect the visibility or audibility of the train, and (2) the warning signs and signals at the crossing. The road user must take the initiative to avoid the collision with the train because the only options available to the train engineer are to provide a timely warning by lights and the train’s horn. In most cases, there is no reasonable opportunity to slow the train to avoid a crash because of its poor braking deceleration (Bentley, 2002) compared with road vehicles. Visibility Factors In the U.S., grade crossings are announced by a “railroad crossing ahead” sign, placed about 229 m before the crossing in rural areas and about 76 m in urban areas. The sign provides no indication of the level of protection at the crossing. In other countries, such as Australia and in Europe, the advance warning sign indicates if the crossing is passive or has some form of active signals or gates. This advance information can be used by drivers to plan their speed as they approach the crossing. At unguarded (passive) crossings, drivers must be able to see along the tracks to detect the train in time to stop or to cross before it reaches the crossing. This would also be desirable when crossing signals and even gates are present at active crossings because, even if the signal indicates that a train is in the vicinity, drivers may cross if they do not see the train. Minimum lines of sight have been suggested based on the speeds of the road vehicles and trains (USDOT, 1978). These lines of sight must take account of obstructions such as buildings and vegetation, although the actual lines of sight may be further impeded by parked vehicles on the road and train cars in sidings and adjacent tracks. When minimum lines of sight cannot be achieved because of fixed structures or the road or track


341

alignment, or if the amount of road and train traffic exceeds some adopted criterion, active warnings must be employed to alert drivers to trains. These normally consist of flashing red signal lamps mounted at the side of the road before the tracks and sometimes on overhead gantries. Newer lamps produce about 3000 cd toward oncoming drivers. Most recently, conventional tungsten filament lamps are being replaced with lightemitting diodes, which offer greater reliability and resistance to vandalism. Although the signal lamps are usually easy to detect at night, depending on the other lights and their background, they are sometimes difficult to detect in daytime, especially when the sun is at a low angle behind the lamp, causing glare, or when it illuminates the lamp directly and dilutes the signal (Mortimer, 1991). Locomotives have lighting equipment in the form of headlights and sometimes ditch lights; those used in suburban service in the U.S. sometimes also have a strobe lamp above the cab. One or two headlights are mounted high up on the centerline of the locomotive and are aimed straight ahead to illuminate the tracks. The lower of the two headlamps may oscillate up to 30° to either side on older locomotives. The central intensity of these lamps is about 200,000 cd, with the major intensities concentrated in a narrow beam of about 10°. The ditch lamps—one on each side, mounted low—are aimed to the sides by up to 45°. A purpose of these lamps is to provide a stronger light signal to motorists approaching the tracks who may be out of the beam of the main headlamps. However, the triangular pattern of the main headlamps and the ditch lamps should aid drivers in assessing the distance and speed of the train if all three are visible. Auditory Factors The primary auditory signal to warn of the approach of a train is its horn. The Federal Railroad Administration requires the horn to have a minimum intensity of 96 dB at 100 ft (30 m). Many train horns provide about 115 dB. Because the intensity of the horn is attenuated by about 6 dB for each doubling of the distance, the horn may provide about 91 dB at 488 m. Although that is a good signal strength for a pedestrian or bicyclist, it is inadequate for the driver of a closed car because of the 20- to 30-dB reduction in intensity of the sound inside the vehicle. When the train is about 183 m distant, the horn may be detectable inside a quiet car, but that allows only about 6 sec before the arrival of a 113km/h train. Locomotives also operate a bell when approaching crossings, stations, or work crews on the tracks. The bell provides 70 to 80 dB (at 30 m) and is a relatively weak signal. There are also bells at crossings that have active signals with or without gates. They have an intensity of about 65 dB (at 30 m) and are intended to warn pedestrians or cyclists and can be a warning to drivers if a window of the vehicle is open (Mortimer, 1994). The audibility of these warning signals is affected by any hearing loss of the observer; such losses usually start beyond about 30 years of age and, by age 50, may be about 5 dB in the frequency range of 500 to 1000 Hz. Train horns have fundamental frequencies in the 300- to 600-Hz range. Using the low end of the sound spectrum ensures that the sound of the horn will travel a maximum distance. The ability to hear the train horn will also be affected by the frequency composition and intensity of the sounds surrounding the observer. Masking of the horn in a vehicle can be due to the radio, speech, wind, and engine and road noises. Low-frequency sounds can mask those of higher frequency but


342

they have little effect on those of lesser frequency. In general, an auditory signal such as a train horn or the siren of an emergency vehicle needs to be about 10 dB more intense than the sound level in the vehicle to be reasonably detectable. To ensure that train horns are detectable, drivers should open the windows of the vehicle when approaching a crossing; otherwise, auditory signals are not much of an advance warning of a train. Other Factors Affecting Driver and Pedestrian Behavior at Railroad Crossings The lines of sight along the tracks available to a driver determine maximum potential visibility. Actual visibility is affected by vertical and horizontal alignment of the road and tracks, ambient lighting, weather factors, artificial lighting, other light sources that can be distractions or mask the crossing’s or locomotive’s signals, color of locomotives and trains, and the background against which they are seen. The behavior of drivers approaching crossings has been studied and indications are that drivers often fail to scan the tracks for trains or look in one direction only, sometimes ignore crossing signals, and have a low expectation that a train may be approaching. Drivers are at particular risk at crossings with more than one track. After waiting for a train to pass through the crossing, drivers have pulled onto the tracks while the warning signals are still activated only to be hit by another train, usually moving in the opposite direction. Signals that warn that more than one train is approaching and show the direction of the train are in use in some European locations and provide important information. Passengers and pedestrians are also at risk when crossing tracks at road crossings or in stations because they misjudge the speed of trains. Passengers in stations have been known to assume that a train approaching the station is stopping. They walk across the tracks to reach another platform only to find, too late, that the train is an express that is not stopping at that station. For this reason, signals are in place at some stations that warn passengers of the approach of a train and alert them not to cross the tracks. The Law and Visibility The Uniform Vehicle Codes of most states in the U.S. have statements concerning the visibility of marker lamps at night, signal lamps in the daytime, and the ability of high and low beams of headlamps to reveal persons and vehicles at certain distances (Chapter 316, sections 237, 238, etc.). For example, the Georgia vehicle code states that the turn signals of vehicles more than 80 in. (203 cm) wide should be visible in normal sunlight at 500 ft (152 m) to the front and rear. It also states that the high-beam headlight should reveal persons and vehicles at 450 ft (137 m) and on low beam at 150 ft (46 m); the low beam “shall be deemed to avoid glare at all times regardless of road contour and loading.” Can the law dictate nighttime visibility? Obviously it cannot. Visibility depends on many factors such as those already discussed. Pedestrians and other low-reflectance objects are often not detectable at the distances specified and there is some exposure to and effect of glare whenever vehicles oppose each other at night, especially on curves and hills.


343

These statements in the vehicle codes are used to support legal arguments that an object should have been detectable at such distances, thus possibly allowing time for a driver to take avoidance action. In that case, the human factors expert must be able to explain the factors that affect visibility that may dispute the codes. An additional attack can come in the form of the “assured clear distance ahead” rule, which holds the driver responsible for ensuring that a speed is maintained that allows obstacles to be seen and avoided. Inherent in that rule is the assumption that drivers are aware of available visibility of hypothetical objects. However, it has been repeatedly demonstrated that drivers routinely “overdrive” their headlights. Drivers tend to base the maximum speed they drive upon the ability to control the vehicle on the road, which is a very different task from being able to see all manner of objects in time to avoid them. The preview distance required for tracking the road lane can be as short as 1 sec of travel time, but that is quite inadequate to avoid an obstacle. At 2 sec of preview distance at 97 km/h, drivers feel quite comfortable in lane tracking, having about 53 m of road in view, but they would not be able to stop before an object that is detectable at that distance. The incongruence between the road-tracking task and the object-detection task is ignored in the assured clear distance ahead rule, which fails to acknowledge normal driver behavior. Suggestions for alleviation of some of these problems have been made by Liebowitz et al. (1998), who have indicated that responsibility for avoiding obscure hazards at night should not rest solely on the driver but also on other road users. Drivers should be educated about their limitations in night driving, make more use of the high beams, and reduce speeds, and other road users should take measures to enhance their own visibility, such as use of retroreflective materials and lighting. 13.4 Pedestrian and Bicycle Crashes The human factors analysis of pedestrian and bicycle collisions with motor vehicles, although similar in many ways to the assessment of vehicle-to-vehicle crashes, differs in at least two important aspects. One difference is that the characteristics of pedestrian and bicycle crashes have been documented in great detail, and excellent taxonomies of crash types exist for both. The existence of crash types provides a normative standard against which to compare the circumstances of any crash analyzed. Once an analyst has concluded that a pedestrian or bicycle crash fits a known type, missing details can often be inferred from knowledge of the modal values associated with the specific type. This is fortunate because the other difference is that police crash reports on pedestrian and bicycle crashes tend to be less complete and insightful than those for crashes involving only motor vehicles. This is not due to any dereliction of duty on the part of the investigating officers. Rather, it is an outgrowth of the lack of in-depth training in the investigation of pedestrian and bicycle crashes that is given to police officers. In light of the foregoing, an understanding of the etiology of pedestrian and bicycle crash types and the limitations of the police crash report (PCR) is essential to the successful human factors analysis of these events.


344

Crash Types Pedestrian and bicycle crashes have likely been studied for as long as there has been an interest in traffic safety. However, knowledge of the causes of these crashes took a quantum jump in the 1970s with the publication of studies by Snyder and Knoblauch (1971) and Cross and Fisher (1977) that identified specific crash types and their main causal factors for pedestrians and bicyclists, respectively. A taxonomy of crash types is a useful analytical tool in the assessment of pedestrian or bicycle crashes. The development of effective countermeasures to crashes is assisted by the ability to disaggregate these collisions into occurrences with similar causal elements, including human factors. In addition to aiding countermeasure development, the existence of crash types can also be a useful tool in the analysis of individual crash events, whether for forensic or research purposes. We rarely have a complete picture of a crash even when specialists have investigated it. Reliable, trained eyewitnesses are present at few crashes. Crash scenes are often disturbed before investigators arrive, frequently as part of

FIGURE 13.5 Function/event sequence as defined by Snyder and Knoblauch. (Snyder, M.H. and Knoblauch, R.L., 1971. Operations Research, Inc., final report U.S. Department of Transportation, National Highway Traffic Safety Administration, report no. DOT HS800 403.) the efforts to render assistance to victims. These and other factors produce gaps in the information available about the causes of the collision. Estimating what may have occurred in the absence of concrete information is greatly facilitated when a crash closely fits one of the predefined and well-researched crash types. This author has successfully argued as an expert witness for the most likely sequence of events in a crash at issue


345

based on a good fit of that crash’s circumstances to one of the defined pedestrian or bicycle crash types. The Function/Event Sequence In order to develop their pedestrian crash types, Snyder and Knoblauch (1971) defined a model of how crashes are generated. The model (shown in Figure 13.5) consists of the following sequence of functions or events that drivers and pedestrians perform in the traffic environment: • Search—scanning the environment for threats • Detection—determining that something is there • Evaluation—assessing whether the detected object is a potential threat • Decision—determining what to do to avoid a collision • Action—executing the selected avoidance maneuver (human and/or vehicle as appropriate) The sequence begins when the driver and pedestrian or bicyclist commence a collision course, thus making a crash inevitable unless some avoidance is taken. If either party to a potential crash performs the sequence successfully, the crash is avoided. Thus, by definition, both parties must suffer a failure of their portion of the sequence for a crash to occur. Cross and Fisher (1977) adopted a virtually identical model in their work to define bicycle crash types. It is also important to note in Figure 13.5 that this sequence is a dependent sequential process. When a participant fails to complete a function successfully (a “no” exit from any of the functions in Figure 13.5), all subsequent functions must also fail. For example, if a motorist fails to detect a bicyclist, the motorist cannot properly perform evaluation, decision, and action. In this example, avoidance of a crash will therefore depend solely on the successful performance of the sequence by the bicyclist. Precipitating Factors The functions in the sequence are generic activities. For example, “search” is actually the process of information gathering and can be accomplished through visual, auditory, or tactile cues. Detection can mean actually seeing the threat or successfully receiving a warning from an artificial detection system such as a radar collision warning device. As part of the process of defining specific crash types, it is necessary to describe how each function failure occurred. For example, a search for threats may never have been performed or the search undertaken may have been inadequate, e.g., because it was made only in one direction. Snyder and Knoblauch (1971) called these descriptions of the function failures precipitating factors. The primary precipitating factor for each party was the one that occurred earliest in the function/event sequence and made completion of the sequence impossible.


346

FIGURE 13.6 Depiction of the backing pedestrian crash type. Defining Crash Types By examining recurring clusters of precipitating factors, Snyder and Knoblauch (1971) and Cross and Fisher (1977) were able to define specific crash types for pedestrian and bicycle crashes with motor vehicles, respectively. Because precipitating factors can be viewed as behavioral errors, whether willful or unintended, the resulting crash type taxonomy is a valuable human factors tool. It is interesting to note that none of the defined crash types for pedestrians or bicyclists is based on the crash-involved population, although some groups, e.g., children and the elderly, are over-represented in certain types of events. Some crash types are characterized by the particular action or maneuver taken by one of the parties. The “backing” type for pedestrians depicted in Figure 13.6 is an example. In this type, the motorist embarks upon a maneuver (backing up), on the street or in a parking lot, that has significant task demands and makes it difficult to conduct a successful search for a pedestrian. The roof pillars of the car may block rearward vision, thereby preventing detection even though a search is performed. Alternatively, the driver may back up without ever scanning for pedestrian threats. The pedestrian in this crash type typically fails to search for a driver in the backing vehicle or assumes that he or she has been seen and that the driver will yield. Other crash types represent combined clusters of location and a particular precipitating factor. The “dartout” type for pedestrians and the analogous “midblock rideout” involving bicyclists occur at non-intersection locations and are precipitated by the pedestrian or bicyclist presenting a short time exposure to the motorist. The driver cannot complete the function/event sequence successfully because time is too short once the pedestrian or bicyclist becomes a threat. The short time exposure can be generated because the victim was running or came out between parked cars that thwarted an effective search, as shown in the depictions in Figure 13.7. It can also occur when a


347

pedestrian or bicyclist on a sidewalk and apparently not in the act of entering the street (e.g., a pedestrian with back to traffic) makes a sudden movement into the roadway. This action invalidates the motorist’s evaluation that the pedestrian or bicyclist is not a threat, but leaves inadequate time for the successful completion of the function/event sequence.

FIGURE 13.7 Depictions of the dartout pedestrian crash type and the midblock rideout bicycle crash type . A third group of crash types is characterized by unique but not necessarily infrequent circumstances. For example, Snyder and Knoblauch (1971) identified a type involving ice cream trucks. As shown in Figure 13.8, the typical situation in this “ice cream/vendor truck-related” type involves a young child leaving an ice cream truck after making a purchase and crossing in front of the truck. The child is concentrating on the ice cream and fails to search. The large truck screens the child from view, thereby preventing the motorist from detecting the child. This is a special variant of the dartout type discussed


348

previously. The important difference is that the involvement of the ice cream vending vehicle provides a focus for countermeasures in addition to efforts to instill correct behaviors in the pedestrian (stop at the edge of the ice cream truck and look left-right-left for oncoming traffic) and driver (slow down, stop if necessary, and search under and in front of the truck). Snyder and Knoblauch (1971) only examined urban pedestrian crashes, whereas Cross and Fisher (1977) included suburban and rural collisions in their study. Additional research by Knoblauch (1977) and Knoblauch and colleagues (1976) augmented the pedestrian crash types with a few additional entries based on crashes occurring in suburban and rural areas and on freeways. The pedestrian and bicycle taxonomies were reexamined by Hunter et al. (1996) to see if they had changed markedly in the almost 20 years since their development. The conclusion was that the basic classification continued to be valid, although some small shifts had taken place in frequency of occurrence and involved populations. Also, some new subtypes were evident, mostly as a result of emerging societal trends, e.g., the use of in-line skates in the roadway.

FIGURE 13.8 Schematic depiction of the ice cream/vendor truck-related pedestrian crash type. The proved utility of crash types in analyzing crash problems, developing countermeasures, and evaluating interventions led to the development of a computer program to assist in the process. This software, entitled Pedestrian and Bicycle Crash Analysis Tool (PBCAT), is readily available and is gaining widespread use (Harkey et al., 1999). PBCAT is particularly helpful to a community in analyzing the nature of its pedestrian and bicycle crash problems. It can also assist in the investigation of a specific crash occurrence by helping the analyst determine the crash type rapidly and accurately. In turn, knowing the crash type provides a wealth of information to guide a more in-depth examination for causes and fault.


349

Predisposing Factors From the foregoing, it is clear that crashes result when a pedestrian or bicyclist and the striking motorist fail to complete the function/event sequence successfully. The fact that these failures result in specific crash types that appear quite universal and recurring is of interest and can help to estimate missing crash details. However, when a particular collision is investigated in order to assign blame or as a guide to remedial actions, crash types and their associated precipitating factors are not sufficient. We need to know why each party failed. Why did that driver and pedestrian or bicyclist collide at that moment when tens, hundreds, or thousands of others under the same circumstances did not? The needed information concerns the predisposing factors for the crash—those situations or conditions that made the errors by the parties in the crash inevitable or at least highly likely. Predisposing factors can arise from the vehicle, the environment, or the condition/state of the involved humans. Few vehicle factors predispose pedestrian and bicycle crashes. In the example of the backing crash type discussed earlier, an extra wide roof pillar or a vehicle designed with poor rear visibility could predispose a detection failure on the part of a driver. A bicycle with a wobbly front wheel might prompt the bicyclist to look down when entering the roadway, thereby predisposing a search failure that could lead to a rideout bicycle crash. Overall, however, the vehicle, whether motor vehicle or bicycle, is not a major cause of pedestrian and bicycle crashes. The environment, on the other hand, predisposes many pedestrian and bicycle crashes. Poor roadway design, road surfaces in disrepair, and, particularly, visual screens are all frequent contributors to crashes. Safety researchers have long known that a visual screen is a major predisposing factor to pedestrian and bicycle crashes. When a driver and pedestrian or bicyclist have their views of each other fully or partially obstructed, detection becomes difficult or impossible even when search is attempted, thereby making a crash more likely. The dartout and rideout types discussed earlier as well as the ice cream/vendor truck-related involve a visual screen. A commercial bus at a bus stop can produce the same dynamics if a pedestrian crosses in front of it. Hedges, road signs, and street furniture are other examples of visual screens that can predispose pedestrian and bicycle crashes. It is worth noting that visual screens do not need to be complete or stationary to predispose a crash. Even when significant portions of a pedestrian can be seen, the interruption of the anthropomorphic shape cue presented to a driver can lead to inconspicuity or improper evaluations after the driver detects something. Research has clearly indicated that the perception of a human-like image greatly enhances the recognition of pedestrians at night (Blomberg et al., 1981). The state or condition of the humans involved in the crash can be a strong predisposing factor. Section 13.6 of this chapter presents significant detail on the effects of alcohol and drugs on human performance. It is well known that impaired drivers run a greatly increased crash risk. It is perhaps less well accepted that pedestrians’ and bicyclists’ use of impairing substances, particularly high doses of alcohol, can also predispose their crash involvement. For example, the relative risk of a pedestrian crash as a function of blood alcohol concentration (BAC) increases according to a function that is almost the same shape as that for motor vehicle drivers except at a higher BAC (Blomberg et al., 1979). This is intuitively reasonable because walking is a less complex


350

task than driving. Although no definitive study has been conducted, evidence also exists that the use of impairing substances by bicyclists can be a significant causal factor in their crashes. Other human conditions and circumstances that can predispose crashes include fatigue, anger, inexperience, reduced visual acuity, and impaired hearing. In addition, a host of pathological conditions ranging from a simple cold to more serious illnesses can interfere with psychomotor performance. Decisions made prior to entering the function/event sequence can also have a profound effect on crash likelihood and therefore must be considered as predisposing factors. The most obvious of these are the speed at which a person chooses to drive, ride, or walk and the route chosen for the trip. With respect to speed, for example, consider a motorist speeding at 50 mph in a 25-mph neighborhood. Everything happens twice as fast for that individual. Thus, when the vehicle is on a collision path with a pedestrian or bicyclist and the function/event sequence begins, that speeding driver has half as much time to execute the functions as would a motorist traveling at the speed limit. The choice of the route to travel can also profoundly affect the likelihood of completing the function/ event sequence successfully and thereby avoiding a crash. Consider, for example, a pedestrian who chooses a route without sidewalks and street lighting at night when an alternative path that is lighted and has a sidewalk is available. That pedestrian is greatly increasing his or her workload in performing the function/ event sequence as well as the task of any motorists encountered. Conspicuity Search and detection are the first two functions in the function/event sequence. Obviously, if a driver fails to search for pedestrians and bicyclists or vice versa, the sequence will fail and a crash will become more likely. Even when a potentially effective search is performed, however, detection may fail if the threat is not conspicuous. Unlike visibility, which is the physical property of an object being viewed, conspicuity deals with the properties of the target and the characteristics of the viewer. Humans tend to exclude things for which they are not looking and be more attuned to detecting things they are expecting or actively seeking. Conspicuity is dealt with here in some detail because it is a major predisposing factor to pedestrian and bicycle crashes and, thus, must be an integral part of the human factors assessment of these crashes. Conspicuity can be simply defined as the attention-getting value of a particular object to a specific viewer. A conspicuous object is one with a high likelihood of being noticed by a viewer even if the viewer is not specifically searching for it. Conspicuity is, in essence, the opposite of camouflage. The objective of camouflage is to reduce the chance that a viewer will detect the treated item. The goal of conspicuity enhancement is to maximize the chance that the particular item will be noticed. Conspicuity is not the same as visibility, although visibility is a necessary condition for conspicuity. An object is visible if it emits sufficient light energy to stimulate a viewer’s eyes (i.e., is above the visual threshold). That object becomes conspicuous if a viewer notices it. For example, a white bed sheet against a snow bank in daylight is highly visible but likely not conspicuous. A black sheet under the same viewing circumstances will emit less light energy but be much more conspicuous because of its contrast with the background.


351

Different colors and varying levels of brightness are two ways to generate contrast and improve conspicuity. Three types of inconspicuity are potentially relevant to pedestrian and bicycle crashes (Blomberg et al, 1984): • Type I—subthreshold targets. When the object in question is at or below the visual threshold for detection by a human, conspicuity is almost impossible and safety will be compromised. A countermeasure to a subthreshold situation would be to increase the level of visual signal that the item produces, e.g., by adding a light to it. For example, a pedestrian carrying a flashlight or a bicycle equipped with a flashing light at night becomes considerably easier to see because the active light source adds visibility. • Type II—suprathreshold targets not seen. When a pedestrian or bicyclist is above the visual threshold but is still not detected, a camouflage or inconspicuity situation exists. This can occur, for example, when the color of a pedestrian’s clothing does not create sufficient contrast with the environment or because a bicyclist is in an unusual or unexpected location, e.g., riding against traffic. Typical physical countermeasures for this type of inconspicuity involve wearing clothing of a strikingly different color from the surrounding environment or treating clothes or the bicycle with high-visibility materials (fluorescent for daytime, dawn, and dusk or retroreflective for night-time). Fire engines are painted bright red or fluorescent lime green to avoid this type of inconspicuity. Behavioral approaches such as promoting bicycling on the right side of the road with the flow of traffic can also help. • Type III—concealed targets. Conspicuity can be compromised by a visual screen that blocks an otherwise potentially detectable object from view. Hedges, mailboxes, or parked cars, as in the dartout and rideout crash types, are common visual screens. Removing the visual screen can cure this type of inconspicuity as can increasing the apparent size of the pedestrian or bicyclist through the attachment of a flag or other extension that can be seen above or around the visual screen. The Police Crash Report The foregoing structure for the categorization and analysis of pedestrian and bicycle crashes is extremely helpful in understanding root crash causes and in guiding an analyst’s search for the specific predisposing factors of any particular crash. Unfortunately, police crash reports (PCRs) are not typically formatted to collect the type of information needed to reach a confident pedestrian or bicycle crash type decision. This is particularly true of the computerized crash archives maintained by all states and some local jurisdictions. For example, information on the presence of visual screens such as ice cream trucks, commercial buses, and hedges is not systematically captured. This pertinent information may appear in the investigating officer’s narrative or on a crash diagram, but these PCR elements are typically not digitized and stored in crash archives. Anyone who has worked with police reports knows that their quality and comprehensiveness varies from extremely sketchy to moderately extensive. They generally focus on describing what happened and determining legal fault rather than assessing the true causal elements of the crash. This is not intended as a criticism of police crash reporting. In an era of reduced budgets and under the pressures of a litigious


352

society, police officers do the best they can. It just is not usually enough to uncover fully the relevant human factors issues in a particular crash. Part of the problem can certainly be traced to the minimal (often nonexistent) training that police officers receive in pedestrian and bicycle crash causation. A better understanding of pedestrian and bicycle crash types almost surely would result in more insightful narratives and diagrams on the police report, if time pressures did not prevent the officer from examining and recording the additional factors. The major area that needs attention to improve the situation for the human factors analyst relates to predisposing factors. Detailed predisposing factor information is rarely available on a police report. This may be because the report was not prepared at the scene, because the investigating officer did not have enough time to examine the scene thoroughly, or simply because the officer was not sufficiently sensitive to the importance of the information. In spite of any of their shortcomings, police reports are usually a primary source document for any human factors investigation or reconstruction of a pedestrian or bicycle crash. It is therefore important to understand the strengths and limitations of these documents. It is this author’s judgment from reviewing thousands of pedestrian and bicycle police reports that superior report quality is associated with careful completion of all preceded information, inclusion of a detailed narrative describing the extent of the investigation and pertinent particulars of the crash generation sequence, and provision of a scene diagram with relevant details and little or no extraneous elements. A good police report can be an invaluable aid to the human factors analysis of a crash. On the other hand, a poor PCR can render some information irretrievable and/or necessitate significant additional data gathering in order to determine what happened. Fundamentals of Pedestrian and Bicycle Crash Analysis Crash types, precipitating factors, and predisposing factors provide an ideal human factors framework in which to analyze pedestrian and bicycle crashes. Many such analyses are conducted in the context of a legal action. Therefore, before turning to the specific analytical steps, it is important to understand the basic distinction between a legal and a causal analysis. Legal and Causal Analyses The vehicle and traffic law (VTL) in all states and many municipalities prescribes and proscribes behaviors for drivers, pedestrians, and bicyclists. For example, pedestrians must walk facing traffic while motorists and bicyclists must drive with the flow of traffic, and pedestrians must not leave a place of safety and move suddenly into the path of a vehicle that is so close as to preclude the driver from taking appropriate evasive action (i.e., successfully completing the function/event sequence). Law enforcement officers and attorneys involved in litigation arising from a crash are often most concerned with violations of the VTL that occurred as part of a pedestrian or bicycle crash sequence. That is their job. It is important to realize, however, that a violation of the VTL per se is not necessarily a causal factor in the crash. It may not even be a predisposing factor at all. Assume a crash in which a pedestrian is crossing in a


353

crosswalk at an unsignalized location on a street with two or more lanes in each direction. A car in an inside lane (near the sidewalk) stops to yield to the pedestrian as required by the VTL in almost all jurisdictions. A car in the center (far) lane overtakes the stopped car and strikes the pedestrian. The driver of the striking vehicle never sees the pedestrian because he is screened from view by the car that stopped to yield to him. This common crash type, known as a “multiple threat,” is depicted in Figure 13.9. With respect to the VTL, the pedestrian and the car that stops for him in the multiple threat example shown in Figure 13.9 have not committed a violation. The striking driver has violated a widely held VTL provision modeled after Section 11–502(d) of the Uniform Vehicle Code, which says that “whenever any vehicle is stopped at a marked crosswalk or at any unmarked crosswalk at an intersection to permit a pedestrian to cross the roadway, the driver of any other vehicle approaching from the rear shall not overtake and pass such stopped vehicle” (National Committee on Uniform Traffic Laws and Ordinances, 2000). In an actual crash of this type, the striking driver would likely receive a citation for this violation. A causal analysis of this same situation in accordance with the function/event sequence described earlier yields a different picture. The multiple threat crash occurs because the pedestrian and the striking driver suffer a detection failure predisposed by the presence of the visual screen generated by the car that stopped to yield to the pedestrian. The striking driver cannot execute his or her duty to stop for a vehicle that is stopped to permit a pedestrian to cross because the critical presence of the pedestrian cannot be determined due to the visual screen. The pedestrian can avoid the crash by stopping at the edge of the stopped car and executing an effective search for overtaking traffic. This would also place the pedestrian in a stationary position in which he is protected by the stopped car and can be seen by the overtaking driver.

FIGURE 13.9 Depiction of the multiple threat pedestrian crash type. The multiple threat crash type is just one example of pedestrian and bicycle crashes in which violations of the VTL and human factors causes of the event are different. When


354

analyzing these crashes, therefore, it is important not to rush to a judgment on causation based on apparent law violations or citations issued. Steps in the Analytical Process The function/event sequence, defined crash types, and associated predisposing factors can be used to structure a systematic process for the human factors analysis of pedestrian and bicycle crashes with motor vehicles. The recommended steps are detailed below. 1. Determine as much factual information about the crash as possible. The police crash report is a good starting point for the basic data such as date, time, location, involved parties, and vehicle types. The narrative and diagram of the PCR should be examined and assessed in the context of the type of examination undertaken by the investigating officer. Crashes with more serious outcomes are typically investigated more thoroughly and will therefore produce more useful PCR data. Routine on-scene investigations produce somewhat less information. PCRs completed at the station house or mailed in by the involved parties are often incomplete and inaccurate. 2. Determine the likely crash type from the PCR information. The PBCAT software discussed earlier can provide assistance in determining the crash type from particulars that can be found on the PCR. If the circumstances can fit more than one type, prepare a list of the alternatives and the additional data needed to select among them. Knowing the crash type accomplishes several important things. First, it serves as a guide to the types of precipitating and predisposing factors that might have played a role in the crash. Second, it provides modal or most likely values for behaviors that are not specifically known from the available data. 3. Run through the function/event sequence for each involved party and enumerate possible failure points. This will help provide an understanding concerning why the crash occurred (actually, why the crash was not avoided). 4. Enumerate potential predisposing factors for each of the possible function/event sequence failures. This list will be an important guide for data gathering. It can be used to: structure observations at the crash scene, interviews or depositions with involved parties, witnesses, and the investigating officer, and any manual or computerized reconstructions attempted. When examining predisposing factors, pay particular attention to factors such as conspicuity and visual screens, which can influence or interfere with search and detection, because they are fundamental to safety. Do not forget to consider environmental and temporal conditions such as the position of the sun and any resulting glare and weather. Be sure to determine if any post-crash changes may have taken place at the crash scene. 5. Gather as much data as possible. A thorough human factors pedestrian or bicycle crash analysis typically requires a visit to the crash scene, preferably under weather and light conditions similar to those at the time of the crash. Look for visual screens or visual clutter that might have prevented detection even if all involved had searched. It is important to examine the scene from the perspectives of all involved parties. Try to look through their eyes and determine what they may have been seeing and doing as they proceeded through the function/event sequence. Access immediate post-crash medical treatment records for involved parties to look for possible impairing substance use or pathology that might have resulted in a participant’s reduced ability to perform.


355

Examine the roadway to see if its condition may have diverted the attention of one or both parties. For example, a broken or excessively high curb may prompt a pedestrian, particularly a senior citizen, to look down to avoid tripping and thus interrupt an adequate search for cars. Look at the striking vehicle in person or in police photos to determine if its design (e.g., extra wide roof pillars) or condition (e.g., dirty windshield) may have interfered with detection. If necessary, find an exemplar that matches the striking vehicle and examine it. Interview the participants (or request that specific questions be asked in depositions) to inquire how they handled each function in the sequence, including whether they attempted a search, what factors may have prevented detection, and so forth. 6. Match the data with the postulated crash types and information on predisposing factors to update the initial assumptions on how each party failed in performing the function/event sequence. Collect additional data as required. This will provide the most complete picture possible of the crash from a human factors perspective. Using the Results The types of analytical processes presented previously have three primary uses: research and development, countermeasure application, and litigation. The same basic approach applies to all three. An example from the author’s experience based on the ice cream/vendor truck-related pedestrian crash type can further illustrate how effective this approach can be. The ice cream/vendor truck-related crash type was initially identified by Snyder and Knoblauch (1971). Using a function/event sequence analysis as part of a research project, Blomberg and colleagues (1974) developed the Model Ice Cream Truck Ordinance (MICTO). Based on the analysis of a large number of ice cream/vendor truck-related crashes, the MICTO required ice cream vendors to install and use flashing lights and a stop swing arm similar to the ones on a school bus. Drivers coming upon an ice cream truck with its lights and swing arm in operation were required to stop, proceed with caution, and yield to any pedestrian going to or from the truck. The MICTO also restricted vending to those streets on which it did not present an unreasonable hazard and dictated other appropriate safety practices such as prohibiting backing the vehicle to make a sale or selling to a pedestrian standing in the roadway. In a subsequent countermeasure application effort, the city of Detroit passed the MICTO. Using the function/event sequence as a guide, Hale and colleagues (1978) evaluated the Detroit implementation and documented a 77% crash reduction in the ice cream/vendor truck-related crash type that was attributable to the new ordinance. More recently, this author has been involved as an expert witness in several lawsuits arising from ice cream/vendor truck-related crashes. Although the circumstances of each case varied, the function/event-based analytical approach proved effective in documenting the underlying causes of all the crashes and in providing a compelling picture to the jury and the litigants As with any analytic technique, the human factors analysis of pedestrian and bicycle crashes is only as good as the available data. Detailed physical evidence and reliable participant and witness statements cannot always be obtained. Even in the absence of complete information, however, the availability of a well-documented taxonomy of crash


356

types for pedestrian and bicycle crashes with motor vehicles, together with the underlying function/event sequence model, is an invaluable aid that should be more widely employed. An Example An example of an actual crash analysis using the approach just outlined will help illustrate the benefits of viewing crashes in terms of a human factors model. The particular crash used in the demonstration

FIGURE 13.10 A replica of the actual crash diagram for the example pedestrian crash. is an actual crash selected because it is simple and fits one of the types discussed earlier. This crash was analyzed by the author as part of a research study, but it will be assumed here that it is of interest in a forensic setting from the perspective of a plaintiff’s attorney. The crash involved a 6-year-old pedestrian struck while crossing the street, not at an intersection, in mid-May at approximately 5:00 p.m. It occurred in full daylight on a residential street. The striking vehicle was a passenger car. Neither the driver of the striking vehicle nor the pedestrian victim was cited for a code violation by the


357

investigating officer. A replica of the actual diagram drawn by the police officer at the scene is shown in Figure 13.10. A review of the diagram shows that the child crossed in front of a stopped vehicle that likely produced a visual screen. The striking driver overtook the screening vehicle and struck the child as his car cleared the front of the screening vehicle. This diagram and the age of the victim suggest a classic dartout situation in which the screening vehicle predisposes the crash by allowing only a short time exposure of the child to the striking driver. It might also fit a multiple threat crash type even though the pedestrian was not in a crosswalk. If the stopped vehicle had halted to yield to the pedestrian in spite of no legal requirement to do so, the collision by an overtaking vehicle would have constituted a multiple threat. In either case, the focus for liability would rest wholly or substantially on the striking driver. In reality, the crash situation was somewhat different, as indicated by the police officer’s narrative shown in Figure 13.11. The narrative indicates that the stopped vehicle was, in fact, an ice cream truck and the crash type is therefore ice cream/vendor truck related. This is of more than academic interest to a plaintiff’s attorney. The extensive research on this crash type has raised the standard of care in the industry as well as spawning implementations of the MICTO. Ice cream vending companies are generally well aware of this crash type and the steps they can take to prevent it. Given that the ice cream truck predisposed this crash, it would behoove a plaintiff’s attorney to examine factors such as the training of the truck driver and the safety equipment installed on the vehicle when creating a theory of liability and planning discovery.

FIGURE 13.11 Actual officer’s narrative for the example crash. Note also that the narrative indicates that the officer believes the pedestrian is at fault even though no citation was issued. Although the pedestrian most likely did not search adequately for vehicular threats, this function/event sequence failure was greatly predisposed by the visual screen presented by the ice cream truck. This visual screen also predisposed the striking motorist’s detection failure. Overall, therefore, the ice cream truck’s visual screen was the major “cause” of this crash. This same example crash can illustrate the assistance possible from PBCAT in determining a crash type. PBCAT leads the user through a series of screens that request


358

specific input on the circumstances of the crash and uses this information to determine a crash type. For this ice cream/vendor truck-related crash, the screens would proceed as follows:

FIGURE 13.12 Opening PBCAT screen.

FIGURE 13.13 PBCAT screen for entering type of location for the example crash.


359

1. The user selects “pedestrian crashes” to start the typing process (Figure 13.12). 2. The user selects “crash typing” from among the tasks PBCAT can perform and enters database identifying information, e.g., report number (screens not shown). 3. The user begins to answer questions about precipitating and predisposing factors leading to a type decision. The first decision concerns where the crash occurred. For this example, “non-intersection location” would be selected from the screen shown in Figure 13.13. 4. The user is asked if the crash involved any unusual circumstances such as a deliberate assault with a motor vehicle. In this example, the user would select “none of the above.” The user is then asked if the crash involved any unique vehicle type or action, including a backing vehicle, driverless vehicle, disabled vehicle, emergency vehicle, or play vehicle. In the example, “none of the above” would be selected (screens not shown). 5. The user is then asked if the crash involved any unique pedestrian actions, one of which is “ice cream/vendor truck related” (Figure 13.14). 6. PBCAT pops up an overlay on the preceding screen that gives the crash type number and name and seeks confirmation from the user. The user then selects “OK,” the data are stored in the database, and the process is complete (Figure 13.15). This simple example shows only part of the power of a crash type analysis. By identifying what would otherwise appear to be another midblock crash involving a young pedestrian as a specific type, a litigant, researcher, or public safety official gains significant additional insights. These insights, in turn, can lead to better and more appropriately focused lawsuits, richer information on which to base countermeasure development and evaluation, and more effective societal responses to pedestrian crashes. 13.5 Positive Guidance Principles 1 Introduction Positive guidance means giving drivers the information they need to avoid hazards, when and where they need it, in a form in which they can best use it. Because positive guidance can be achieved only through 1

© Copyright by Positive Guidance Applications, Inc. Permission is expressly granted to Taylor & Francis for any use within its purview. Permission is further granted to quote excerpts with customary attribution. Permission to reproduce or republish any substantial portion of text must come from Mr. Alexander.


360

FIGURE 13.14 PBCAT screen on which “ice cream/vendor truck related” is selected

FIGURE 13.15 PBCAT type number determination screen.


361

the understanding and integration of human factors and highway engineering technologies, some basic human factors principles are included in this section. The materials here are drawn from Federal Highway Administration’s series on positive guidance (1975) (Post et al., 1977, 1981; Lunenfeld and Alexander, 1990) as well as materials developed by the author for the State of Maryland (undated) and the Province of Ontario (Alexander and Lunenfeld, 1998). In terms of driver behavior, optimum highway design is achieved when drivers know what to expect from the highway and their attention is naturally attracted to the most important sources of information, and they have adequate time to respond to conditions and situations as they arise. To attain this objective, it should be evident that all elements of highway planning, design, construction, maintenance, and operations consider expectancy and primacy. Definition and Concept Any information carrier, including the highway, that helps or directs drivers in making speed and/or path decisions, transmits guidance information. Positive guidance is provided when that information is presented unequivocally, unambiguously, and conspicuously enough to meet decision sight distance criteria and enhances the probability of drivers making appropriate speed and path decisions. Control, Guidance, and Navigation Levels of Performance Control refers to task performance related to a driver’s interaction with the vehicle; vehicles are controlled in terms of speed, path, and direction. Drivers exercise control through the steering wheel, accelerator, and brake. Information about how well drivers perform at the control level comes from the vehicle and its displays as well as visual observation of changes in speed, path, and direction. Drivers receive continual feedback through vehicle response to various control manipulations. Overt response to hazards is part of the control level of performance. Guidance refers to task performance related to a driver’s selection and maintenance of a safe speed and path. Control subtasks require action by the driver. Guidance requires decisions involving judgments, estimates, and predictions. The driver must evaluate the immediate environment and translate changes in alignment, grade, and traffic into control actions needed to stay in the appropriate lane at an appropriate speed for the prevailing conditions. Information at this level comes from the highway—its alignment and grade, geometric features, hazards, shoulders, etc.; from traffic—speed, relative position, gaps, headway, etc.; and traffic control devices—regulatory and warning signs, signals, and markings. Navigation refers to the activities involved in planning and executing a trip from origin to destination. Drivers generally evaluate route numbers and/or names, street names, interchange or intersection designations, cardinal directions, and landmarks. They make guidance-level decisions at choice points and ultimately translate those decisions into control actions. Offline information sources include maps, verbal directions, and prior experience. Online information input comes from the full range of guide signs, verbal directions, and landmarks. Through the implementation of intelligent


362

transportation systems (ITSs), map information is increasingly presented online in the vehicle. (Noy, 1997; Barfield and Dingus, 1998). Task Complexity Information and task performance associated with the three levels of performance form a hierarchy of complexity. At control, the lowest level, information processing and vehicle handling are relatively simple and so completely overlearned by experienced drivers that they are performed almost without conscious thought. At the guidance and navigation levels, information handling is increasingly complex and demanding, and drivers need more processing time to make decisions and respond to information inputs. The need for more time frequently occurs in urban locations, at intersections and interchanges and where traffic demand is heavy. The nature and number of hazards and of the available information displays also affect task complexity. The scale of complexity increases from control through navigation. Primacy It should be evident that at any given location on the highway, some information is more important than other information. Primacy refers to the relative priority of each level of the driving task and of the information associated with a particular activity within each level. Here, too, is a hierarchical scale. The major criterion upon which primacy is assessed is the consequence of driver performance error. Because loss of vehicle control, the results of which can be catastrophic, is of the greatest immediate concern to the driver, the control level is assigned the highest level of importance. A guidance level failure also is assigned a high primacy because errors in speed and path selection frequently result in accidents (Haworth et al., 1997). In navigation, where errors usually result in delayed, lost, or confused motorists, the lowest priority is assigned. The scale of primacy decreases from control through navigation. Primacy is a most important consideration when information competes for drivers’ attention. If there is inadequate time to process all the information at one location, high primacy needs should be satisfied first and lower primacy needs should be deferred. Information should be placed so as to spread the information challenge at locations where critical driver decisions and/or actions are required. Information Handling While driving, drivers do many things simultaneously or nearly so. They monitor traffic, follow the road, stay in the lane, read signs, listen to the radio, and accelerate and decelerate their vehicles. At any given point in time, drivers may have several overlapping needs associated with each level of performance. To handle this array of needed and available information, drivers must search the environment for information sources, detect their presence, recognize their message, make decisions, and perform control actions safely and efficiently. Thus, information must be available when and where needed, and in a form best suited for driver comprehension. In situations in which information competes for drivers’ attention, unneeded and low-primacy information is


363

shed. Errors can occur when drivers process less important information and miss or shed more important information. Perception Reaction Time PRT includes the components of information processing, detection, recognition, and decision making plus action initiation. It not only varies from individual to individual but also is a function of decision complexity, information content, and driver expectancy. The more complex the decision is or the more information is needed to make a decision, the more time it takes. Clearly, long PRTs reduce the time available to load shed, attend to other information sources, and respond to other task requirements, increasing the probability of error. Although 2.5 sec is the constant used for PRT in design and sight distance calculations (AASHTO, 1994), it is hardly a constant. Even for something as simple as brake-reaction time, there are substantial ranges in PRT, with the range of responses to unexpected signals higher than the range of responses to expected signals. (See also section titled “Response Time.”) Driver Expectancies and Surprises The nature of the driving task and drivers’ information-handling characteristics emphasize the importance of expectancies. Reaction to an unexpected event takes longer than when the event is expected (Johannson and Rumar, 1971). Conversely, drivers are less likely to become confused or commit errors when their expectancies are reinforced. Because the key to safe, efficient driving-task performance is rapid, error-free information handling, what drivers expect and do not expect has a major influence on task performance, particularly under time pressures and high task loading. Expectancy relates to a driver’s readiness to respond to conditions, situations, events, and information. It influences the speed and accuracy of information processing and is one of the more important driver-related characteristics in the design and operation of highways. Configurations, geometric features, traffic operations, and traffic control devices that meet or reinforce expectancies help drivers to respond quickly, efficiently, and without error. Roadway Design The configuration of a fully directional freeway-to-freeway interchange is a good example of an unusual design configuration that may contribute to driver expectancy problems. In a fully directional interchange, the exit to go left, i.e., the north-to-west, the east-to-north, the south-to-east, and the west-to-south movements, are all left exits. Because in conventional interchanges almost all exits are on the right, unfamiliar motorists expect to exit from the right lane of the freeway. Without conspicuous, specific advance notice, motorists unfamiliar with the road who want to exit to “go left” at the exit will move to the right lane. Here, motorists will miss their exit or perform an erratic late lane change to get to it. When the left-lane exit movement is also a lane drop, there is double expectancy violation. More drivers are affected, interactions in the traffic stream are more turbulent, and the potential for driver error is greater. Wherever this geometric


364

design feature has been located, from Maryland to California, it is usually a source of operational problems. Conventional guide signs have been shown to be ineffective for left exits and left exit lane drops, and diagrammatic advance guide signs are recommended by the Manual on Uniform Traffic Control Devices (MUTCD) (Federal Highway Administration, 2000). At most freeway exits, motorists must move into a deceleration lane to exit the facility. It is therefore an expectancy violation when a lane that had been a through lane exits the facility directly. Instead of needing to change lanes to leave the freeway, motorists are required to change lanes to stay on the freeway. The black-on-yellow “EXIT ONLY” panel has the needed conspicuity when placed on the white-on-green guide sign to gain drivers’ attention. The uniform application of the panel at interchange lane drops serves to structure the appropriate expectancy. Freeway tangential exit ramps create expectancy problems for drivers. Interchange exits with this configuration are the scenes of many unintentional erratic maneuvers and other errors. Drivers find themselves leaving the freeway by going straight ahead on the tangent while the freeway curves to the left. The tangential off movement is thus an unexpected feature and one that creates perceptual problems whether the tangential exiting movement is at the beginning of the curve or within it. No traffic control device has yet been found that can adequately warn drivers about tangential exit ramps. Lacking superior sight distance, tangential exit ramps are best treated by configuring the diverge area so that the off movement does not appear as the continuation of the main roadway. The desired effect could be achieved if the diverge area could be relocated as little as 100 ft up- or downstream of the curve or the divergence angle were no longer tangent to the curve, so that drivers would be required to make a steering adjustment to exit. Rural two-lane road situations function similarly to the freeway tangential exit ramp. Off-road features, such as a line of trees or railroad tracks that run parallel and adjacent to the highway, create drivers’ expectancy that the condition will continue straight. A similar situation occurs where a tangent roadway intersects at the point of curve of a turning roadway. A common freeway design feature with the potential for violating expectancies is a variant of the interchange lane drop: the split or bifurcation. Two kinds of split surprise drivers: any split in which the off-route movement is to the left of the through-route movement, and the optional lane split. The optional lane split creates expectancy problems for many drivers. Because it is a lane drop, drivers in the exiting lanes are affected. Additionally, drivers in the optional lane do not expect to be faced with a lane choice by staying in lane. This situation can be described as a classical dilemma—the choice between equal alternatives. When drivers make a late choice—or worse, no choice—it is foreseeable that something undesirable will happen, such as an erratic maneuver, a gore crossover, a fixed object struck in the gore, or a truck jackknife. Any reduction in width of the road represents an expectancy violation and a hazard to drivers. Situations such as mainline lane drops, work zones, and narrow bridges are common sources of pavement width reduction. Although all are expectancy violations, narrow bridges are particularly difficult because of the many configurations they take, from those that are short box culverts to long bridges with trusses. Their narrowness


365

ranges from loss of shoulder to narrowing of a lane width to a one-lane bridge that handles two-way traffic. Narrow bridges occur on curves (horizontal and vertical) and in sags beyond the crest vertical, making them hard to see. Thus, not only are they unexpected, but they may also be difficult to detect, recognize, and negotiate in the presence of oncoming traffic. Positive guidance treatments for a variety of narrow bridge configurations are contained in Appendix A of the “Yellow Book” (AASHTO, 1974). Traffic Control Devices Traffic control devices serve to structure expectancies about downstream features and operations. They also structure expectancies about information treatments at similar locations. The key to effective expectancy structuring is uniformity and standardization. Inconsistently applied standard devices create expectancy problems for drivers. If upstream curve warning signs underestimate maximum safe speed, drivers will expect similar underestimations for similar curves downstream. When a downstream curve is more realistically signed, drivers may be unprepared or unable to respond properly. Traffic control devices not only serve to structure expectancies but also tend to violate expectancies if they are misapplied, inconsistently applied, absent when needed, present when unneeded, and/or ambiguous. Traffic signals often violate expectancies. At many signalized intersections, motorists who are stopped for the red can see the signal display for the crossing roadway. This is particularly true at intersections where the roadways cross at something other than a 90° angle. From the stopped position, the lenses on the signal face for the crossing movement are frequently clearly visible. Invariably, a nonlocal driver at the head of the queue will move into the intersection when the crossroad signal changes from yellow to red, expecting to get the green; all too frequently, however, his signal indication does not get the green. Lagging greens, protected turning movements, pedestrian phases, and even clearance intervals surprise the unfamiliar motorist. Local drivers know the condition and stay put until the light changes to green. However, visibility of the crossing movement signal induces inappropriate behavior of those unfamiliar with the signal operation. Another example of an unexpected traffic signal indication is the mid-block signal. In most instances, drivers do not expect a traffic signal anywhere but at an intersection. When a mid-block signal is used, they will not be prepared without conspicuous advance warning and may not react in time or may hit the rear end of another vehicle stopped at or in advance of the crosswalk. Signs that provide information at the guidance level (regulatory and warning signs) as well as at the navigational level (guide signs) have the potential to structure and to violate driver expectancies. These expectancies are, of course, based on the driver’s experience. For example, drivers generally expect to be able to exceed the advisory speed safely when one is posted beneath a curve warning sign. Advisory speed warning plates are usually conservative when it comes to a safe speed under most conditions. On overhead guide signs, down arrows generally mean one needs to be in the specific lane to get to the signed destination, but not always. In some locations, the warning message requires special emphasis. Curves where the advisory speed should be adhered to, even in dry weather, and unexpected situations beyond a curve or vertical crest that would surprise drivers are two examples in which


366

such special emphasis is justified, i.e., something that says, “This time we really mean it.” Special display treatments such as chevron alignment signs, oversize warning signs, and flashing beacons are used to good effect at such locations. For the sake of continued credibility of these special emphasis devices, however, their use should be reserved for locations in which gaining drivers’ attention is important. Key Considerations The development of appropriate highway designs and traffic control devices that meet driver expectancies or that tell drivers what to expect is the principal way to aid performance and enhance safety and efficiency on all highways. Attention should be given to assure consistent design from one segment of highway to another. When drivers get the information they expect from the highway and its information system, driver response tends to be rapid and error free. When drivers get what they do not expect or do not get what they do expect, longer response times, inappropriate responses, confusion, and errors are the predictable result. Key considerations about expectancies include: • Expectancies are associated with all levels of the driving task and all phases of the driving situation. • Drivers experience problems and commit errors when they are surprised. • Drivers anticipate upcoming situations and events that are common to the route they are driving. • The more predictable the design, information display, or traffic operation is, the less likely will be the chance for driver error. • In the absence of information to the contrary, drivers assume they will need to react only to standard (expected) situations. • The roadway, the information system, and the environment upstream will structure expectancies of downstream conditions. The objective in helping drivers overcome the effects of expectancy violation is to structure the appropriate expectation through advance warning. When it is not possible to give drivers what they expect, it is imperative to tell them what they should expect. Failures at the Guidance Level A guidance-level failure occurs when the driver chooses an inappropriate speed and/or path. The failure translates to improper, inadequate, or inappropriate control actions and does not imply driver fault. When an accident occurs because of wrong control action, it may be caused by certain inadequacies in relevant information available to the driver. Such inadequacies include too much information, too little information, ambiguity, conflicting information, improper location of information, and information not visible under ambient conditions. It is the function of positive guidance to enhance safe driver performance by providing appropriate, usable information that would reduce failures that are not caused by the driver. Two limitations to the potential effectiveness of positive guidance are important here. First, driver failures due to driver impairment are not necessarily amenable to correction by providing improved highway information. Drivers who are drunk, drugged, or drowsy


367

or whose normal performance is otherwise impaired have problems that are not usually solved by better signs and markings, although exceptions to this generality do exist. An example is the use of wider lane edge lines to aid drivers impaired by alcohol at night. Second, certain highway design features exceed driver response capabilities. When the design complexity is such that drivers do not have enough time to make all the judgments required, no solution short of redesign will eliminate frequent accidents. Information at the Guidance Level At the guidance level of performance, drivers select and process information with the objective of selecting and maintaining a speed and path they consider safe, efficient, and comfortable. Roadway Environment Drivers gather considerable information from the roadway itself. Drivers’ ability to select an appropriate speed and path depends on their ability to see the road. They must see the road directly in front and see enough of it at some distance ahead to predict its alignment, grade, width, and several other factors with a high degree of accuracy. Drivers’ view of the road includes a view of the immediate environment including the shoulder and any obstacles. This also includes shoulder surface and width, sign supports, bridge piers, abutments, guardrails, and median barriers. Information received from the roadway and its immediate environment is used continuously during performance at the guidance level. Traffic Control Devices Three kinds of devices directly affect guidance performance: pavement markings and delineators, regulatory and warning signs, and signals. Guide signs, associated with the navigation level, indirectly affect performance at the guidance level and, as such, are an important source, although lower in primacy. Traffic Maintenance of a safe speed and path in response to other vehicles in the traffic stream is a major activity at the guidance level. Drivers must process information received from other vehicles at the same time as other information related to the driving task. Traffic information may be intermittent or continuous, but in either case, it must be integrated with other guidance information to assure adequacy of speed and path decisions. Hazards at the Guidance Level A hazard is any object, condition, or situation that tends to produce an accident when drivers fail to respond successfully. Object hazards, of course, can be fixed or moving. Condition hazards refer to conditions of the major system elements: driver, vehicle, or roadway environment. Situation hazards are combinations of conditions and objects,


368

usually with a temporal feature, e.g., a wet pavement or a train approaching a highwayrailroad grade crossing. It is well beyond the scope of positive guidance to deal with driver or vehicle condition hazards. Guidance-level condition hazards are only those of the roadway environments. Fixed Objects This type of hazard is the most obvious because it includes objects recognized as killers: bridge rails, piers and other bridge elements, sign supports that do not break away, large trees, etc. In general, any object that is stationary and accessible is included. The fact that some objects are protective devices, like guard rails and median barriers, does not mean they are without hazard to motorists. Moving Objects Anything that can move into a driver’s path falls into this category (e.g., other vehicles, pedestrians). Driver assessment of what is hazardous is relatively simple for the fixedobject hazard but somewhat more complex for the moving-object hazard. Furthermore, the decision process involved in avoiding a moving-object hazard is also more complex in that drivers are required to evaluate the speed and path of the moving object, make corrections in their own speed and path, and reevaluate. This iterative process is well within the capability of most drivers but can consume much processing time and information-handling capacity. When seen in time, the object hazard is the simplest hazard with which to deal. Although the decision-making process is complex under some situations, the identification of what is hazardous is usually rapid and error free. Unfortunately, perception of highway condition and situation hazards is neither simple nor without error. Highway Conditions The condition of the highway, its design features, and its state of maintenance or repair, irrespective of any obstacles, contribute to consideration of the roadway environment as a hazard. Included are: • Design features such as tangential off-ramps or lane drops and other expectancy violations as mentioned earlier • Accessible roadway features that make it difficult to maintain or regain control of an errant vehicle, such as potholes, pavement edge drops, and curves with inadequate superelevation All of these features create perceptual problems and expectancy violations for drivers. Without positive guidance, they can be expected to induce driver error. Any location in which the condition of the highway or its immediate environment needs to be interpreted as a cause for extra caution or a cause to modify speed or path significantly should be considered as a highway condition hazard.


369

Situations This hazard category includes combinations of conditions with or without objects and may include a temporary condition such as darkness or rain. A situation hazard could include conditions that, taken individually, may be of only moderate concern but in combination are treacherous. Combining such elements as rain, a polished surface, a vehicle with bald tires, a curve with not quite enough superelevation, and an insufficient recovery area for errant vehicles leads to the kind of situation hazard responsible for many skidding and single-vehicle running-off-the-road-type-accidents. Highway-railroad grade crossings are good examples of the difference between highway condition and situation hazards. Many crossings have several hazards associated with them. Elevated tracks, crossings at angles other than 90°, and rough crossings are conditions that warrant extra driver caution. When they all exist at the same crossing, the problem is much more serious for drivers. It is the approach of a train, however infrequent, coming from the acute angle that takes the crossing to another level of hazard. Strategic Improvements Looking at the range of hazards, it is possible to define the obligation of any state or highway authority to motorists on its highways. First, if possible, practicable, and within the financial and programmatic ability of the relevant traffic authority, the hazard should be eliminated. If that cannot be done (and there are many valid reasons for that to be the case), then the hazard should be made inaccessible or forgiving (move it, screen it, or make it breakaway). If that cannot be done (again, there are many valid reasons for that to be the case, particularly with design features), motorists must be given enough information to avoid the hazard. It is that information that provides positive guidance. Short of closing the roadway to traffic, no other alternatives exist. A final word about the term hazard: increasingly in personal injury litigation, the term is used to imply negligence. As used here, however, that is not the case. Having a hazard on the highway does not make a cognizant agency negligent. Negligence accrues when there is knowledge of the hazard and little or nothing is done to ameliorate its effects. Drivers Avoiding Hazards Successful performance by drivers is dependent upon their ability to detect a hazard, recognize it or the threat it poses, decide on an appropriate speed and path, and act on that decision. The principles of positive guidance would require that drivers be given all the information needed to maximize the selection of appropriate speeds and paths. Detecting the Hazard Hazards range in detectability from very easy (seeing a fixed object in the road) to very difficult (seeing a 3-in. pavement edge drop-off at night). How seeable a hazard is depends on many factors, including the interaction among its visibility, its conspicuity, and the number of competing information sources. Also included are the driver’s scanning behavior, visual acuity, prior knowledge, and expectancy. This interaction defines how detectability of a hazard can be enhanced. In the case of a fixed object,


370

making it more visible makes it more detectable. With objects and highway conditions, reducing the number of information sources competing for drivers’ attention gives drivers more time to detect the hazard. In the case of any hazard, increasing driver expectancy of seeing the hazard will improve its detectability. Here, signing and marking play an important role in the hazard detection task. Recognizing the Hazard Hazard recognition is a simple name for a complex mental process. Once something has been seen, the driver must decide what it is. Because recognition follows detection in time, the driver is closer to the hazard and can see it better (or more of it) and get more information from it. That information is compared with the driver’s store of prior knowledge. Prior driving experience becomes increasingly important in recognizing highway condition- and situation-type hazards, although some of the knowledge can be gained through driver training. Some knowledge is too situation specific to be taught in driver education, and in those cases hazards are recognized through personal experience, flagged with a device (usually a warning sign), or not recognized. Warning signs prepare motorists to detect and recognize hazards. When the hazard is present, the value of the warning is reinforced. However, when the hazard is not there or not apparent to drivers, the credibility of the warning or regulation is reduced. A typical example is work zone signing, including reduced speed warnings with no work or workers in sight. Deciding What to Do After the hazard is recognized, drivers need to determine if modification of speed and/or path is necessary and, if so, define alternative courses of action. If more than one course of action is considered, drivers evaluate the probability of success as well as the ease and comfort of implementation. Here, too, experience plays a big role; we all tend to repeat past behavior that has been successful. Finally, drivers select the speed and path they consider to be the most appropriate for the situation. These decisions are frequently made under great time pressure. Here, it can be seen that those who are inexperienced and those whose information-processing abilities have deteriorated through advanced age or impairment are at a disadvantage. Doing It Helping motorists to select the appropriate speed and path is as far as the positive guidance process can go. Vehicle control to implement the decisions is entirely in the hands of the driver. After taking an action, the driver evaluates its adequacy, applies a speed or path correction if required, and continues the process until the hazard no longer poses a threat or a higher primacy need intervenes. A great deal of information handling occurs in this process—some at the guidance level and much at the control level.


371

Planning, Design, and Construction Although the positive guidance procedures in the literature are designed to be used by traffic operations personnel, the concept and principles, as defined here, are equally important in the planning, design, and construction phases of project development. Traffic control devices are usually considered the principal means of communicating with motorists. However, the highway itself conveys more information to its users than any other single source. Planners and designers whose job is to determine what the highway will look like play a key role in the development of highway-related information. In fact, any activity whose output conveys to highway users information related to the driving task is an activity with potential for providing positive guidance, whether or not information giving is its primary intent. For example, although the principal purpose of guardrail placement is the physical protection of motorists from fixed object hazards and the redirection of errant vehicles, its placement and appearance also give motorists important guidance information. The information imparted is positive guidance when it assists highway users in making correct speed and path decisions to avoid hazards. Here, it is no more or less of an example of positive guidance than a line of retroreflectorized barrels delineating the temporary edge of travel lane in a work zone. The alignment and profile relationships of any highway are crucial to the formulation of accurate driver expectancy. Together, and in combination with other features such as superelevation, signs, markings, and roadside grading, they provide the positive guidance drivers need to conduct the task safely and efficiently. It is essential, therefore, that all the elements act in concert. For example, an urban or suburban facility should not be planned or designed to give the impression of a higher type facility than the posted speed limit would warrant. The ambiguity created by an apparent high type facility and relatively low speed limit violates driver expectancy, creates credibility problems, and invites speeds higher than can safely be accommodated. Positive Guidance Analysis of Two Accident Locations A Passively Protected Railroad Grade Crossing The Uniform Vehicle Code specifies that drivers approaching a passively protected railroad grade crossing be required to yield to oncoming trains in hazardous proximity to the crossing. They are required to maintain a speed from which they can stop within 15 ft of the track if, at any point on their approach, they see an oncoming train that is or will be in hazardous proximity to the crossing by the time they get there. Because no warning sign is present to specify or even suggest an appropriate approach speed, it is essential that drivers know how much distance to the track they will have when conditions permit an approaching train from either direction to be seen so that they may select an appropriate approach speed. In many, if not most, rural grade crossings, long “corner sight distances” are available, providing early detectability of an oncoming train and making the selection of an appropriate approach speed simple for all but the most novice drivers. In the absence of very long sight distances, however, three conditions are necessary for drivers to make the


372

approach speed decision reliably: familiarity with visibility conditions at the crossing, experience at making speed judgments relative to rates of closure, and good visibility conditions (daytime or high ambient lighting) in the vicinity of the crossing. At 10 p.m. on a cold Friday night in December of 1994, a 16-year-old driver and two friends went for a ride after work. They lived in a small town in a large midwestern state and drove to a neighboring town about 5 miles away. They had not driven there before, so they decided to look around. As she made a turn onto a quiet residential street, the driver saw a railroad advance sign and a crossbuck two blocks away. In the first block, homes were on both sides of the street. In the second block, buildings were on the right side and homes and a stand of trees on the left. These impediments to the corner sight distance did not permit an unobstructed view in either direction until the driver was about 50 ft from the crossing. The crossing had no external illumination; neither did the street nor any location in the vicinity of the track. There was a single track, skewed at a 75° angle to her right and 105° to her left. It was elevated 5 ft above the approach roadway with an 8% grade over the last 50 ft. Because of its elevation and the lack of illumination, the track was not visible. A passenger train, traveling about 80 mph, approached from the right, blowing its horn. There was no posted speed limit on the street. Clear visibility to the right for the approaching motorist does not occur until the driver’s eye position is within 50 ft of the track, and even then vision would be obscured by the vehicle’s B pillar. The vehicle speed at which 50 ft is an adequate stopping sight distance is about 5 mph. For vehicles traveling at the speed limit of 25 mph and trains traveling at 75 mph, a driver would need to be able to see an approaching train when they are no closer than 165 ft (or 4.5 sec) from the crossing in order to bring his vehicle comfortably to a safe stop 15 ft from the track. At 4.5 sec from the crossing, trains traveling at 75 mph (or 110 ft/sec) are still 500 ft from the crossing and completely hidden from view. Although the engineer was blowing the train’s horn, several factors mitigate against that as an effective or even adequate warning. Even when effective as a general warning, train whistles do not unequivocally identify the specific location, speed, or direction of the train. Seeing the train is the only way to get that information. Ambient temperature and radio use play a role in the ability to hear well enough to identify the train horn or whistle. In mid-December, it is probable that windows are rolled up and heater fans turned on as well as the vehicle occupants engaging in conversation. In all, seeing the train early enough to respond has been recognized as the only way to assure the adequacy of information at passively protected crossings. According to Lerner et al. (1990), “Given such factors as raised windows, noise from heaters or fans, wind noise, engine noise, radios or tape decks and other outside noise sources, acoustic train signals cannot be considered useful means of providing accurate information on dynamic aspects of trains.” The Lerner report further indicates that physical barriers may reflect or absorb sound, such as buildings, trees, and foliage, thus attenuating sound level at the ear of the listener. Insertion loss of the train’s horn sound into the vehicle would, under these conditions, completely mask the sound until the last few seconds of its approach to the crossing. (See also section titled “Railroad Crossings.”) Differences in elevation as well as smoothness of the crossing play a role in the driving task at railroad grade crossings in general. Because of the steep grade in the


373

immediate vicinity of the railroad crossing, drivers must pay close heed to negotiating it. This makes the task more complex than at flat, smooth, and 90°crossings. Drivers must ascertain an appropriate speed and path to traverse the elevated crossing and determine whether there are any oncoming vehicles, in addition to searching left and right for train approach. Unless motorists come to a virtual stop at the crossing, in the last few seconds of this task attention must be paid to negotiating the crossing. In observing traffic operations at the crossing, it became evident that almost all motorists came to a virtual stop before entering the crossing. Given poor night visibility, such behavior requires prior knowledge of the conditions at the crossing. The train and the car arrived at the crossing at about the same time. There was no physical evidence of braking by the driver. She survived the collision but had no recollection of the event. The passenger in the right front seat was killed and the driver and rear seat passenger were seriously injured. To paraphrase a U.S. Senator asking questions at the Watergate Hearings, what information did the driver need in order to avoid the accident, and when did she need it? In other words, what information would have provided her with positive guidance? At some point in her approach to the crossing, what did she need to know? • That she was on a collision course with a train and, if she continued at the speed limit, her vehicle and the train would arrive at the crossing at the same time? • The approach speed and distance of the train from the crossing? • The approach speed at which she would have enough time to see a train, recognize its threat, decide on a course of action, and bring her vehicle safely to a stop? Did she need any of that information, or was this one of those locations where a simple STOP sign at the crossing would be the most appropriate guidance? It seems evident that the simpler and more unequivocal the information, the more likely it is to be the right answer. We need merely to turn to the Manual on Uniform Traffic Control Devices for guidance: To be effective, a traffic control device should meet five basic requirements: Fulfill a need. Command attention. Convey a clear, simple meaning. Command respect of road users. Give adequate time for proper response. However, none of this information was available to the driver that December evening. An Urban Intersection Work Zone Work zones are stress-inducing situations. Workers perform their tasks in proximity to the frequently high-speed and high-volume traffic stream, where one errant vehicle can cause injury and death. They rely on temporary traffic control devices to keep them safe by giving drivers a clear understanding of the appropriate speed and path through the work zone.


374

For drivers, work zones are equally stressful, whether or not workers are present. Because temporary traffic control information is superimposed on the existing highway and traffic control information, the quantity and complexity of information available to process are increased. As such, drivers need more time to determine appropriate speed and path, particularly on approach and at decision points. However, operating speeds are usually not reduced significantly (if at all), giving drivers even less time to handle the increased complexity of information. Furthermore, because of the increase in complexity, potential ambiguities between the existing information and the superimposed work zone information are also increased. The layout of traffic control devices changes during the course of the work, sometimes frequently. The benefits of familiarity and driver expectancy are lost, and drivers are left to interpret a sometimes-ambiguous meaning of a different array as conveyed by the layout and spacing of channelizing devices and temporary signs. In Figure 13.16, a line of barricades closes the left-turn and adjacent lanes (lanes 1 and 2), although the protective/permissive signal controlling the left-turn movement has not been moved to the temporary left-turn lane (lane 3), two lanes to the right. For the driver who intends to turn left at the intersection, the decision is made more difficult by the presence of the barricades extending well beyond the intersection.

FIGURE 13.16 A line of barricades closes the left-turn and adjacent lanes; the protective/permissive signal controlling the left-turn movement has


375

not been moved to the temporary leftturn lane. A driver familiar with the intersection shown in Figure 13.16 and Figure 13.17 attempted to make the left turn from lane 3 at night during the permissive signal phase—a green ball in the left-turn signal. She knew she was required to yield the right-of-way to any vehicle coming through the intersection in the opposite direction. In testimony, she stated that she had stopped twice in the intersection in an attempt to determine whether any vehicles were coming from the opposite direction. An eyewitness corroborated her testimony. As she got to the opposing traffic lane seen in Figure 13.17, her vehicle was struck broadside, killing her daughter in the front passenger seat and severely injuring her twin 3-year-old boys in car seats in the back. She further testified that when she had been there earlier in the day, the original left-turn lane was still open, so she had been surprised by the new configuration. It is evident that the line of barricades obscured her vision. However, other issues pertaining to highway engineering and human factors played a significant role in the litigation:

FIGURE 13.17 Line-up of the far-side barricades at night. • Is it reasonable to have the left-turn signal over lane 1 and the turn from lane 3? The arrangement is highly ambiguous and violates the MUTCD. From a positive guidance perspective, such ambiguity can lead to confusion, delayed decisions, and errors. Inasmuch as lane 3 was not an exclusive turn lane, drivers turning left needed to determine whether the left-turn signal still controlled the left-turn movement. For


376

example, if the left green arrow is illuminated over lane 1, and the signal over lane 3 is red, is a left turn legal? Furthermore, would it be appropriate to wait at the stop line when the signal is green until the protected phase was activated? It is at least foreseeable that errors will be made in this decision process. • Is it reasonable to have a permissive green for a 7-sec turn? The left turn from lane 3 would take several seconds longer than the left turn from lane 1. What does this mean in terms of how much farther a driver would need to see to determine whether clearance was adequate to make the turn? If it takes 7 sec to initiate and complete the turn, at least 10 sec of visibility should be available to provide a comfortable safety margin. With an operating or 85th-percentile speed of 45 mph or 66 ft/sec, 660 ft of visibility distance would be necessary. Not only was that distance unavailable because of the barricades, but it is also considered unlikely, even given full visibility at the location, that drivers stopped at or approaching the intersection would appreciate how far they would need to see in order to make the judgment. • Is this one intersection or two? The original roadway was a four-lane divided arterial with exclusive left-turn lanes. Two new outside lanes were being added and the median was being modified. On the day of the accident, the four inside lanes were temporarily closed to traffic and the two new outside lanes were opened. The effect was to create a six-lane divided arterial with a 56-ft-wide median (four 12-ft lanes and the 8-ft median). Treating the temporarily reconfigured intersection as two, with appropriate striping and signing, would have eliminated the need for protecting the left turn; drivers who were turning left would have been required to stop at a red signal and stop line before crossing the opposing traffic lanes. The purpose of the positive guidance analysis in these two litigation examples was to assist the court in determining whether the roadway and the information-processing task improperly burdened the drivers and played a role in accident causation. Furthermore, the analysis was helpful in defining alternative designs that met relevant design or traffic control standards and provided drivers with clear, unequivocal information about appropriate speed and path decisions to avoid hazards. 13.6 Alcohol, Other Drugs, and Driving Traffic collisions are a major cause of morbidity and mortality in the world. Alcohol and, to a lesser extent, other drugs have been implicated as causal factors in crash involvement. The research on the relationship between alcohol and driving began in the 1930s; research interest in other drugs and driving appeared in the 1970s, initially with illicit drugs, particularly cannabis. Current research interest is focused on licit, medicinal drugs as an emerging problem. Interest in drugs 2 and driving is driven by two questions: • To what extent is it a problem? • If it is a problem, what can we do about it? Studies examining the extent of the problem have found that, despite the increased interest in medicinal and illicit drugs and driving, alcohol is still the most frequently found drug in fatally injured drivers, although polydrug use is common. For example, in a


377

3-year sample of seriously injured drivers consecutively admitted to a regional trauma unit in which the drug screening rate was 90.0%, 59.3% of drivers tested positively for alcohol and/or other drugs, 16.5% tested positively for alcohol in combination with 2

The word drug will be used to connote alcohol and licit, medicinal, and illicit drugs unless otherwise indicated.

other drugs, 18% tested positively only for alcohol, and 24.8% tested positively for a variety of other drugs (Stoduto et al., 1993). The two most commonly found drugs after alcohol were cannabinoids (13.9%) and benzodiazepines (12.4%). The fact that almost 60% of all seriously injured drivers were positive for alcohol and/or other drugs would suggest that drugs could be a problem if evidence were found that drugs did indeed impair driving ability. Current evidence suggests that drugs can impair mental and physical functions and thus contribute to road crashes (Ferrara et al., 1994; Moskowitz et al., 2000). Although more information is available on the role of alcohol in motor vehicle collisions, most evidence suggests that other drugs have a less detrimental impact on traffic collisions than alcohol (Xie et al., 1996); however, other drugs may still have an effect. For example, de Gier (1993) has estimated that at least 10% of persons injured or killed in road crashes in the European Union were taking some type of psychotropic medication. Determining what to do about this problem inevitably leads to the issue of detection, adjudication, and sanctioning of drug-impaired drivers. To date, the issue of detection is most developed for alcohol. For other drugs, the problem is that effects are not fully understood, can be additive, and can interact with alcohol (Crowley and Courtney, 1999). Similarly for adjudication and sanctioning, legal codes for alcohol-impaired driving are the most developed. Industrialized countries have impaired-driving and per se laws, with the per se blood alcohol concentration (BAC) legal limit in most countries set below 0.10% per 100 ml of blood. Sweden has a limit of 0.02% BAC; Australia, Finland, France, the Netherlands, and Norway have set 0.05% as the legal limit; while Austria, Canada, Great Britain, Switzerland, and all U.S. states have 0.08% as the BAC limit. In addition, a number of countries have laws for refusing breath tests. Driving under the influence of drugs other than alcohol is more difficult to detect and prosecute. Drugs can be determined by chemical or behavioral methods: blood, urine, breath, saliva, sweat, hair, field sobriety tests, and drug recognition experts (although internationally and in the U.S., state variations exist for the legal admissibility of blood, breath, and urine samples; field sobriety tests; and drug recognition experts) (National Highway Traffic Safety Administration, 1999). Blood tests can correlate more closely with impairment depending on the drug; however, they are more invasive (Kapur, 1994) and are typically only used in fatal crashes/trauma in North America although commonly used evidentially in the E.U. (e.g., Maes et al., 2002; Willekins et al., 2002). Although many countries have impaired driving legislation that includes licit and illicit drugs and some countries have recently introduced new laws on driving under the influence of illegal drugs (Maes et al., 2002; Willekens et al., 2002), assessment of drug impairment and of accuracy of interpretation is still a challenge. The challenges for assessment and accuracy require three types of research studies to: (1) produce a drug dose-related impairment of skills associated with driving or related psychomotor


378

functions, (2) link these drug-dose effects with driving ability, and (3) link driving skills and any impairment to the skills that result from taking drugs to actual road crashes (Crowley and Courtney, 1999; McBay, 2002). The objectives of this section are to • Describe methodological strengths and weaknesses of the three types of research studies used to assess drug impairment • Provide a simplified, therapeutic categorization of drugs in three groups depending on the ways in which they affect the central nervous system (CNS), with a particular focus on alcohol and other commonly used drugs of concern: benzodiazepines, methadone, cocaine, and cannabis • Review the extent of drug use, including alcohol, in collisions and violations • Provide current evidence on drug impairment and collision risk Methodological Approaches Used to Assess Drug Impairment The combination of laboratory/experimental and epidemiological studies provides the best information for understanding the causal connection between drugs and traffic crashes (Friedel and Staak, 1993; Simpson and Vingilis, 1992). Studies to Produce Drug Dose-Related Impairment of Skills Associated with Driving Experimental studies determine the precise nature of impairment produced by specific drugs and its impact on performance tasks. These studies usually require subjects to perform laboratory psychomotor tasks or driving tasks on simulators or on closed-road circuits or regular roadways. The best research design typically consists of a double-blind experimental approach in which subjects are randomly assigned to an active drug group or to a nonactive drug (placebo), which is replicated using varying drug dosages. Performance differences between these two groups are assessed. The strength of this method is in its internal validity in controlling for extraneous factors, like expectancy of alcoholic effects, that could inadvertently affect outcome (Cook and Campbell, 1979). Thus, the experiment is the method of choice for determining drug and dose response effects on skills performance. The main limitations of laboratory and simulator experiments relate to external validity or generalizability of the findings to the real world. Drug use and the driving environment in experimental designs may not reflect actual patterns of drug use and driving (Del Rió and Alvarez, 1995a; van Laar et al., 1993; Volkerts and van Laar, 1993). Therefore, experiments that assess a driver’s performance may not be applicable to real driving situations. Generalizability problems of experimental research to driving and road crashes (i.e., external validity) can also emerge in relation to the use of healthy volunteers (vs. heavy drinkers, drug users, or actual medication users with health conditions); types of control drugs used; duration and dosages of drugs; types of psychomotor tasks; and other statistical and methodological issues (de Gier and Laurell, 1992; Del Rió and Alvarez, 1995a).


379

Studies to Link Drug Dose Effects with Driving Ability In recent years, a second type of experimental study has been developed to link these laboratory effects with driving ability. Experimental research on drugs and driving through the use of “on-the-road” driving experiments and drug-using individuals as subjects has reduced many of the problems described earlier (Ferrara and Giorgetti, 1992; van Laar et al., 1993; Vermeeren et al., 1994). Studies to Link Driving Skills and Drug Impairment to Actual Road Crashes Epidemiological studies address the issue of incidence/prevalence of drug use among various subpopulations of drivers in the real world. The two major epidemiological research goals are descriptive and analytic. Descriptive epidemiology provides an indication of the extent or magnitude of the problem and, as such, guides experimental research by detecting the substances in persons involved in collisions. If a particular drug is not detected in crash-involved drivers, studying the impairing properties of the drug in a laboratory may be of little practical value (Friedel and Staak, 1993; Simpson and Vingilis, 1992). Descriptive epidemiology can also provide important trend data on the changing patterns of drug use and driving (Simpson and Vingilis, 1992). Analytic epidemiology seeks to determine which drugs are over-represented in crashinvolved drivers. The more methodologically valid approach is the case-control study in which crash-involved drivers are compared to non-crash-involved drivers matched for age, sex, location, and time of crash. Another related approach most typically used for fatal crashes is through responsibility or culpability analysis. Within a group of fatally injured drivers, the proportion of those judged responsible for crashes in the drug-positive group is compared to the proportion responsible in the drug-negative group. The resulting odds ratio helps in understanding whether the specific drug in question was related to the crash. Both these methods have strengths in that they use a systematic and consistent approach to gather data and control for extraneous factors that could affect outcomes. Limitations also exist with the interpretation of causality because the relationship established is one of association (Friedel and Staak, 1993; Simpson and Vingilis, 1992). Thus, although the relative risk of crash involvement for drug-using drivers could be higher than for non-drug-using drivers, other factors, such as a sensation-seeking personality, may explain the drug use and the high-risk driving. A methodological problem with culpability analysis relates to the difficulty in assessing 100% culpability to one driver in multivehicle crashes. A major challenge for any drug studies of crash-involved drivers relates to the measurement of psychoactive drugs in the crash victim (Ferrara, 1992; Simpson and Vingilis, 1992; Staak et al., 1993). Drug tests detect psychoactive drugs as well as metabolites (waste products) of the drug. Blood tests are better able to detect psychoactive drugs than urine and other tests because blood tests measure the drug circulating throughout the body, which can affect the brain and other tissues and cause impairment (Morgan, 1988). When drugs are used, whether ingested, injected, inhaled, etc., they pass into the blood by absorption. The drug is then metabolized by the liver and the metabolites are excreted in urine. Urine tests measure these byproducts of the liver’s metabolic processes (NIAAA, 1997).


380

The challenge is to link the presence or quantity of a particular drug to impairment because some drugs have long half-lives, meaning that metabolites can be found in the body well after initial use. For example, the half-life of THC carboxylic acid (cannabis metabolite) can range from 3 days to 20 days after ingestion (Dalén et al., 1997). Some exceptions exist, as with alcohol; most of the ingested alcohol is metabolized in the liver and turned into metabolites, but the small amount that remains unmetabolized permits alcohol concentration to be measured in breath and urine (NIAAA, 1997). An added challenge in establishing the link between drug use, impairment, and crash involvement relates to the growing evidence that longer term use of certain drugs can cause brain damage that could also impair driving ability. That is, the impairment may not be caused by the psychoactive properties of the drug per se, but rather by the brain damage and cognitive and other impairments that occur after longer use. The neurological damage associated with chronic alcohol abuse is well documented (NIAAA, 2001). Other drugs have also been associated with neurological damage. Recent animal studies have shown that methamphetamine damages brain cells involved in transport of dopamine, a chemical messenger involved in movement and cognition (Cadet et al., 1997; Deng et al., 1999; McCann et al., 1998). Research on long-term methamphetamine abusers who have abstained for 2 months have found reduction in dopamine transporters that appears to be linked to slowed motor skills and weakened memory (Volkow et al., 2001a, b)—two skills needed for safe driving (NIAAA, 1994). Thus, findings of increased crash rates of drivers who have consumed drugs compared to non-crashed control drivers are not able to disaggregate the cause of the increased crash risk because it could be due to the acute affect of the drug on the brain, the longterm effect of the drug causing neurological damage, or other factors, such as sensationseeking personality, that could cause drug use and risky driving. Another epidemiological approach involves survey research in which subjects are asked to report on drug use and driving. For example, they might be asked to indicate if they believe their drug use contributed to a traffic crash. Here again, any identified relationship between self-reported drug use and traffic collisions would be correlational and not causal. Survey methods are generally considered less valid than methods using drug tests because many people under-report or lie about their drug use. The pathogenesis of drug use may be helpful for understanding the impact of drugs on traffic crashes. The degree to which casual or infrequent use of drugs leads to dependence varies based on the type of drug (McKim, 1986). Also, withdrawal and hangover effects from these substances, which vary from drug to drug, may differentially affect the likelihood of adverse driving outcomes (Burns and Anderson, 1995; McKim, 1986). Another problem of determining crash risk associated with different drugs in epidemiological studies relates to multiple drug use, which is common among drug users. Multiple drug use makes it difficult to disentangle individual drug use risks and to assess possible additive affects. Long-term use can also affect personality characteristics, such as paranoia, which in turn could elevate accident risk (Coambs and McAndrew, 1994). A final element with regard to medicinal drugs relates to the therapeutic impact of the drug. Studies are needed in which the potential dangers of drugs, such as behavioral impairments, are weighed against the potential improvement in the condition for which the drug was prescribed (Moskowitz, 1985a). As Del Rió and Alvarez (1995a) summarize, drugs can cause deterioration of driving skills of varying degrees depending


381

on factors such as drug taking, abuse and dependence issues, expectations and factors associated with drug taking, multiple drug taking, and issues of dependence related to withdrawal and overdose. All of this is exceedingly hard to measure in order to determine the causal link between drug use and motor vehicle collisions. Therapeutic Categorization of Drugs and Pharmacological Effects Psychoactive drugs affect the CNS and interfere to various degrees with the psychophysical conditions necessary for driving (Del Rió and Alvarez, 1995a). Drug effects can vary among individuals. The effects are influenced by history of drug use (chronic or naïve user), tolerance, overall health, individual sensitivity to the drug metabolism, and other factors. Many drugs, especially those that affect the central nervous system, can impair driving. These include alcohol and illicit drugs, as well as therapeutic and over-the-counter medications. Many therapeutic drugs that are available with or without a prescription can have unwanted side effects that can impair driving performance. The three simplified categories of psychoactive drugs are depressants, stimulants, and hallucinogens (see DeLong, 2002, for a comprehensive and detailed description of drug psychopharmacology and effects). Depressants Depressants, which include alcohol, tranquillizers (e.g., benzodiazepine derivatives, barbiturates), narcotics (e.g., opiates, methadone) and inhalants, produce the predominant effects of relaxation, sedation, and a sensation of well-being (Del Rió and Alvarez, 1995a). Use of depressants can cause confusion, poor divided attention, slowed reaction times, memory effects, mental clouding, poor psychomotor skills, poor coordination, slurred speech, ataxia, disorientation, and decreased pulse and blood pressure (New Mexico Department of Health, 2002). Stimulants Stimulants include drugs such as amphetamines, cocaine, caffeine, and nicotine. The main effect of these drugs is to stimulate transmission at the synapses that use epinephrine, norepinephrine, dopamine, or serotonin as a transmitter (McKim, 1986). Although the drugs in this category have similar neurochemical effects, their behavioral effects can differ considerably. Use of stimulants can cause hypervigilance, anxiety, agitation, paranoia, self-absorption, obsessive activity, and elevated pulse and blood pressure (New Mexico Department of Health, 2002). Hallucinogens Hallucinogens include a wide range of drugs such as cannabis, LSD, psilocybin, mescaline, and many designer drugs (McKim, 1986). The main characteristic of hallucinogens is that they produce a distorted perception of reality. Although other types of drugs can produce altered perceptions, hallucinogens are categorized as such because they can produce these effects at low doses, without toxic effects. Thus, their effects are a


382

direct result of the drug and not an effect of drug poisoning (McKim, 1986). They can produce a sense of euphoria together with disturbances in perception and varying intensities of hallucinations, difficulties with paying attention, and a diminishing of reflexes and movement coordination (Del Rió and Alvarez, 1995a). Due to their CNS effects, all the drugs described here have the potential to impair skills needed for driving and, consequently, to be a causal factor in motor vehicle collisions. Various drugs can impair coordination, reaction time, judgment, tracking, divided attention, and perception. Extent of Drug Use in Collisions and Violations All three therapeutic drug categories—depressants, stimulants, and hallucinogens—have been found among drug-tested drivers. Depressants, particularly alcohol, are most prevalent. Although testing methods and rates differ among countries, the percentage of alcohol-related fatalities declined in the late 1980s and 1990s and ranged from 13.6% in 2000 in Germany (Kroj and Lerner, 2002) to 40% in 2000 in the U.S. (Sweedler, 2002). Studies on the proportion of drivers in traffic collisions testing positive for cannabinoids find prevalence rates between 2.7 and 13.9% (Stoduto et al, 1993; Soderstrom et al, 1995; Mercer and Jeffery, 1995). A review article reported rates of benzodiazepines generally ranging from 2 to 9.6% among arrested, injured, and fatally injured drivers (Woods et al., 1992). Studies in which drivers were tested for cocaine generally show that between 4 and 8% test positive (Stoduto et al., 1993; Soderstrom et al., 1995; Mercer and Jeffery, 1995). In his review of drugs and casualty driver studies, Lillsunde (1998) said that cannabinoids and benzodiazepines generally were the most frequently detected drugs, after alcohol, although in the U.S. benzodiazepines were less common and alcohol, cannabis, cocaine, PCP, and opiates were common. Amphetamines were most commonly detected in Nordic countries, while opiates were more frequent in central European countries (Vingilis and MacDonald, 2002). Similar patterns have been found for arrests for impaired driving. The challenges of drug testing (Verstraete, 2002) and the lack of per se legislation for drugs other than alcohol means that drugs other than alcohol were far less likely to be tested. The prevalence of drugs among suspected drug-impaired drivers is similar to that in the driver casualty studies. Cannabis or benzodiazepines (including tranquillizers) are typically the most commonly found drugs after alcohol (Dussault et al., 2002; Gerostamoulos et al., 2002). These studies of incidence of drug use among collision-involved drivers or violators do not provide information on the role that drugs may have had in the collision or violation. In order to determine whether the various drugs identified previously are causal factors in the collisions, it is necessary to review experimental studies that examine drug and dose response effects on skills performance, as well as analytic epidemiological casecontrol studies that examine the risk of crash involvement of various drugs. Current Evidence on Drug Impairment and Collision Risk Experimental psychopharmacological studies have shown adverse effects of many drugs on driving in controlled environments and driving-related performance variables (Del Rió


383

and Alvarez, 1995a, b; Ferrara, 1992; Moskowitz, 1985a; Moskowitz and Fiorentino, 2002). Based on current knowledge, certain classes of drugs capable of producing impairment in experimental studies are (Burns, 1996; Chesher, 1990; Ellinwood and Heatherly, 1985; Gengo and Manning, 1990; Linnoila and Seppala, 1985; Moskowitz, 1985b; Moskowitz and Fiorentino, 2000; O’Hanlon and Ramaekers, 1995; Roth and Roehrs, 1985; Simpson, 1985): • Alcohol • Anesthetics • Antidepressants • Antihistamines • Cannabinoids • Cardiovasculars • Hallucinogens • Hypnotics • Narcotics • Psychosometics • Sedatives • Solvents • Stimulants • Volatiles Experimental studies have also shown that some drugs produce large performance deficits, while other drugs produce minor changes in performance or even performance improvements. Few analytic epidemiology studies have been conducted in recent years in which different drugs are examined in the same study, using either the case-control method (Dussault et al., 2002; Ferrara et al., 1990; Honkanen et al., 1980, Marquet et al., 1998; Meulemans et al., 1996) or methods used to ascribe crash responsibility (Drummer, 1995). Both types of studies have found increased odds ratios for a variety of crashrelated measures, such as crash involvement, injury severity, and fatality for some drugs. For example, a recent case-control study conducted in Quebec, Canada, obtained blood and urine samples from 354 fatally injured drivers that were compared to samples of drivers who were stopped in two roadside surveys. The odds ratios, which constituted the ratio of the two rates of drug use, indicated that the odds of alcohol-positive drivers being fatally injured were 9.2. times as great as the odds for a sober driver. Similarly, the odds of cannabis, cocaine, and benzodiazepine drivers being fatally injured were 4.6, 12.2, and 4.2 times as great, respectively, as the odds for a sober driver (Dussault et al., 2002). An example of a culpability analysis study was conducted by Drummer (1995), who examined blood samples of driver fatalities in Australia. He found those with positive tests for alcohol were about 6.8 times more likely to be judged to be responsible for the car crash than those who tested negative. Those with opiates were 2.4 times more likely to be judged to be responsible for crashes; however, this was not significant. Those with stimulants or benzodiazepines had odds ratios of 1.4 and 1.0, respectively; however, these odds ratios were not significant. Interestingly, those testing positive for cannabis were less likely than those without drugs to be judged to be responsible for crashes (odds ratio=0.6).


384

Depressants, Impairment, and Risk of Collision Involvement Alcohol Numerous studies and reviews have been conducted examining the effects of alcohol on driving-related performance skills by different dosage amounts. Moskowitz and Robinson (1988) and Moskowitz and Fiorentino (2000, 2002) summarized more than 300 experimental studies from 1950 to 1997 that met acceptable scientific standards such as placebo treatments, statistical significance, and ability to determine BAC levels at performance skill testing periods. The results were indexed by BAC and performance skill and two analyses were conducted. The first analysis determined the lowest BAC at which impairment was reliably present for a particular performance skill and the second determined thresholds of impairment for different performance skills (Moskowitz and Fiorentino, 2000, 2002). The results of the three reviews found that critical flicker fusion and simple reaction time studies were most insensitive to the effects of alcohol, while tests of divided attention and of driving exhibited impairment by 0.01% BAC. The majority of studies measuring physiological sleep tendency found increased drowsiness by 0.02%; the majority of vigilance tests exhibited impairment by 0.04%. Moreover, the results of these different studies were congruent in that nearly all studies produced evidence showing impairment at nearly all of the BACs examined. However, greater variability of results was found for tracking, perception, visual function, cognitive task, psychomotor skill, and choice reaction time. This variability seemed to be due to the range of tasks tested within certain performance domains. Areas that remain to be explored are alcohol effects on risk taking, aggression, and motivation. However, Moskowitz and Fiorentino (2000, 2002) conclude that domains crucial to driving, such as vigilance, drowsiness, and divided attention, are impaired at BACs of 0.01% and higher. The studies linking alcohol use to crash involvement were conducted from the 1930s onward. Early culpability and case-control studies of alcohol crash risk compared the incidence of alcohol in collision and noncollision cases. For example, a study conducted in Toronto, Canada, in 1951 and 1952 had research assistants accompany special traffic police cruisers investigating collisions each week from Monday to Saturday (Lucas et al., 1955). For each driver involved in the collision, four or more noncollision drivers passing the scene of the crash in vehicles of the same vintage at approximately the same time provided breath samples with the understanding that such samples were for research purposes only. Their research showed an increasing exponential curve in which relative risk for crash involvement increased as BACs increased, with an accelerated rise at BACs above 0.08% (Lucas, 1955). Subsequent case-control studies have replicated this exponential curve (Borkenstein et al., 1964; Compton et al., 2002; Krüger et al., 1995; McLean and Kloeden, 2002). The relative risk of crash involvement at BACs of 0.05% is 1.4 times the risk of a sober driver, while at 0.10% the relative risk is 4.8 times. At a BAC of 0.16% (the average BACs of drivers arrested for driving under the influence or of crash-involved drivers), the relative risk of crash involvement is nearly 30 times the risk of a sober driver (Compton et al., 2002).


385

Benzodiazepines The second most commonly studied subclass of depressants on psychomotor performance are benzodiazepine derivatives (Coambs and McAndrew, 1994; Ferrara, 1987; Woods et al., 1992). Woods and colleagues (1992) have noted more than 20,000 papers on benzodiazepines since the 1960s and concluded that the degree of effect depends on the type of benzodiazepine and the types of psychomotor tasks used. Also, Berghaus and Guo (1995) conducted a meta-analysis of more than 1000 experimental studies of druginduced performance problems for prescribed drugs and also concluded that performance varied, depending on the type of benzodiazepine. Laboratory psychomotor tasks find less impairment with anxiolytics, while the hypnotic sedatives show more performance impairment (Hobi and Gerhard, 1993). Friedel and Staak (1993) in their review also found simpler tasks such as reaction time showed mixed results, while real on-the-road performance studies all reported some impairment. Similarly, van Laar et al. (1993), in their comparison of on-the-road and simulated driving tasks, found greater sensitivity in measuring impairment for on-theroad tests of benzodiazepine use than for simulator driving. Additionally, different benzodiazepines show differing degrees of impairment, with older benzodiazepines showing more behavioral side effects than newer ones (Del Rió and Alvarez, 1995b; Friedel and Staak, 1993). For example, O’Hanlon et al. (1995) found significant impairment of driving performance tests for on-the-road tests of benzodiazepines (diazepam, lorazepam) and benzodiazepine-like anxiolytics (alpidem, suriclone). Moreover, these authors found no placebo or drug differences between clinically anxious patients and healthy volunteers. Driving impairment has been found to be significant and equivalent to impairment found for BACs over 0.10% for the most commonly used benzodiazepines of lorazepam, diazapam, and flurazepam (Giorgetti et al., 2000; McBay, 2002; O’Hanlon et al., 1995). Several case-control studies have been conducted in which traffic records of groups with prescriptions for tranquilizers were compared to those of nonprescribed groups. The majority of these studies found increased likelihood of crash involvement for persons taking tranquilizers compared to drivers not taking tranquilizers, particularly for long half-life drugs (Hemmelgarn et al., 1997; Neutel, 1995; Ray et al., 1992; Skegg et al., 1979), although Leveille et al. (1994) found no relationship between benzodiazepine use and crashes. Narcotics Another subclass of depressants is narcotics, which include a wide class of drugs, from natural substances such as opium, morphine, and codeine to synthetically produced substances such as methadone and meperidine. Opiates depress neurotransmitter functioning, yet low to moderate doses of opiates do not greatly affect human performance (McKim, 1986). Tolerance to these drugs develops quickly, and first-time doses are much more likely to produce cognitive impairments than subsequent doses (Coambs and McAndrew, 1994). Some studies have been conducted on the impact of methadone on driving performance and crashes. Berghaus et al. (1993), in a case-control laboratory study on methadone users and non-drug-using controls, found that the methadone group exhibited poor performance on all psychophysical performance tests compared to the age-, sex-,


386

and education-matched controls. However, the study could not conclude whether the methadone or the personal characteristics of the users caused the poorer performance. Friedel and Berghaus (1995) concluded that heroin addicts treated with methadone are unfit to drive. By contrast, Chesher et al. (1995) found clients in a methadone maintenance program showed no evidence of performance deficits associated with driving. In another epidemiological study, driving records of treatment patients for opiates were compared to a matched control group (Blomberg and Preusser, 1974). The two groups did not differ significantly in terms of traffic citations; however, the opiate group had more equipment problems and documentation violations. In a literature review, Gordon (1976) found that narcotics by themselves were not a significant factor in poor driving. Stimulants Two commonly used stimulants are cocaine and amphetamines, but this category also includes such widely used substances as caffeine. Most laboratory studies have failed to find performance deficits associated with this class of drugs (Ferrara et al., 1994). In fact, performance improvements have been found in some studies for endurance tasks (Burns, 1993; Coambs and McAndrew, 1994; McKim, 1986). Laboratory evidence on the performance-enhancing effects of cocaine and other stimulants is inconclusive and researchers have suggested that subjects may perceive performance improvements while no real improvements have been noted (Fischman, 1987). Experiments have shown that five types of stimulants, including caffeine, did not differ in terms of their impact on performance (Mascord et al., 1997); however, as Burns (1993) states with regard to her research on cocaine, the effects are not unidimensional in direction in that the effects of overstimulation on performance may be qualitatively, as well as quantitatively, different from the effects of mild to moderate stimulation. Time of day, as well, appears to be a critical variable. Ellinwood and Nikaido (1987) write, “The realistic understanding of the nature of stimulant-induced effects requires the scientific perspective of all aspects of drug use, including acute and chronic dosing conditions, behavioral sensitization and tolerance and drug withdrawal.” No analytic epidemiological studies that tested only the role of stimulants in driving crashes were found. None of the analytic epidemiological studies previously reviewed found cocaine was associated with increased crash risk. Hallucinogens Hallucinogens include a wide array of drugs such as cannabis, LSD, psilocybin, and mescaline. Cannabis Cannabis is classified as an hallucinogen, but is often treated as a separate category because its effects are somewhat different and the prevalence of use is considerably higher than with the other drugs. A recent review of the experimental research on the immediate effects of cannabis use has been conducted by Schwenk (1998). Several


387

components of performance, such as attention and psychomotor skills, were examined. Most of the research has shown that subjects’ ability to concentrate and learn is reduced when they are under the influence of cannabis; however, long-term memory appears to be unaffected. In terms of psychomotor skills, the findings from research are mixed, especially in relation to driving simulation tasks. For example, one study found that marijuana had little effect on subjects’ ability to control a car, but impaired their ability to attend to peripheral stimuli (Moskowitz et al., 1976). Coambs and McAndrew (1994) have noted that studies in which deficits were not found involved more experienced marijuana users. For specific tasks, such as reaction time or spatial and temporal judgments, the preponderance of evidence suggests cannabis intoxication impairs performance, particularly in more complex tasks. Berghaus and Guo (1995) conducted a meta-analysis of experimental studies on the effects of cannabis on psychomotor skills and driving performance. More than 120 studies were collected, of which 60 fulfilled their methodological inclusion criteria. Their meta-analysis revealed that Smoking of marijuana causes to a more or less obvious extent impairment of every performance area connected with safe driving of a vehicle. Thus, performance areas such as tracking, psychomotor skills, reaction time, visual functions, attention, en-/decoding and performance in simulated or real driving experiments are involved. In each of these performance areas significant deterioration in dependency on the post smoking interval—that is to say on the THC concentrations in plasma—is found after smoking marijuana. THC-related impairment is concentrated within the first two hours after the beginning of the smoking procedure. Attention, tracking and psychomotor skills reveal the highest percentage of significant deterioration (p. 405). A recent meta-analysis of the experimental research literature on the impact of cannabis on performance was conducted by Smiley (1999). She noted that marijuana does impair one’s ability; however, drivers are quite aware of this impairment, which prompts them to slow down and drive more cautiously to compensate. However, in their review, Del Rió and Alvarez (1995a) stated that some studies found that psychomotor functions were disturbed up to 24 h after cannabis preparations were smoked and that the smokers were not aware of these deficits. In another review, Robbe and O’Hanlon (1993) examined laboratory and epidemiological studies of cannabis and traffic crashes and concluded that no clear relationship has been shown between marijuana smoking and seriously impaired driving performance or the risk of accident involvement. Later, they indicated that marijuana may well be the least harmful of the many psychotropic drugs. LSD, Psilocybin, and Mescaline With respect to other hallucinogens such as LSD, psilocybin, and mescaline, although less experimental research is available on their impairing properties, research shows these drugs cause major performance deficits (see McKim, 1986, for a review of experimental studies). No epidemiological studies were located in which the impact of hallucinogens


388

on crashes was examined. These drugs are particularly difficult to detect through urine analysis techniques. The prevalence of use of these drugs is generally restricted to youth and prevalence of its use is fairly low based on survey data. Conclusions One clear conclusion from a large number of laboratory and experimental studies is that the pharmacological properties and effects on psychomotor performance of the drugs examined in this chapter vary quite considerably. Laboratory studies have shown that alcohol, hallucinogens, and benzodiazepines impair performance, while the findings for methadone are mixed and stimulants may enhance performance. The adverse effects on driving skills of one drug, alcohol, have been well established. Impairment effects can be clearly demonstrated based on blood and breath concentrations in experimental and laboratory studies. Relative risk of crash involvement can also be determined based on blood and breath concentrations. The epidemiological evidence of other drugs as causal agents in collisions, however, is largely inconclusive; however, this evidence, along with experimental studies, allows one to draw tentative conclusions. Cannabis is typically the most commonly used illicit drug in most industrialized countries. Although experimental studies have shown that it moderately reduces performance, epidemiological research has failed to show it is a major contributor to traffic crashes. This conclusion was reached in several review articles on the topic and case—control studies have failed to find it a significant predictor of crashes. Two explanations for this finding are likely. First, drivers under the influence of marijuana may be reluctant to drive, thereby avoiding a risk of crashes. Second, if they do drive, they may pace their driving by driving more slowly and reducing the likelihood of collisions (Chesher, 1995). The research evidence on the role of benzodiazepines and traffic crashes is more extensive and more conclusive than that of cannabis research. Several studies have linked health records for benzodiazepine use and driving records (that show traffic crashes). Experimental studies show decreased performance with use and epidemiological studies generally show that benzodiazepine users are up to six times more likely than nonusers to be in crashes, although some studies failed to find it a significant risk factor. The risk of crashes likely varies, depending on the specific type of benzodiazepine, the half-life of the drug, and duration of use. The results of methadone studies are more variable and definitive conclusions cannot be drawn regarding the risks related to driving. Stimulants have not been shown to affect performance adversely in experimental studies and can actually enhance performance, particularly for endurance tasks. However, long-term use could affect personality characteristics, thus possibly increasing the likelihood of crashes. Also, these drugs can be addictive and withdrawal could enhance traffic risks. Epidemiological studies have not conclusively shown stimulant use to be a major contributor to crashes. Similar conclusions can be drawn for methadone, although performance deficits have been demonstrated in some experimental studies. Experimental studies show hallucinogens (i.e., LSD, mescaline) produce measurable reductions in performance; however, virtually no epidemiological research has been conducted to demonstrate that it represents a significant crash safety problem. Such


389

research faces severe methodological challenges because urine tests typically are not used to detect these substances. This review has been restricted to three categories of drugs. Other types of drugs may also impair performance and pose a traffic safety hazard. Also, this review focused on the independent effects of each drug and the probable impact on driving. Synergistic and dangerous effects on performance can occur when some drugs are used in combination with each other, particularly with alcohol. Better and more methodologically sound research in this area is required in order to understand the role of drugs other than alcohol in motor vehicle collisions. 13.7 Concluding Remarks This chapter has tried to provide insights into factors that need to be considered in the forensic evaluation of traffic crashes. Just as in any epidemiological study, examination of traffic crashes must proceed with a multidisciplinary team orientation to ensure that the important variables are properly considered. Human factors professionals are key members of such a team and it is their task to use their knowledge to ensure that the relevant human behaviors are evaluated and their role discerned in the sequence of events that led to the failures in the system and to crashes. This approach is exemplified by crash analyses by multidisciplinary accident investigation teams, which can arrive at crash sites, often within a few minutes of the event, and photograph, measure, and document the scene as well as talk with witnesses while the material is still fresh and unspoiled. By the time the human factors professional is brought into a case, the opportunities for getting to the scene while it is still fresh are rare. Other, secondhand materials must be relied on. Although some large trucks and buses and most railroad locomotives have event recorders that provide a few parameters such as speed and time (and, in the case of locomotives, also the time of brake and horn application and throttle position), most vehicles have no such records available. Such data could be of great value to researchers and forensic investigators and proposals have been made to equip more classes of vehicles with recorders. Human factors has traditionally been the discipline that examines a system from end to end including its human operators, equipment, ambient environment, and operating rules. The typical role of the human factors specialist is to help integrate the various system components into a harmonious and productive entity. This role is particularly germane to the forensic analysis of traffic crashes. Using human factors principles such as those enumerated in this chapter will invariably provide a more complete picture of crash causation, including salient interactions among possible causal factors. Thus, for example, the human factors analyst will not only determine if a driver made an error but also assess vehicle, environmental, personal, and organizational elements that may have induced the error. This, in turn, can provide a plaintiff with new theories for a complaint or a defendant with improved arguments to deflect liability. The human element is the one about which expert and nonexpert witnesses have been allowed to make subjective evaluations, sometimes with little foundation. This is probably because less is known about the behavior of drivers and other road users in events that lead to crashes than about most of the other factors involved. For example,


390

even though the effects of alcohol on many aspects of human performance and behavior have been studied extensively and there is a relatively good understanding of some doseresponse relationships for alcohol, many aspects of driving and the effects of alcohol are still not clear, particularly at low concentrations. Our best information probably comes from epidemiological studies and secondly from experimental studies. However, the latter bear the burden of generalization to a particular set of traffic circumstances and the former fail to pinpoint the nature of the performance affected. Although the information on alcohol and driving performance is far from complete, it is much more tenuous for other drugs such as cannabis, methadone, stimulants, and hallucinogens. Evidence of impairment for benzodiazepines, particularly for older and longer half-life drugs, is more conclusive. There is a great need for models of human performance to aid the forensic analyst. These may take the form of check lists or taxonomies of types of crashes, such as described for the evaluation and analysis of pedestrian and bicycle crashes or such as the concepts of positive guidance. Positive guidance has an interesting historical development because it arose as a result of two narrow-bridge crashes and the congressional hearings that followed to find solutions to “America’s narrow-bridge problem.” It soon became apparent that the principles being developed were not only appropriate for narrow bridges but equally applicable to the broad range of hazardous locations on the nation’s highways. Then Federal Highway Administrator Norbert T.Tiemann testified that, if we could not physically protect motorists at all hazardous locations, “we must give them enough information so that they can protect themselves.” Positive guidance is an integrated human factors/traffic engineering tool designed to enhance safety at hazardous locations. As an analytic tool, it enables forensic human factors and traffic engineering practitioners to determine whether highway-related information, i.e., the design and layout of the highway and its devices, plays any underlying role in contributing to crashes. At a higher level of analytical modeling and simulation of human performance are models such as night driving visibility models. They can at least be used as good first approximations of the ability of drivers to detect objects in darkness even when confronted by glare of other headlamps and street lamps. Models of driver headway change detection and relative velocity detection should also be useful tools in evaluating perceptual aspects of rear-end or head-on crashes. As the science of human factors continues to develop, more and better tools will become available that can aid the forensics human factors analyst and lead to a greater understanding of the numerous interactions among road users and their environment. The human factors professional will play a more important role in the future because of the growth in complexity of the traffic system and increasing demands on the human’s information processing. The need to understand these interactions and their positive and negative effects on driver and system safety performance, as well as the ability to explain them to a jury, will remain a continuing and growing challenge to human factors forensic practitioners.


391

13.8 Checklist of Main Factors to Consider in Traffic Collisions Primary factor Considerations Line of sight

Visibility

Audibility

Detection

Perception

Decision

Primary factor Response

Expectancy

Stress

Drugs/alcohol

Obstructions external to the vehicle Buildings, trees, vegetation, signs, other vehicles Obstructions internal to the vehicle and vehicle structure Vertical and horizontal alignment of road Illumination, street lighting, vehicle lighting, sun, moon Visual adaptation Glare Contrast Reflectivity, fluorescence Horns, bells, emergency vehicle sirens Background sounds, masking Hearing loss Stimulus Threshold Differential threshold Alertness, workload Movement Salience Recognition, encoding Ambiguity Interpretation Comprehension Exposure, learning Evaluation of information Evaluation of alternative responses Costs/effectiveness, effort Determination of response

Considerations Slow, stop, accelerate, reverse Turn, swerve Walk, run, look, listen Modulate control, turn on or off Familiarity Exposure Mental model Assumption, publicity Training Maturity, risk level acceptance Behavior of others Design-induced conditions Experience Knowledge Risk evaluation, training Personal, psychological Exposure


392

Time of consumption, time from end of consumption Type of drug Amount consumed Amount measured in body fluids, what and when measured Effects on performance and behavior Tolerance Driver/pedestrian characteristics Age, sex, height, weight Body fat Experience, prior driving record and violation record Training Hearing, vision and force/reach abilities Physical disabilities Familiarity with traffic regulations Fatigue Hours of service Sleep deprivation, sleep apnea Boredom, vigilance, attention External environment Rain, fog, snow, heat, cold, clear, dry, wet Available friction Road characteristics, horizontal and vertical alignment Speed limit Traffic density Traffic controls, signs, railroad crossing Road delineation Law enforcement, automated enforcement Signs and signals Guide signs, regulatory signs, warning signs Sign location, illumination, reflectorization, cleanliness Traffic signals, railroad crossing signals Pedestrian signals Construction zone signs Advance warning signs Warrants for signs and signals Roadway markings Edge line delineation, center line markings No-passing zones Construction zone markings Delineators Channelization devices, barricades Stop line, railroad crossing marking, turn lane

Primary factor

Considerations

In-vehicle Controls, displays Road noise, heater/air-conditioning blower Radio, cell phone, environment map displays, conversation, children, other passengers Temperature, fogged windows Insulation from outside warning-sound sources Isolation from road and other external conditions Ride conditions, vehicle loading Driver’s seated position Eye location Accessibility of controls/displays Vehicle Condition of windows, headlamps, other lamps, reflective devices Brake system: total or partial loss, wear Antilock brakes, air brake system lag Tire condition, inflation Steering system condition, free play, power-assist failure Control force requirements


393

References AASHTO, 1974, Highway Design and Operational Practices Related to Highway Safety (2nd ed.), Washington D.C.: American Association of State Highway Officials. AASHTO, 1994, A Policy on Geometric Design of Highways and Streets , Washington, D.C.: American Association of State Highway Officials. Alexander, G.J. and Lunenfeld, H., 1998, Positive Guidance Toolkit, Ontario Traffic Manual, Book 1C, Ministry of Transport, Ontario. Alexander, G.J. (Ed.) (undated), Positive guidance in Maryland, guidelines and case studies, Maryland State Highway Administration, Hanover, MD. Barfield, W. and Dingus, T., (Eds.) 1998, Human Factors in Intelligent Transportation Systems , Erlbaum: Mahwah, NJ. Becker J. and Mortimer, R.G., 1974, Further development of a computer simulation to predict the visibility distance provided by headlamp beams. University of Michigan, report UM-HSRI-HF74–26, Ann Arbor, MI. Beirness, D.J. and Simpson, H.M., 2002, The Safety Impact of Lowering the BAC Limit for Drivers in Canada. Traffic Injury Research Foundation, Ottawa, Canada. Bentley, J., 2002. An introduction to train brakes, http://www.train-dynamics.com/. Berghaus, G. and Guo, B-L., 1995, Medicines and driver fitness—findings from a meta-analysis of experimental studies as basic information to patients, physicians and experts, in Alcohol, Drugs, and Traffic Safety—T’95 , Eds. C.N.Kloeden and A.J.McLean, Adelaide, Australia, 295–300. Berghaus, G., Staak, M., Glazinski, R., Høher, K., Joo, S., and Friedel, B., 1993, Complementary empirical study on the driver fitness of methadone substitution patients, in Utzelmann, H.D., Berghaus, G., and Kroj, G. (Eds.), Alcohol, Drugs, and Traffic Safety—T’92 , Verlag TÜV Rheinland GmbH, Köhn, 120–132. Bhise, V.D., Farber, E.I., and McMahan, P.B., 1976. Predicting target detection with headlights. Transportation Research Record 611, Transportation Research Board, Washington, D.C. Blackwell, H.R., 1952, Brightness discrimination data for the specification of quantity of illumination. Illuminating Eng. , 47, 2, 602–609. Blomberg, R.D. and Preusser, D.F., 1974, Narcotic use and driving behavior. Accident Anal. Prev. , 6, 23–32. Blomberg, R.D., Fell, J.C., and Anderson, T.E., 1979, A comparison of alcohol involvement in pedestrians and pedestrian casualties. Proc. 23rd Annu. Conf. Am. Assoc. Automot. Med. , 23, 1– 17. Blomberg, R.D., Hale, A., and Kearney, E.F., 1974, Development of model regulations for pedestrian safety. Dunlap and Associates, Inc. final report to the U.S. Department of Transportation, National Highway Traffic Safety Administration, report no. HS-801 287, November 1974. Blomberg, R.D., Hale, A., and Preusser, D.R, 1984, Conspicuity for pedestrians and bicyclists: definition of the problem, development and test of countermeasures. Dunlap and Associates East, Inc. final report to the U.S. Department of Transportation, National Highway Traffic Safety Administration, contract no. DTNH22–80-C-07052. Blomberg, R.D., Leaf, W.A., and Jacobs, H.H., 1981, Detection and recognition of pedestrians at night. Paper presented at the Transportation Research Board Symposium on Conspicuity on the Highway, June 24, 1980. Transportation research circular no. 229. Borkenstein, R., Crowther, R., Shumate, R., Ziel, W., and Zylman, R., 1964, The Role of the Drinking Driver in Traffic Accidents . Bloomington: Indiana University. Broadbent, D.E., 1978, Decision and Stress . New York: Academic Press. Burns, M., 1993, Cocaine effects on performance. In Utzelmann, H.D., Berghaus, G., and Kroj, G. (Eds.), Alcohol Drugs, and Traffic Safety—T’92 , Verlag TÜV Rheinland GmbH: Köhn, 612– 619.


394

Burns, M. and Anderson, E., 1995, A Colorado validation study of the standardized field sobriety test (SFST) battery, Colorado Department of Transportation, Denver, Colorado. Burns, M., 1996, Driving skills impairment by prescription and OTC medications: the evidence from laboratory experiments. 7th Annual Meeting of the Transportation Research Board, Washington, D.C. Cadet, J.L., Ordonez, S.V., and Ordonez, J.V., 1997, Methamphetamine induces apoptosis in immortalized neural cells: protection by the proto-oncogene, bcl-2. Synapse , 25, 176–184. Chamberlain, E. and Solomon, R., 2002, The case for a 0.05% criminal law blood-alcohol concentration for driving. Unpublished report, June, 2002. Chesher, G.B., 1990, Qualitative differences between drugs demonstrated by psychomotor testing. In Perrine, M.W. (Ed.), Alcohol Drugs, and Traffic Safety—T’89 , Chicago: National Safety Council, 151–156. Chesher, G.B., 1995, Cannabis and road safety: an outline of the research studies to examine the effects of cannabis on driving skills and actual driving performance, in the effects of drugs (other than alcohol) on road safety, Incorporating Collected Papers. Melborne, LV North, Government Printer, 67–96. Coambs, R.B. and McAndrew, M.P., 1994, The effects of psychoactive substances on workplace performance, in Macdonald, S. and Roman, P. (Eds.), Drug Testing in the Workplace , New York: Plenum Press. Cook, T.D. and Campbell, D.T., 1979, Quasi-Experimentation Design and Analysis Issues for Field Settings , Boston: Houghton Mifflin. Cross, K. and Fisher, G., 1977, A study of bicycle/motor vehicle accidents: identification of problem types and countermeasure approaches. Anacapa Sciences, Inc. final report to the U.S. Department of Transportation, National Highway Traffic Safety Administration. Three volumes, report nos. DOT HS 803 (315, 316, 317). Compton, R.P., Blomberg, R.D., Moskowitz, H., Burns, M., Peck, R.C., and Fiorentino, D., 2002, Crash risk of alcohol impaired driving. In Mayhew, D. and Dussault C. (Eds.) Alcohol, Drugs and Traffic Safety (Montreal: Canada CD-ROM). Crowley, J. and Courtney, R., 1999, The relation between drug use, impaired driving and traffic accidents—the results of an investigation carried out for the European Monitoring Centre on Drugs and Drug Addictions (EMCDDA), Lisbon, Illicit Drugs in Road Traffic Seminar, Strasbourg, April, 1999. Dalén, P., Villén, T, and Stahle, L, 1997, Kinetics of 11-NOR-d-9-THC-carboxylic acid (THCCOOH) in urine. XXXV The International Association of Forensic Toxicologists (TIAFT) Annual Meeting. www.tiaft.org/tiaft97/proceedings/abstracts/posters/. Deese, J., 1959, Some problems in the theory of vigilance. Psych. Rev. , 62(5), 359–368. de Gier, J.J. and Laurell, H., 1992, Studies on man-machine interaction. Methodology in manmachine interaction and epidemiology on drugs and traffic safety. Experiences and guidelines from an international workshop. ARFI Mongraph Series 6. Padova, Italy: Addiction Research Foundation of Italy, 1992, pp.35–50. de Gier, J.J., 1993, Driving licences and known use of licit and illicit drugs. Maastricht: Institute for Human Psychopharmacology, University of Limburg. DeLong, J.V., 2002, Dealing with drug abuse: a report to the Ford Foundation, Schaffer Library of Drug Policy. Del Rió, M.C. and Alvarez, F.J., 1995a, Illegal drugs and driving. J. Traffic Med .. 23(1), 1–5. Del Rió, M.C. and Alvarez, F.J., 1995b, Prescribing medication for the driver: the role of health professionals. J. Traffic Med. , 23(3–4), 123–128. Deng, X., Ladenheim, B., Tsao., L.I., and Cadet, J.L., 1999, Null mutation of CFOS exacerbation of methamphetamine-induced neurotoxicity, J. Neurosci. , 19 (22), 10107–10115. Drummer, O.H., 1995, A review of the contribution of drugs in drivers to road accidents, in The Effects of Drugs (other than alcohol) on Road Safety, Incorporating Collected Papers, Melborne, LV North, Government Printer, 1–28.


395

Dussault, C., Brault, M., Bouchard, J., and LeMire, A.M., 2002, The contribution of alcohol and other drugs among fatally injured drivers in Quebec: some preliminary results. In Mayhew and Dussault (Eds.) Alcohol, Drugs and Traffic Safety—T’2002 (Montreal, Canada, CD-ROM). Ecker, H., Fischer, A., Vavryn, K., and Winkelbauer, M., 2000, Braking performance of motorcyclists: Results of a reliable test procedure. Proc. 3rd Intl. Motorcycle Safety Conf. , September 11–12, 2000, Munich, Germany. Ecker, H., Wasserman, J., Hauer, G., Ruspekhofer, R., and Grill, M., 2001, Braking deceleration of motorcycle riders. Proc. Hum. Element, Int. Motorcycle Safety Conf. , Orlando, March 1–4. Ellinwood, E.H. and Heatherly, D.G., 1985, Benzodiazepines, the popular minor tranquillizers: dynamics effect on driving skills. Accident Anal Prev. , 17(4), 283–290. Ellinwood, E.H. and Nikaido, A.M., 1987, Stimulant-induced impairment: a perspective across dose and duration of use. Alcohol, Drugs, and Driving , 3(1), 19–24. Farber, E.I. and Matle, C., 1989, PC detect: a revised version of the detect-seeing distance model. Transportation Research Board, Washington, D.C. Farber, E. and Silver, C.A., 1967, Conceptualization of Overtaking and Passing on Two-Lane Rural Roads, vol. 2. The Franklin Institute Research Laboratories, Report 1–193–11. Federal Highway Administration, 1975, Positive guidance in traffic control, Washington, D.C.: FHWA. Federal Highway Administration, 2000, Manual on Uniform Traffic Control Devices , Millennium Edition, Washington, D.C. Ferrara, S.D., 1987, Alcohol, drugs and traffic safety. Br. J. Addiction , 82, 871–883. Ferrara, S.D., Zancaner, S., Snenghi, R., and Berto, E, 1990, Psychoactive drug involvement in traffic accidents in Italy. In Alcohol Drugs, and Traffic Safety (Chicago: National Safety Council), 260–264. Ferrara, S.D., 1992, The status of methodological problems and the search for solutions. Methodology in man-machine interaction and epidemiology on drugs and traffic safety. Experiences and guide-lines from an international workshop. ARFI Monograph Series 6. Padova, Italy: Addiction Research Foundation of Italy, 5–31. Ferrara, S.D. and Giorgetti, R., 1992, Methodology in man-machine interaction and epidemiology on drugs and traffic safety. Experiences and guidelines from an international workshop. Padova, Italy: ARFI Research Monograph Series 6. Ferrara, S.D., Giorgetti, R., and Zancaner, S., 1994, Psychoactive substances and driving: state of the art and methodology. Los Angeles: UCLA Brain Information Service/Brain Research Institute. Fischman, M.W., 1987, Cocaine and the amphetamines, in Psychopharmacology: a Third Generation of Progress , H.Y.Meltzer (Ed.), New York: Raven Press. Friedel, B. and Staak, M., 1993, Benzodiazepines and driving performance, in Utzelmann, H.D., Berghaus, G., and Kroj, G. (Eds.), Alcohol, Drugs, and Traffic Safety—T’92 , Verlag TÜV Rheinland GmbH, Köhn, 539–545. Friedel, B. and Berghaus, G., 1995, Methadone and driving, in Utzelmann, H.D., Berghaus, G., and Kroj, G. (Eds.), Alcohol, Drugs and Traffic Safety , Verlag TÜV Rheinland GmbH, Köhn, 307– 310. Fry, G.A., 1954, Evaluating disabling effects of approaching automobile headlights. Highway Research Board Bulletin, 89, 38–42. Gengo, F.M. and Manning, C, 1990, A review of the effects of antihistamines on mental processes related to automobile driving. J. Allergy Clin. Immunol. , 86(6), 1034–1039. Gerostamoulos, J., McCaffrey, P., Drummer, O.H., and Odell, M., 2002, Drug profiles of apprehended drivers in Victoria. In Mayhew, D. and Dussault, C. (Eds.), Alcohol, Drugs, and Traffic Safety—T’2002 (Montreal, Canada, CD-ROM). Giorgetti, R., Zancaner, S., Tambuscio, S., and Ferrara, S.D., 2000, Effects of single doses of etizolam and lorazepam on psychomotor performance in healthy volunteers. In H. Laurell (Ed.), Alcohol, Drugs, and Traffic Safety—T’2000 (Stockholm, Sweden, CD-ROM: 605).


396

Gordon, N.B., 1976, Influence of narcotic drugs on highway safety. Accident Anal Prev. , 8, 1–7. Haddon Jr., W., 1970, On the escape of tigers: an ecological note. Am J. Pub. Health , 60, 2229– 2234. Hale, A., Blomberg, R.D., and Preusser, D.F., 1978, Experimental field test of the model ice cream truck ordinance in Detroit. Dunlap and Associates, Inc. final report to the U.S. Department of Transportation, National Highway Traffic Safety Administration, report no. DOT HS-803 410. Harkey, D.L., Mekemson, J., Chen, M.C., and Krull, K.A., 1999, Pedestrian and bicycle crash analysis tool (PBCAT), software and user’s manual. U.S. Department of Transportation, Federal Highway Administration, publication no. FHWA-RD-99–192. Haworth, N., Vulcan, P., Bowland, I., and Pronk, N., 1997, Fatal single vehicle crashes study, Monash University. Hemmelgarn, B., Suissa, S., Huang, A., Boivin, J.-F., and Pinard, G., 1997, Benzodiazepine use and the risk of motor vehicle crash in the elderly, JAMA , 278, 27–29. Hobi, V. and Gerhard, U., 1993, Effects of other than alcohol on psychomotor performance, in Utzelmann, H.D., Berghaus, G., and Kroj, G. (Eds.), Alcohol, Drugs, and Traffic Safety—T’92 , Verlag TÜV Rheinland GmbH, Köhn, 527–538. Hoffmann, E.R., 1968, Detection of vehicle velocity changes in car-following, Proc. Australian Road Res. Board , 4(1), 821–837. Hoffmann, E.R. and Mortimer, R.G., 1996, Scaling of relative velocity between vehicles. Accident Anal. Prevention , 28(4), 415–421. Holohan, R.D., Billing, A.M., and Murray, S.D., 1989, Nighttime photography—tell it like it is. Society of Automotive Engineers report 890730. Honkanen, R., Ertama, L, Linnoila, M, Alha, A., Lukkari, I., Karlsson, M., Kiviluoto, O., and Puro, M., 1980, Role of drugs in traffic accidents. Br. Med. J. , 281, 1309–1312. Home, J.A. and Barrett, P.R., 2001, Over-the-counter medicines and the potential for unwanted sleepiness in drivers: a review. Department of Transport, Local Government and the Regions, http://%20www.dft.gov.uk/Stellent/groups/dft_rdsafety/documents/page/dft_rdsafety_026928.h csp. Hunter, C.E., Lokan, R.J., Longo, M.C., White, J.M., and White, M.A., 1998, Appendices to: the prevalence and role of alcohol, cannabinoids, benzodiazepines and stimulants in non-fatal crashes. Forensic Science Department for Administrative and Information Services, South Australia. Hunter, W.H, Stutts, J.C., Pein, WE., and Cox, C.L., 1996, Pedestrian and bicycle crash types of the early 1990’s, U.S. Department of Transportation, Federal Highway Administration, publication no. FHWA-RD-95–163. Hurt, H.H., Ouellet, J.V., and Thom, D.R., 1981, Motorcycle accident cause factors and identification of countermeasures, University of Southern California, Report DOT HS-5–01160. Hyzer, W.G., 1993, The eye, brain and camera in nighttime accident reconstruction, Photo Electron. Imaging , 36(1), 8–9. Illuminating Engineering Society, 1972, IES Lighting Handbook . New York. Jehu, V.J., 1955, A method for evaluating seeing distances on a straight road for vehicle meeting beams. Trans. Illuminating Eng. Soc. , London, 20(2), 57–68. Johansson, G. and Rumar, K., 1971, Driver’s brake reaction time. Human Factors , 13(1), 23–27. Kapur, B., 1994, Drug-testing methods and interpretations of test results. In Macdonald and Roman, (Eds.), Drug-Testing in the Workplace (New York: Plenum Press). Knoblauch, R.L., 1977, Causative factors and countermeasures for rural and suburban pedestrian accidents: accident data collection and analysis. Biotechnology, Inc. final report, U.S. Department of Transportation, National Highway Traffic Safety Administration, report no. DOT-HS-802 266. Knoblauch, R.L., Moore, W., Jr., and Schmitz, P.R., 1976, Pedestrian accidents occurring on freeways: an investigation of causative factors. Biotechnology, Inc. final report to the U.S. Department of Transportation, contract no. DOT-HS-355–3-718.


397

Kroj, G. and Lerner, M., 2002, Alcohol related accidents in the Federal Republic of Germany— Status till 2000. In Mayhew, D. and Dussault, C. (Eds.), Alcohol Drugs, and Traffic Safety—T02 , Quebec, CD-ROM. Krüger, H.P., Kazenwadel, J., and Vollrath, M., 1995, Grand Rapids effects revisited: accidents, alcohol and risk. Schaffer Library of Drug Policy. http://www.druglibrary.org/schaffer/MISC/driving/%20s9p2.htm. Lerner, N., Ratte, D., and Walker, J., 1990, Driver behavior at rail highway grade crossings, report no. FHWA-SA-90–008, Washington, D.C. Leveille, S.G., Buchnew, D.M., Koepsell, T.D., McCloskey, L.W., Wolf, M.E., and Gagner, E.H., 1994, Psychoactive medications and injurious motor vehicle collisions involving older drivers. Epidemiology , 5, 591–598. Liebowitz, H.W., Owens, D.A., and Tyrrell, R.A., 1998, The assured clear distance ahead rule: implications for nighttime traffic safety and the law. Accident Anal. Prev. , 30(1), 93–99. Lillsunde, P., 1998, Commentary: alcohol-drug users and traffic accidents—a review on epidemiological studies. J. Traffic Medi. , 26(1–2), 5–10. Linnoila, M. and Seppala, T, 1985, Antidepressants and driving. Accident Anal. Prev. , 17(4), 297– 301. Lucas, G.H., Kalow, W., McColl, J.D., Griffith, B.A. et al., 1955, Quantitative studies of the relationship between alcohol levels and motor vehicle accidents. Alcohol and Road Traffic, Proc. 2nd Intl. Conf. On Alcohol and Road Traffic (1953), Garden City Press Cooperative, Toronto, Canada, pp. 139–142. Lucas, G.W., Kalow, W, McColl, J.D., Griffith, B.A., and Ward Smith, H., 1955, Quantitative studies of the relationship between alcohol levels and motor vehicle accidents. In Alcohol, Drugs and Road Traffic (Toronto: Garden City Press), 139–142. Lunenfeld, H. and Alexander, G., 1990, A users’ guide to positive guidance, 3rd ed., report no. FHWA-SA-90–017, Federal Highway Administration, Washington, D.C. Maes, V., Samyn, N., De Boeck, G., Willekens, M., and Verstraete, A., 2002, First experiences with the new law on DUID in Belgium: alcohol and medicines in drug negative cases. In Mayhew and Dussault (Eds.), Alcohol, Drugs, and Traffic Safety—T’2002 (Montreal, Canada, CD-ROM). Mann, R.E., MacDonald, S., Stoduto, G., Shaikh, A., and Bondy, S., 1998, Assessing the potential impact of lowering the legal blood alcohol limit to 50 mg% in Canada. Transport Canada, report TP 13321 E. Marquet, P., Delpla, P.-A., Kerguelen, S., Bremond, I., Facey, F., Garnier, M., Guery, B., Lhermitte, M., Mathé, D., Pelissier, A.-L, Renaudeau, C., Vest, P., and Seguela, J.-P., 1998, Prevalence of drugs of abuse in urine of drivers involved in road accidents in France: a collaborative study. J. Forensic Sci. , 43(4), 806–811. Mascord, D.J., Dean, A., Gibson, J., Springall, J., Starmer, G.A., and Christie, M.J., 1997, Comparison of five commonly abused stimulants which are used to reduce fatigue. In MercierGuyon (Ed.), Alcohol, Drugs, and Traffic Safety—T’97 (Annecy: France), 99–104. McBay, A.J., 2002, Drugs and impairment. Schaffer Library of Drug Policy, http://druglibrary.org/schaffer/%20MISC/driving/ddimp.htm. McCann, U.D., Wong, D.F., Yokoi, F., Villemagne, V., Dannals, R.F., and Ricaurte, G.A., 1998, Reduced striatal dopamine transporter density in sabstinent methamphetamine and methcathinone users: evidence from positron emission tomography studies with [11C]WIN-35, 428. J. Neurosci . 18(20), 8417–8422. McKim, W., 1986, Drugs and Behavior . Englewood Cliffs, NJ: Prentice Hall. McLean, J. and Kloeden, C., 2002, Alcohol, travelling speed and the risk of crash involvement. In Mayhew and Dussault (Eds.) Alcohol, Drugs, and Traffic Safety—T’2002 (Montreal, Canada, CD-ROM). Mercer, G.W. and Jeffery, W.K., 1995, Alcohol, drugs and impairment in fatal traffic accidents in British Columbia. Accident Anal Prev. , 27(3), 335–343.


398

Meulemans, A., Hooft, P., Van Camp, L, De Vrieze, N., Buylaert, W., and Wenning R., 1996, Belgian toxicology and trauma study: une étude portant sur la presence d’alcool, de médicaments et de drogues illicites chez des conducteurs victimes d’accidents de la route. Toxicological Society of Belgium and Luxembourg. Miller, N.D., Baumgardner, D., and Mortimer, R.G., 1974, An evaluation of glare in nighttime driving caused by headlights reflected from rearview mirrors. Society of Automotive Engineers report 740962, Warrendale, PA. Morgan, J., 1988, The “scientific” justification for urine drug screening. Schaffer Library of Drug Policy. http://www.druglibrary.org/schaffer/MISC/sciurin2.htm. Mortimer, R.G. and Becker, J., 1973, Development of a computer simulation to predict the visibility distance provided by headlamp beams. University of Michigan, report UM-HSRI-HF73–15, Ann Arbor, MI. Mortimer, R.G., 1974, Some effects of road, truck and headlamp characteristics on visibility and glare in night driving. Society of Automotive Engineers, report 740615, Warrendale, PA. Mortimer, R.G., 1975, Eye fixations of drivers in night driving with three headlamp beams. Proc. Soc. Photo-Optic. Eng. , 57, 81–88. Mortimer, R.G., 1976, Implications of some characteristics of drivers for brake system performance. Proc. Int. Conf. Braking Road Vehicles , Institution of Mechanical Engineers (London), Loughborough University, 187–195. Mortimer, R.G., 1981, Car following evaluation of braking deceleration signals. Society of Automotive Engineers, report 810188, Warrendale, PA. Mortimer, R.G., 1991, Visual factors in rail-highway grade crossing accidents. Proc. Hum. Factors Soc. 35th Annu. Meeting , 600–602. Mortimer, R.G., 1994, Oh! say, can you hear that train coming to the crossing. Proc. Hum. Factors Ergonom. Soc. 38th Annu. Meeting , 898–853. Mortimer, R.G., 1996, The effect of expectancy on visibility in night driving. Proc. 40th Annu. Meeting Hum. Factors Ergonom. Soc. , 506–510. Mortimer, R.G., 1998, Motorcycle braking controls—an ergonomic dilemma, Proc. Silicon Valley Ergonom. Conf. Exposition , ErgoCon ‘98, 102–106. Mortimer, R.G., 2001, The nighttime pedestrian accident: human factors issues and a case study. Proc. 45th Annu. Conf. Hum. Factors Ergonomics Soc. , 833–837. Mortimer, R.G. and Schuldt, R.C., 1980, Field test evaluation of gap acceptance of drivers as a function of motorcycle front lighting. Proc. Int. Motorcycle Safety Conf. , Washington, D.C., May 18–23, 945–954. Moskowitz, H., 1985a, Guest editor’s introduction. Accident Anal. Prev. , 17(4), 281–282. Moskowitz, H., 1985b, Marijuana and driving. Accident Anal. Prev. , 17(4), 323–345. Moskowitz, H., Hulbert, S., and Mcglothin, W.H., 1976, Marijuana: effects on simulated driving performance. Accident Anal. Prev. , 8(1), 45–50. Moskowitz, H. and Robinson, C.D., 1988, Effects of low doses of alcohol on driving-related skills: a review of the evidence. Report no. DOT HS 807 280. Washington, D.C.: National Highway Traffic Safety Administration, SRA Technologies, Inc. Moskowitz, H., Burns, M., Fiorention, D., Smiley, A., and Zador, P., 2000, Driver characteristics and impairment at various BACs. National Highway Traffic Safety Administration, Washington, D.C., DOT HS 809 075. Moskowitz, H. and Fiorentino, D., 2000, Driver Characteristics and Impairment at Various BACs. U.S. Department of Transportation. National Highway Traffic Safety Administration, DOT HS 809 075, August. Moskowitz, H. and Fiorentino, D., 2002, A review of experimental studies of low BAC effects on skills performance. In Mayhew, D. and Dussault, C. (Eds.), Alcohol Drugs, and Traffic Safety— T’2002 , (Montreal, Canada, CD-ROM) National Committee on Uniform Traffic Laws and Ordinances, 2000. Uniform Vehicle Code . Alexandria, VA.


399

National Highway Traffic Safety Administration (NHTSA), 1999, Horizontal gaze nystagmus state case law summary, http://www.nhtsa.dot.gov/. National Institute of Alcohol Abuse and Alcoholism (NIAAA), 1994, Alcohol-related impairment. Alcohol Alert , 25, PH 351, July 1994. NIAAA,1997, Alcohol metabolism. Alcohol Alert , 35, PH 371, January 1997. NIAAA, 2001, Cognitive impairment and recovery from alcoholism. Alcohol Alert , 53, July 2001. National Safety Council, 2001, Injury Accident Facts , 2001 ed. Chicago: NSC. National Safety Council, 2002, International Accident Facts , 3rd ed. Chicago: NSC. Neutel, C.I., 1995, Risk of traffic accident injury after a prescription for benzodiazepine. Ann. Epidemiol , 5(3), 239–244. New Mexico Department of Health, 2002, Toxicology bureau fact sheet: drug-impaired drving. www.sld.state.nm.us/drug. Noy, I.Y., 1997, Ergonomics and Safety in Intelligent Transportation Systems , Erlbaum: Mahwah, NJ. O’Hanlon, J.F., 1993, Ten ways for physicians to minimize the risk of patients causing traffic accidents while under the influence of prescribed psychoactive medication. Primary Care Psychiatry , 1,77–85. O’Hanlon, J.F. and Ramaekers, J.G., 1995, Antihistamine effects on actual driving performance in a standard test: a summary of Dutch experience, 1989–94. Allergy , 50, 234–242. O’Hanlon, J.F., Vermeeren, A., Uiterwijk, M.M.C., Van Veggel, L.M.A., and Swijgman, H.F., 1995, Anxiolytics’ effects on the actual driving performance of patients and healthy volunteers in a standardized test. Neuropsychobiology , 31, 81–88. Olson, P.L. and Sivak, M., 1983, Improved low beam photometries. University of Michigan, report UMTRI 83–9, Ann Arbor, Michigan. Olson, P.L., Cleveland, D.E., Fancher, P.S., Kostyniuk, and Schneider, L.W., 1984, Parameters affecting stopping sight distance, University of Michigan Transportation Research Institute Report UMTRI-84–15. Olson, P.L., 1988, Minimum requirements for adequate nighttime conspicuity of highway signs. University of Michigan report: UMTRI-88–8, Ann Arbor, MI. Perchonok, K., 1972. Accident Cause Analysis. Cornell Aeronautical Laboratories, Report ZM5010-V-3. Post, T, Alexander, G., and Lunenfeld, H., 1981, A users’ guide to positive guidance, 2nd ed., report no. FHWA-TO-81–1 Federal Highway Administration, Washington, D.C. Post, T, Robertson, D., Price, H., Alexander, G., and Lunenfeld, H., 1977, A users’ guide to positive guidance, Federal Highway Administration, Washington, D.C. Ray, W., Fought, R., and Decker, M., 1992, Psychoactive drugs and the risk of injurious motor vehicle crashes in elderly drivers. Am. J. Epidemiol , 136, 873–883. Robbe, H. and O’Hanlon, J., 1993, Marijuana and actual driving performance. U.S. Dept. Transportation, National Highway Traffic Safety Administration, 9. Roper, V.J. and Howard, E.A., 1938, Seeing with motorcar headlights. Trans. Illumination Eng. Soc. , 33, 417–427. Roth, T. and Roehrs, T, 1985, Determinants of residual effects of hypnotics. Accident Anal. Prev. , 17(4), 291–296. Rowan, NJ. and Walton, N.E., 1976, Lighting of traffic facilities. In Baerwald, J. (Ed.), Transportation and Traffic Engineering Handbook , (Englewood Cliffs, NJ: Prentice Hall), 912–943. Rumar, K. and Berggrund, V., 1973, Overtaking performance under controlled conditions, Proc. 1st Intl. Conf. on Driver Behavior Res. , Zurich, Switzerland. Schwenk, C., 1998, Marijuana and job performance: comparing the major streams of research. J. Drug Issues , 28(4), 941–970. Simpson, H.M., 1985, Polydrug effects and traffic safety. Alcohol Drugs Driving , 1(1–2), 27–44.


400

Simpson, H.M. and Vingilis, E., 1992, Epidemiology and special population surveys, in Ferrara, S.D. and Giorgetti, R. (Eds.), Methodology in Man-Machine Interaction and Epidemiology on Drugs and Traffic Safety , Centre of Behavioural and Forensic Toxicology: Padova, Italy. Skegg, D.C.G., Richards, S.M., and Doll, R., 1979, Minor tranquillisers and road accidents. Br. Med. J. , 1, 97–919. Smiley, A., 1999, Marijuana: on-road and driving-simulator studies. In H.Kalant, W.Corrigall, W.Hall, and R.Smart (Eds.), The Health Effects of Cannabis (Toronto, Ontario Addiction Research Foundation), 173–191. Snyder, M.H. and Knoblauch, R.L., 1971, Pedestrian safety: the identification of precipitating factors and possible countermeasures. Operations Research, Inc., final report U.S. Department of Transportation, National Highway Traffic Safety Administration, report no. DOT HS-800 403. Soderstrom, C.A., Dischinger, PC., Kerns, T.J., and Trifillis, A.L., 1995, Marijuana and other drug use among automobile and motorcycle drivers treated at a trauma center. Accident Anal. Prev. , 27(1), 131–135. Staak, M., Kaferstein, H., and Sticht, G., 1993. Importance of quantitative estimation of opiates and cannabinoids for determination of driving performance. In Utzelmann, H.D., Berghaus, G., and Kroj, G. (Eds.), Alcohol, Drugs and Traffic Safety—T’92 . H-D. Verlag TÜV Rheinland GmbH, Köhn, 490–496. Stoduto, G., Vingilis, E., Kapur, B., Shen, W., McLellan, B., and Liban, C., 1993, Alcohol and drug use among motor vehicle collision victims admitted to a regional trauma unit: demographic, injury and crash characteristics. Accident Anal. Prev. , 25(4), 411–420. Stokes, A. and Kite, K., 1999, Grace under fire: the nature of stress and coping in general aviation. Chapter 4 in O’Hare (Ed.), Human Performance in General Aviation ,, Aldershot: Ashgate. Summala, H., 1981, Driver/vehicle steering response latencies. Hum. Factors , 23(6), 683–692. Swain, A.D. and Guttman, H.E., 1983, Handbook of Human Reliability Analysis with Emphasis on Nuclear Power Plant Applications , Sandia Laboratories, NUREG/CR-1278, U.S. Nuclear Regulatory Commission. Sweedler, B., 2002, Worldwide trends in drinking and driving: has the progress continued? In Mayhew and Dussault (Eds.) Alcohol, Drugs, and Traffic Safety—T’2002 (Montreal, Canada, CD-ROM). Triggs, T.J. and Harris, E.C., 1982, Reaction time of drivers to road stimuli. Monash University, report HFR-12. USDOT, 1978, Railroad-Highway Grade Crossing Handbook , Federal Highway Administration, FHWA-TS-78–214, U.S. Department of Transportation, Washington, D.C. USDOT, 1998, Traffic safety facts 1997, National Highway Traffic Safety Administration, U.S. Department of Transportation, Washington, D.C. USDOT, 2001, Traffic safety facts, 2000, National Highway Traffic Safety Administration, U.S. Department of Transportation, Washington, D.C. van Laar, M.W., van Willigenburg, A.P.P., and Volkerts, E.R., 1993, Over-the-road and simulated driving: Comparison of measures of the hangover effects of two benzodiazepine hypnotics. In Utzelmann, H.D., Berghaus, G., and Kroj, G. (Eds.), Alcohol, Drugs, and Traffic Safety—T’92 , Verlag TÜV Rheinland GmbH, Köhn, 672–677. Vermeeren, A., de Gier, J.J., and O’Hanlon, J.F., 1994, Methodological guidelines for experimental research on medicinal drugs affecting driving performance: an international expert survey. J. Traffic Med. , 22(4), 173–174. Verstraete, A.G., 2002, Roadside drug testing: the results of the ROSITA Project. In Mayhew and Dussault (Eds.), Alcohol, Drugs, and Traffic Safety—T’2002 (Montreal, Canada, CD-ROM). Vingilis, E.R. and MacDonald, S., 2002, Review: drugs and traffic collisions. Traffic Injury Prev. , 3(1), 1–11. Volkerts, E.R. and van Laar, M.W., 1993, A methodological comparative study of over-the-road and simulated driving performance after nocturnal treatment with lormetazepam 1 mg and


401

oxazepam 50 mg. In Utzelmann, H.D., Berghaus, G., and Kroj, G. (Eds.), Alcohol, Drugs, and Traffic Safety—T’92 , Verlag TÜV Rheinland GmbH, Köhn, 664–671. Volkow, N.D., Chang, L, Wang, G.-J., Fowler, J.S., Leonido-Yee, M., Franceschi, D., Sedler, M.J., Gatley, S.J., Hitzemann, R., Ding, Y.-S., Logan, J., Wong, C., and Miller, E.N., 2001a, Association of dopamine transporter reduction with psychomotor impairment in methamphetamine abusers. Am. J. Psychiatry 158(3), 9414–9418. Volkow, N.D., Chang, L., Wang, G.-J., Fowler, J.S., Franceschi, D., Sedler, M.J., Gatley, S.J., Miller, E., Hitzemann, R., Ding, Y.-S., and Logan, J., 2001b, Loss of dopamine transports in methamphetamine abusers recovers with protracted abstinence. J. Neurosci. , 21(23), 9414– 9418. Welford, A.T., 1973, Stress and performance. Ergonomics , 16, 567–580. Willekens, M., Samyn, N., De Boeck, G., Maes, V., and Versraete, A., 2002, First experiences with the new law on DUID in Belgium: Field results and plasma levels of illicit drugs. In Mayhew, D. and Dussault, C. (Eds.), Alcohol Drugs, and Traffic Safety—T02. D , Quebec, CD-ROM. Wolf, E., 1960, Glare and age. Arch. Opthalmol , 64, 502–514. Woods, J.H., Kats, J.L., and Winger, G., 1992, Benzodiazepines: use, abuse and consequences. Pharmacol. Rev. , 44, 151–347. Worm, K., Steentoft, A., and Toft, J., 1996, Drugs and narcotics in Danish drivers, J. Traffic Med. , 24(1–2), 39–42. Xie, X., Rehm, J., Single, E., and Robson, L., 1996, The economic costs of alcohol. Tobacco and illicit drug abuse in Ontario: 1992. Toronto: Addiction Research Foundation.

Further Information Alexander, G.J. and Lunenfeld, H., 1986, Driver expectancy in highway design and traffic operations, U.S. Department of Transportation, report FHWA-TO-86–1. Deese, J., 1955, Some problems in the theory of vigilance. Psych. Rev ., 62(5), 359–368 Forbes, T.W. (Ed.), 1972, Human Factors in Highway Traffic Safety Research , New York: John Wiley & Sons. Green, D. and Swets, J.A., 1966, Signal Detection Theory and Psychophysics , New York: John Wiley & Sons. Henderson, R.L. (Ed.), 1987, Driver performance data book, Vector Enterprises, report DOT-HS807–121, U.S. Department of Transportation, NHTSA. Johansson, G. and Rumar, K., 1971, Drivers’ brake reaction times. Hum. Factors , 13(1), 23–27. Mace, D,. Garvey, P., Schwab, R., and Adrian, W., 2002, Countermeasures for reducing the effects of headlight glare. Washington, D.C.: AAA Foundation for Traffic Safety. National Safety Council, 2002, Injury Facts 2001 Edition . Chicago: Illinois. Olson, P.L., 1996, Forensic Aspects of Driver Perception and Response , Tucson, AZ: Lawyers and Judges Publishing Co. Peters, G.A. and Peters, B.J. (Eds.), 1988, Automotive Engineering and Litigation , New York: Garland Law. Sens, M.J., Cheng, P.H., Wiechel, J.F., and Guenther, D.A., 1989, Perception/reaction time values for accident reconstruction, Society of Automotive Engineers report 890732. Shinar, D., 1978, Psychology on the Road: the Human Factor in Traffic Safety , New York: John Wiley & Sons. Staplin, L., Lococo, K., Byington, S., and Hartley, D., 2001, Guidelines and recommendations to accommodate older drivers and pedestrians, Federal Highway Administration, report FHWARD-01–051. Wickens, C., 1984, Engineering Psychology and Human Performance , Columbus, OH: Merril.


402

See Also (complete references are in reference list): Compton et al. (2002); Beirness and Simpson (2002); Chamberlain and Solomon (2002); Mann et al. (1998); and Moskowitz and Fiorentino (2000) for reviews on the effects of low doses of alcohol on driving-related skills Del Rió and Alvarez (1995a) for a general review on the influence of illegal drugs on driving within the context of European Community regulations Ferrara et al. (1994) for a review of laboratory and epidemiological studies of psychoactive substances and driving Home and Barrett (2001) for a review of over-the-counter medicines Hunter et al. (1998) for a detailed, tabulated review of samples, methods, and findings of epidemiological and experimental studies Krüger et al. (1995) and Compton et al. (2002) for alcohol dose-response crash risk Lillsunde (1998) for a review of epidemiological studies McBay (2002) and DeLong (2002) for a detailed review of drugs and their effects Woods et al. (1992) for a comprehensive review of benzodiazepines use and consequences, including averse driving outcomes

Additional Literature of Interest: Horizontal Gaze Nystagmus (HGN) Oct 2, 2002 Horizontal Gaze Nystagmus State Chart Summary http://www.nhtsa.dot.gov/people/injury/enforce/nystagmus/app_c.html Oct 2, 2002 Melethil S., Breath tests for blood determination ratio: partition ratio, http://www.forensicevidence.com/%20site/Biol_Evid/Breath_Tests.html, Oct 2, 2002 http://www.druglibrary.org/schaffer/MISC/sciurin2.htm http://www.druglibrary.org/schaffer/MISC/driving/ddimp.htm http://www.druglibrary.org/schaffer/Library/studies/dwda/staffl.htm http://www.toxicologyassociates.com/toxicsubreg3.htm

14 Estimating Driver Response Times Jeffrey W.Muttart Accident Dynamics Research 0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

14.1 Introduction This chapter addresses automobile driver response times and the factors that influence them. Previous research has reported driver response times from 0.5 to 10 sec for various tasks. This chapter will offer reasons for different response time results. Muttart (2003a, b, 2004) found that different results can be explained and quantified by methodology and substantive variables. Thus, a mean response time for various scenarios can be estimated. The precision of such estimates will also be addressed. First, we must clarify and explain terms. Perception in the case of a driver’s response is something beyond simple visibility; it is the organization and interpretation of sensations (visual, tactile, or auditory). Before an investigator can estimate a driver’s “perception”—response time, he should understand the processes that may delay the organization and interpretation of what may be visible (or heard or felt). The terms driver response time, brake reaction time, perception—reaction time, and perception—response time have been used almost interchangeably at times. This inappropriate use of the terms could lead to confusion. The reader should understand that a driver’s response does not follow an orderly step-by-step process. Research has shown that even the response may influence the perception (Bekkering and Neggers, 2002) and that drivers experience response inertia when the perception complexity likely influences the motor response time (Muttart, 2004). However, as a means to compare responses, Figure 14.1 was prepared to show the traveling distance components of a driver’s response.

FIGURE 14.1 The distance components of a driver’s response. (Reprinted with permission from SAE paper 2003–01–0885. © 2003 SAE International.)


404

Driver response is a generic term used to describe any portion of a driver response time. Driver response time encompasses from first visibility (“something” may be visible but nothing has been recognized as a hazard) to first vehicle response. Perceptionresponse time is from first perception of an immediate hazard until first vehicle response. Perception-reaction time is from first perception of a hazard until first reaction (foot off the accelerator) by the driver, and brake reaction time or brake response time is the time from first perception of a hazard until the foot touches the brake (there is no effective deceleration at that point). As Figure 14.1 shows, the investigator must assure that the estimated response time he is using is applicable to the particular situation. Furthermore, if relying upon a research study, the investigator should make sure that he knows how the response time was measured and should not rely upon the terminology used by an author. Also, in most situations, before a driver can perceive an object as an immediate hazard, he must attend to it. Therefore, we should address attention. 14.2 Attention vs. Inattention Police reports frequently cite “inattention” as a cause in many collisions without any support for such a conclusion. Some investigators have supported “inattention” based upon a lack of skid marks on the road or because the driver was not able to recall seeing traffic that the investigator (with hindsight bias) may believe is a salient object. An investigator must understand what “attention” is before being able to evaluate a driver’s response. In simple terms, attention is a narrowing of focus. If investigators use the word focus rather than the word attention, a great many misunderstandings may be avoided. Any driver who is conscious is attentive. The question that must be asked is what was the driver attentive to (focused upon)? There are appropriate and inappropriate causes for inattention to a road hazard. An example of an inappropriate attention is a driver talking on a cellular telephone, eating fast food, and steering with his legs. A driver focused upon a possible path-intruding vehicle to his left will not be focused upon a previously hidden pedestrian who emerges suddenly from between two vehicles on the right and is struck. Other examples of appropriate inattention would be the allocation of attention to a vehicle in the immediate area at the expense of a pedestrian or other traffic that is downstream and not an immediate hazard. Summala and colleagues (1998) performed a study of drivers who were required to respond to the sudden slowing of a lead vehicle when focused on the instrument panel or rear view mirror. These researchers found that drivers responded much more slowly to the lead vehicle when the driver’s attention was allocated to these areas. A safe driver must navigate, guide, and control his vehicle (Alexander and Lunenfeld, 1975) and monitor the environment and vehicles in the area. Therefore, instances of acceptable inattention will occur. It is up to the investigator to attempt to differentiate between acceptable and unacceptable inattention. Consequently, an investigator must consider several things before evaluating a driver’s response time or response choice—basically, 10 considerations that account for the differences between what the investigator sees and knows and what the driver saw and knew. These considerations can be recalled by the acronym IN CAR CREED

Estimating driver response times

405

(immediacy, neutralized, contrast, anticipation, recall, context and priming, relative velocity, eccentricity, expectancy, driving tasks). 1. Immediacy. Was there any immediate reason to respond before the driver perceived the hazard or was the object a secondary stimulus to one being mentally processed (Batchelder et al., 2003; Harrell, 1994)? A sight line of 1 km has no bearing upon a collision if the object was a hazard for only 25 m before the collision. 2. Neutralized. Was there reason to believe that the driver may have mentally “neutralized” the stimulus by an earlier corrective action (Fuller, 1984)? (For example, a driver slowed and moved away slightly in response to a roadside pedestrian, but the pedestrian suddenly ran into the road when there was no time to respond.) 3. Contrast. Does the object to be responded to contrast with its background when not being focused upon? Is the object easily identifiable and intense enough to grab a driver’s attention away from that to which he is already attending (Mack, 2003; Most et al., 2001; Simon and Chabris, 1999; Muttart, 2003)? This is particularly true in situations in which the driver may not be adapted to the ambient level of light or if the ability to discriminate between an object and its background is degraded significantly by glare. 4. Anticipation. Drivers come to expect certain things with time. Hole and Tyrell (1995) conducted a laboratory study and found that when slides depicting traffic were shown to subjects rapidly, their responses to a motorcycle without a headlight decreased as the likelihood of a motorcycle having a headlight turned on increased. The subjects developed a “set” to allow for a quicker search technique by looking for headlights rather than motorcycles in that laboratory study. Therefore, (right or wrong) if a driver learns that it is more efficient to look for headlights, he is more likely to miss an approaching vehicle without headlights. Furthermore, if he knows the stimulus and knows the appropriate response, response time will improve. 5. Recall. Do not evaluate the driver’s recall of traffic or signs but ask (through the investigation) if the driver’s response was appropriate (Fisher, 1992; Carpenter, 2001; Louma, 1988; Resnick, 2004). 6. Context and priming. When a driver is alerted to what he is going to experience and what it means to him, then performance will likely improve. Simply because there may have been a sign or the lack of a sign may not in and of itself be evidence that a driver will know what is coming and what to do about it. Additionally, a driver may not be able to determine that the stray light or flicker of movement seen today is relevant to safe driving because similar nondescript lights have never been a problem in the past. 7. Relative velocity detection threshold. Drivers will not be able to detect the rate of closure on a lead vehicle without other cues until the subtended arc angle of the object expands in the observer’s field of view at a rate exceeding 0.003 rad/sec (Hoffman and Mortimer, 1996; Todosiev, 1965). Also, most road lines are 4 in. (1.6 cm) wide. A person with 20/20 visual acuity can read a 4-in. (1.6 cm) letter from approximately 220 ft (70 m). Therefore, it may be difficult to determine exact tire positions relative to a center line from distances greater than this. 8. Eccentricity. Considering that his central vision is a very narrow cone reported to be no more than a few degrees in each direction from straight ahead, it is unlikely that a


406

driver will be looking directly toward the hazard. Thus, the investigator should assume that the hazard had to be detected in the driver’s peripheral visual field. 9. Expectancy. This has been cited as a primary reason for slow or inappropriate responses (Alexander and Lunenfeld, 1986; Green, 2000; Dewar and Olson, 2002). “Unexpected” could mean needing to respond to a satellite falling from the sky or it could mean not knowing when a traffic signal will change. Investigators should not evaluate the subjective term of “expectancy,” but should address the manner in which expectancy influences a response in a more direct and objective way. A single correction factor (Dewar and Olson, 2002, p. 372; Roper and Howard, 1938) to account for expectancy is not scientifically justifiable. “Lack of expectation” may have a different influence in each case, based upon current research (Muttart, 2003a, b). If a driver does not expect a hazard, he may experience any or all of the following: • Eccentricity: looking in a different direction • Context: looking for or expecting a different hazard • Anticipation: may not know what to look for or where to look and may not be adapted to the necessary level of lighting • Operant conditioning: may be anticipating a different response and misinterpret the situation or pattern that he is seeing • Response complexity: may not have a conditioned response available (Bekkering and Neggers, 2002) or may interpret the traffic pattern as something that he has seen before when in fact it is not • Contrast: the target may not be easily distinguishable when compared to its background and other objects in the area • Number of stimuli: despite several things to look at, the driver who has expectancy can weed through all the visual noise and focus specifically upon the hazard, while most real-life drivers do not have this luxury 10. Driving task. Drivers do not view the scene as a stationary observer. They are viewing an ever-changing scene and typically do not gaze in any one direction. The typical river changes fixation location approximately one to two times per second (OCED, 1970; Lee et al., 2003). They also must devote mental resources to the tasks of navigation, guidance, and control of their vehicle (Alexander and Lunenfeld, 1975). Lassiter et al. (Lassiter, 2002; Lassiter et al., 2002) found that people attribute unwarranted influence to a stimulus simply because it is more noticeable (to them) than other available stimuli. They referred to this as the “illusory-causation phenomena,” which could cause prejudicial effects relative to how people evaluated certain types of evidence. Objects that stand out in the visual field or are the focus of attention are more likely to be judged as originators of the event, even when no objective basis for such a conclusion exists. Therefore, it is common for postcrash investigators to suffer from the illusory-causation phenomena. To avoid this, the investigator should understand and consider the 10 considerations listed earlier. More than a few expert witnesses have been known to use a rule-of-thumb estimate for expectancy based upon the research by Roper and Howard (1938). These researchers conducted a test in which they placed a mock pedestrian in the path of the subjects, noted their response time, backed up and again approached the “pedestrian,” and noted the time necessary to respond to the pedestrian when they knew he was there. Roper and Howard


407

may not have measured the effects of “expectancy” but instead measured the effects of their methodology; they were measuring perception-response, so someone had to determine when the pedestrian was perceivable. Expectancy is not an on or off switch; rather, there are degrees of expectancy. Muttart (2004) found that as the complexity of the situation increased, braking latency (limb movement) increased, and brake force decreased. Therefore, at least part of Roper and Howard’s results can be explained by their methodology. Furthermore, their findings could not be corroborated in a literature review and the research by Muttart has shown that a single correction factor is not justifiable. Muttart (2000) indicated that the acronym of CASE CR might be used to understand the probability of detection. The acronym stands for contrast, anticipation (or bias toward that object), strength of sensory stimulus, eccentricity, cognition, and response complexity. Muttart based his acronym upon Bundensen’s theory of visual attention (1990) and Bailey’s probability of detection (Peterson and Dugas, 1972). Factors such as alcohol use and fatigue may influence a driver’s ability to detect hazards based upon how they influence each of the variables in CASE CR. If we look at the CASE CR acronym, we can see that even the anticipated response may influence the driver’s ability to detect an object. According to Bekkering and Neggers (2002), in general, we are better at finding objects to which we are prepared to respond. This selection for action mechanism has also been supported by the research of others (Airport, 1987; Deubel and Schneider, 1996).

FIGURE 14.2 Response times of drivers responding to different stimuli during daylight and darkness.


408

If a driver is attentive toward the direction of a hazard, this does not guarantee that the object will be detected, as is the case with inattentional blindness research (Mack, 2003; Most et al., 2001; Simon and Chabris, 1999). Research regarding inattentional blindness suggests that drivers who are mentally engaged in a primary task may not identify a noncontrasting second object. The most common reasons for drivers’ failure to detect are due to lighting and contrast difficulties when responding to nonilluminated objects at night. Simply stated, if the hazard is not easily identifiable as an immediate hazard, then a driver is not likely to perceive and respond to it until it is. Delayed detections are more common at nighttime, but nighttime is not associated with much slower response times in all cases. Driver response times to illuminated objects such as traffic signals and illuminated vehicles within lighted intersections have not been much greater at night than during daylight (See Figure 14.2). Muttart (2003a, b) compiled the response times of subjects and real-life drivers by what they were responding to and day vs. darkness (dusk, dawn, and nighttime responses). In Figure 14.2, there is a significant difference in response times and range of responses to pedestrians and bicycles at night compared to daylight. Therefore, objects that are not illuminated (or objects that do not offer an easily identifiable pattern) usually offer the greatest difficulty for a driver at night. For investigation of a collision involving an easily identifiable hazard, proceed to Section 14.5; otherwise, read on. 14.3 Driver Response to Objects Not Easily Identified A forensic investigator may account for a delayed detection interval (a term used by Olson, 1996) depicted in Figure 14.1 in two ways. The first method is to determine a threshold discernibility distance to account for the detection interval (depicted in Figure 14.1) and to apply a perception—response time from that location. Threshold discernibility, also known as the detection interval, detection threshold, or the absolute threshold, is the distance at which 50% of subjects will be able to identify the object. The second method is “distance to impact” research. Distance to impact accounts for the detection interval and perception but does not account for the driver’s motor response and the vehicle latency. An assumed starting point (where perception occurs) is necessary when using perception-response times in scenarios involving a response to an object that is not easily identifiable. The perception starting point is most commonly based upon visibility distance models or illumination criteria, or by measuring the headlight beam illumination pattern of an exemplar vehicle. Each method has inherent flaws and is based upon assumptions that certain visibility or illumination levels are associated with a driver’s ability to perceive a hazard in a real-world setting. Most “visibility” models involve numerous specific measurements. Visibility models frequently require several lighting, contrast, and reflection measurements. Objective and accurate measurements should be utilized whenever possible. The investigator must also consider the tasks of driving relative to the manner in which the model is based.


409

An example of an illumination model is the twilight distance method (Liebowitz and Owens, 1991; Owens et al., 1989), which is based upon the premise that the end of civil twilight is when there is a need

TABLE 14.1 Twilight Illumination Distance Applied to SAE Ground Vehicle Lighting Standard J1383 Candela Twilight distance 25″ Headlight height 10 to 90° up and 45° left to right= 125 20.4 ft 20.4 ft 1° up and 1.5° left= 700 48.3 ft 48.3 ft 1.5° up and 1° right= 1400 57.7 ft 57.7 ft 0.5° up and 1–1.5° left= 1000 68.3 ft 68.3 ft 0.5° up and 1–3° right= 2700 94.9 ft 94.9 ft 0.5° down and 1–1.5° left= 3000 100.0 ft 100.0 ft 4° down and 4° right= 8000 163.3 ft 29.8 ft Max: 0.5° down, 1.5° right= 20000 258.2 ft 239.5 ft Typical max output (2° down and 2° 30000 316.2 ft 59.7 ft right) Notes: Column three shows the minimum of the beam aim distance or the twilight distance if the headlight height is 25 in. (10 cm).

for greater lighting during outdoor activities and that the ambient illumination at civil twilight is 0.3 fc (3 lx). Thus, the conclusion was that a reasonable threshold for visibility is when objects ahead are illuminated to 0.3 fc (3 lx) or more by headlights. When testing various vehicles, this threshold distance varies due to the nonsymmetrical beam of vehicle headlights. It is recommended that an exemplar vehicle (with a similar load) be used to determine the headlight beam characteristics and to consider where within the asymmetrical beam pattern the hazard was at different times during the vehicle approach to the crash. The equation utilized in the twilight method is: TD=(cd/0.3)1/2 (14.1) where TD is the distance at which headlight illumination is equal to or greater than ambient lighting at twilight (0.3 fc) and cd is intensity in candela. As an example, Table 14.1 shows the twilight distances applied to Ground Vehicle Lighting Standard SAE J1383. With this type of model, contrast, glare, and adaptation should also be considered and accounted for if necessary. Visibility models are typically based upon the research by Blackwell (1971, 1980a, b). Blackwell had stationary observers make observations in l/30th of a second. When similar types of studies were conducted, it was noted that subjects could be responding to changes as well as visibility (VISCON, 1998). If subjects were stationary and viewing the same scene, it is not clear how Blackwell determined that they were responding to mentally seeing a target or detecting a change in the scene. Researchers from the Ford


410

Motor Company (Bhise et al., 1976) attempted to validate the Blackwell data by conducting an experiment of their own. However, Farber (1988, p. 14) indicted that “we do not as yet have a complete working model [for object detection distance determination].” Farber and Matle (1988) indicated that additional work was needed to substantiate more thoroughly the field factors (environmental factors) and assumptions used in DETECT and PCDETECT computer detection distance programs. Neither the DETECT nor PCDETECT program has been released to the public. Therefore, this model is not recommended for use in a forensic setting until it is released to the public and analyzed well beyond 12 subjects. Visibility models generally account for many more variables than does an illumination model. Visibility models may consider glare, reflectivity, and contrasts luminance, and they may be used to meet the requirement for scientific support, but they usually involve a great deal of work to develop a credible estimate that has no assurance of being more or less accurate than the much simpler twilight distance model due to individual variance in low-light situations (at this time). Furthermore, visible does not equate to discernible in a real-life environment by an unsuspecting nonstationary driver. Therefore, any visibility model or method must account for the limitations of a person who is engaged in the task of driving.

TABLE 14.2 Relationship of Driver Identification Distance to Clothing Brightness Material

Feet SD Meters SD

Ref.

Red/Red-Or/Or. retroreflective Red retroreflective Yell-Grn/Yell/Flr.Yell/Grn retroreflective Lime retroreflective White retroreflective White vest

Right Left Right Left Left Right

258 267 226 184 143 180

110 138 289

79 81 69 56 44 49

White vest

Left

120 114

37

Yellow Shirt Beige pants-green plaid—more unexpected Beige pants-green plaid—more unexpected Gray pants-blue jean shirt Gray pants-blue jean shirt Denim

Left Right

110 84 139 101

33 42

Turner et al., 1997 49 Muttart, 2000 Turner et al., 1997 34 Muttart, 2000 42 Muttart, 2000 88 Olson and Sivak, 1983 44 Olson and Sivak, 1983 26 Muttart, 2000 31 Muttart et al., 2003 24 Muttart et al., 2003

162

Left

84

78

26

Right Left Right

160 117 80

95 18 85

49 36 24

29 Muttart et al., 2003 5 Muttart et al., 2003 26 Olson and Sivak, 1983 Denim Left 60 33 18 10 Olson and Sivak, 1983 Notes: Speed of the vehicle ranged from 42 to 50 km/h, low beams, and the drivers were told to look for pedestrians. Right refers to the curbside and left is the opposite side.


411

As mentioned earlier, visibility and illumination methods are used to determine a “point of perception” from which a perception-response time can be used. The alternative method to this would be to utilize “distance-to-impact” research results. Distance-toimpact results can be obtained from studies involving subjects in moving vehicles identifying when they could identify an object. The distance-to-impact method accounts for the time necessary to perceive an object; it is not necessary to determine a starting point or a “point of perception.” However, distance-to-impact results are from subjects who responded to the presence of a pedestrian or object and knew what they were looking for. Distance-to-impact research involves a driver or observer responding verbally or by pressing a button in response to pedestrians or objects in the road. This methodology includes the time necessary for identification and a simplified response (verbal response or time to press a button). Therefore, only the time necessary for formulating a potentially more complex response, the motor response of the driver (limb-movement time), and the inherent delay of the vehicle are not accounted for in distance-to-impact research. Motor response and vehicle latency usually account for approximately 0.75 sec (Muttart, 2004); subjects in distance-to-impact research are specifically looking for objects (but do not know when or where they will appear). The formulation of a response in distance-to-impact research is likely to be more complicated than “press the button when you see a pedestrian.” Therefore, an adjustment time of 0.75 to 1.5 sec is added at the point of identification to account for mental processing, limb movement, and vehicle latency. The identification distance is the distance to impact of an object of similar contrast and lighting identified in research. Several researchers have collected data regarding visibility distances. Research collected on a stark closed course or research from stationary observers will likely overestimate the identification distance. Laboratory studies using videotape have reported much shorter identification distances than those conducted with moving observers in more realistic environments (Olson and Sivak, 1983, 1984; Muttart, 2000; Muttart et al., 2003; Turner et al., 1997; refer to Table 14.2). The distances listed in Table 14.2 were those of observers in vehicles that were traveling 42 to 50 km/h and who were told to identify when they could see pedestrians (they were alerted to the purpose of the test); however, they did not know when or where the pedestrian(s) would appear. The investigator must effectively account for the visual noise, dynamic viewing, cognitive factors, and sensory filter that are inherent when dealing with unexpected objects in a real-life response situation. Thus, a seeing-distance study will not be an effective tool to predict the distance or likelihood of detection in most cases if the cognitive factors involved in real-world driving are not accounted for. Once the detection interval is accounted for or if there is no measurable detection interval, an estimated response time may be utilized. This leads to the question of what an appropriate perception response time is for a given situation. 14.4 Influence of Research Methods on Previous Results Previous authors have incorrectly concluded that driver response times varied and therefore could not be predicted (Sohn and Stepleman, 1998; Dilich et al., 2002).


412

However, these studies mixed several different stimulus-response situations—many involving different methodologies—and then claimed that because the results were so different, response times could not be predicted. Muttart (2001, 2003a, b, 2004) found that the different results could be accounted for if the methodology of the study and the stimulus (object or hazard) were accounted for. The response time results reported will not only change due to the stimulus (hazard) to which the subjects responded, but the methodology of the experiment could also explain and predict the different results obtained among studies. The methodology variables examined in the research by Muttart (2003a, b, 2004) include the driving task, anticipation, experiment type, response complexity, and the amount of the transition time (limb movement and vehicle latency) accounted for. In some studies, the subjects did not have the burdens of the driving task. Anticipation varied in that a few researchers told the subjects what to look for and what to do in response and others did not warn the subjects at all. In some studies, subjects were told to brake in response and in others the driver was allowed to select his or her response. The differences among laboratory, simulator, closed-course, and road testing were a significant factor, as was the method used to measure the response. Some studies stopped the clock when the foot came off the accelerator (perception-reaction time), some measured brake reaction time, and others measured the full response. Therefore, the methodology of the study was a significant factor in the results. Alexander and Lunenfeld (1975) indicated that the driving task was a significant factor to consider when evaluating the response of drivers. These researchers claimed that the task of driving involves navigation, guidance, and control tasks, all of which contribute to the workload on a driver. Driver response results (Cohn, 1987; Yoo et al., 1999; Fambro et al., 1998) have subsequently corroborated Alexander and Lunenfeld. Overall, drivers needed more than twice as much time to respond to a light stimulus when engaged in the task of driving than when seated in a stationary vehicle. Therefore, to address properly how drivers may respond, the workload involved in the task of driving must be accounted for. Other methodology variables that Muttart found to be statistically significant were experiment type, transition time, and anticipation. The manner in which Muttart defined his variables can be found in Appendix A at the end of this chapter. In general, as the methodology of the experiment became closer to that of real life, the response times increased. Therefore, if the subject knew the stimulus (“When you see the light go on…”) and knew the appropriate response (“…depress the brake pedal”), the response times were faster. Response times increased from laboratory to simulator to closed course and then to road studies. Lastly, if a study measured a perception—“reaction,” or a brake reaction time, it would underestimate a perception-response time. Muttart also examined substantive variables such as age, movement, response options, response choice, number of stimuli, contrast, eccentricity, speed, road type, topography, and others (a complete list can be seen in Appendix A) to determine which variables prove to be significant influences upon a driver’s response (using ANOVA) or have linear trends (using linear regression). ANOVA allows for the evaluation of interactions and is also a relatively robust experimental design. Linear trends were used to evaluate the amplitude of the influence


413

of each variable. Therefore, a primary goal of the research was to determine which variables influence driver response time and to what extent. A few of the variables that Muttart (2001, 2003a) found were not significant predictors of change in response times were noteworthy. These variables include alcohol consumption, age, fatigue, and gender. The primary reason that alcohol did not reach significance was due to lack of data. To understand the influence of alcohol properly, it is necessary to consider the acronym DATT, which stands for dosage, absorption, tolerance, and task. As we have found, when the stimulus-response scenario changes, so does the response time and also the influence of each variable (see the adjustments to the response time in Table 14.3). Therefore, alcohol dosage will likely have a different influence in car following, nighttime response, path intrusions, and responding to vehicles changing lanes. One adjustment for DATT

TABLE 14.3 Amplitude of Adjustments for Response Times in Various Collision Configurations and Variablesa Vehicle following Contrast Natural lighting Driving Eccentricity Experiment location Headway Number of stimuli Response complexity Topography Transition

583 ms/unit 97.5 ms −740 ms 26 ms/degree 56.5 ms/unit

On lighted roads Day vs. night Add for driving, subtract for not driving Measured from directly ahead 1—lab; 2—simulator; 3—closed course; 4—road

279 ms/sec 509 ms −210 ms

On straight and level road One vs. multiple Button—brake only—brake when other options were available—brake and steer Straight vs. curves (horz/vert)/intersections/cues 1—throttle release; 2—brake appl.; 3—full braking Path intrusions

Anticipation Contrast Natural lighting Driving Eccentricity

24 ms/unit 743 ms/unit 224 ms/unit 350 ms 17.5 ms/degree

Experiment location Road Number of stimuli Topography Transition

79.5 ms/unit

Natural lighting Driving

303 ms/unit −91 ms/unit

−703 ms 375 ms

63 ms/unit 716 ms/stimulus −496 ms/unit 375 ms/unit

On lighted roads Day vs. night Driving vs. not driving One lane 3°, instruments 16°, rear-view 27°, side-view 35 to 45°, radio 50° 1—lab; 2—simulator; 3—closed course; 4—road 1—rural; 2—urban; 3—highway One vs. multiple Straight vs. curves (horz/vert)/intersections/cues 1—throttle release; 2—brake application; 3—full braking Vehicles changing lanes Day vs. night Driving vs. not driving

Handbook of Human factors in litigation Eccentricity

414

13 ms/sec

One lane 3°, instruments 16°, rear-view 27°, side-view 35 to 45°, radio 50° Lanes crossed 340 ms/lane 1 or 2 (multiple) Number of stimuli 254 ms/stimulus Up to 2 Transition 375 ms/unit 1—throttle release; 2—brake application; 3—full braking a In milliseconds. Notes: The amplitude of the adjustment for a single variable may change in different collision configurations. These adjustments are designed to supplement a research study analogous to a reallife crash situation so that the investigator can directly compare the research with the real-life situation.

combination is not justifiable. Instead, the response of a driver suspected of being intoxicated should be compared with the manner in which drivers have responded in research and real-life situations in order to estimate the influence of alcohol. The same type of analysis may be done relative to fatigue. No research is available that can enumerate the response delay of a fatigued driver and it is unclear what the definition of a fatigued driver is. Again, it is best to compare the response of the driver against the manner in which drivers have responded to analogous circumstances in research to determine the influence of fatigue, if any. When all the data (more than 20,000 responses) involving men and women were compared, men responded slightly less than 1/10th of a second faster, but that difference was not significant. One study measured the response time of drivers responding to a known audible alarm when backing a vehicle at slow speeds in a parking lot (Lerner et al., 1997). This study reported very fast response times and involved only men. There was no similar study involving all women. When that study was removed from the database, response times of men and women showed virtually no difference. Many believe that younger drivers will respond faster than older drivers. Mourant and Rockwell (1972) showed that younger drivers have underdeveloped search patterns; the meta-analysis of the data and the research by Muttart (2000) have shown that they also have poor discrimination skills and underdeveloped conditioned responses (Muttart, 2004) when compared to more experienced drivers. Although younger drivers are capable of faster limb movements (Olson and Sivak, 1986), the underdeveloped search patterns, discrimination skills, and conditioned responses could cause response delays much greater than the 1/10th of a second that they may gain in the motor response portion of the response. Overall, response times relative to driver age followed a very shallow Ushaped curve, with younger and older drivers responding marginally more slowly, although this did not reach statistical significance. Again, this research primarily involved responses on lighted roads and in daylight. Even though experienced drivers may have better search skills and discrimination patterns, the results may be different when evaluating response times to dark objects on unlighted roads. Muttart (2003a) completed a meta-analysis of 145 studies that reported driver response results. This involved reading each study and entering reported response times, short comments, the type of response, the stimulus (car, light, pedestrian, etc.), and appropriate codes (as listed in Appendix A) into a database. Although Muttart found that several variables were statistically significant predictors of driver response time, no single variable accounted for more than 15% of the variance and most accounted for less than 5% of the variance of the response. Therefore, the question that Muttart attempted to


415

answer was: “How do the variables fit together and is it possible to develop a mathematical equation to predict driver response times based upon the manner in which drivers have responded in research?” 14.5 Perception-Response Time Muttart (2001, 2003a, b, 2004) developed a model for estimating driver response times. A series of equations that mathematically estimate driver response times based upon a meta-analysis of previous research and real-life responses was developed. Also, a method for estimating driver response times using analogous research studies was developed (Muttart, 2003b). Multiple linear stepwise regression (MLSR) was used to develop the equations (Muttart, 2003a, b). MLSR has been referred to as the surest path to the best prediction equation (Tabachnick and Fidell, 1996). Furthermore, Muttart (2003b, 2004) offered the statistical distribution of real-life driver responses. This information is extremely valuable because it offers insight into what could be considered a reasonable (85th percentile) response time range. The 85th percentile response is the maximum response time at which 85% of the population will respond and provides a benchmark to which to compare driver responses. After developing the equations using MLSR using data from the meta-analysis of research studies that measured driver response, Muttart compared the estimates using the equations to real-life driver responses. The real-life drivers were video-recorded during responses to emergency traffic incidents in Kentucky (2000 to 2003) and Helsinki (1991 through 1999). There were 147 real-life traffic responses and a total of many more than 10,000 measured responses in research studies. Of the emergency responses, 75 involved crashes, while the remaining incidents involved emergency responses. If an emergency response was not necessary, then the data were not included in the analysis. Emergency response was defined as skidding, steering, or horn use in an attempt to avoid a collision that would most likely occur without an avoidance response. Researchers video recorded and analyzed 62 responses to vehicles changing lanes, 12 responses to a lead vehicle, and 58 responses to a path intrusion. In cross-street (broadside configuration) events, the perception-response time started when the intruding vehicle moved beyond the stop line. In the Helsinki path intrusions, the stop line was a significant distance from the intersection. Therefore, the starting point in these videos was when the vehicle crossed beyond the crosshatched crosswalk. When the response to vehicles changing lanes was measured, the time was started at the first frame that showed lateral movement toward the responder’s lane. In instances in which the response of the turning driver was measured, the time started when the vehicle first crossed the stop line (not all were starting from a stop). In the carfollowing situation, the time was started when the lead vehicle began to decelerate (when the brake lights of the lead vehicle went on). When a subject was responding to pedestrians, the time was started when the pedestrian first started to move deliberately toward the road (up to 1.5 steps from the road edge) or when the pedestrian first emerged from behind a parked vehicle.


416

TABLE 14.4 Simplified Estimates of Full Braking Response Times in Path-Intrusion Situations Looking….

Ahead Day Night

Instruments Day Night

Rear view Day Night

Single hazard Intersection 1.3 1.6 1.8 2.0 2.0 2.2 Single hazard Midblock 1.8 2.1 2.3 2.5 2.5 2.7 Multiple stimuli Intersection 2.1 2.3 2.5 2.7 2.7 2.9 Multiple stimuli Midblock 2.6 2.8 3.0 3.2 3.2 3.4 Notes: For objects that are easily recognizable as an immediate hazard, ±40% has accounted for the middle 70% of drivers. The response time varies in different situations. This chart should not be used without consideration of other variables, corroboration by other means, and consideration of individual variability.

The first video frame to show lateral movement, noticeable slowing (vehicle pitch) with sound (braking), or brake light activation identified the end of the response. In two instances, there was lateral movement with horn activation or noticeable deceleration with horn activation. In all other instances, sound refers to brake application noises (the short grinding sound followed by tire squealing). In all such events, the response could be defined within one frame (.03 sec). With the Helsinki videos, a change in forward movement per frame that was closely followed by vehicle pitch was included. However, in most cases, braking was identified by the brake lights. Only if tire squealing was concurrent with lateral movement, vehicle pitch, or first noticeable deceleration was that event included in this analysis. To estimate a driver’s response accurately, Muttart found that analogous situations should be used and defined an analogous situation as a response to an object that emerged from the same general direction—for instance, in-line, path-intrusion, or sideswipe situations. A study that measures the response time to a red light or the response time to a lead vehicle will not be an effective tool for estimating the response to a path intrusion. Furthermore, even with a similar scenario (path-intrusion study to estimate a pathintrusion response), the investigator should make sure the study is measuring the same time period to which he is comparing it. For instance, if the study measured a perception—“reaction” time (up until the foot first moves from the throttle), then it should not be used as a perception—“response” time (up until full braking or start of skid marks or first lateral movement) without an appropriate adjustment. Most importantly, perception-response time is exactly that—it is not vision-response time. Therefore, the time period starts from when the hazard is easily identifiable as an immediate hazard until the vehicle first starts to respond due to driver response choice. Objects that cannot be easily identifiable as an immediate hazard should be analyzed further before perception-response times are utilized, if utilized at all. The stepwise equations have estimated the actual response time within 0.5 sec nearly 75% of the time to date. Based upon the manner in which drivers have responded in 145 research studies that measured driver response times and also upon how the 147 real-life drivers have responded, tables were made that show the manner in which driver response time fluctuates in different situations. The tables (Table 14.4 and Table 14.5) show the estimate and the range of the middle 70th percentile, which would mean that 85% of the


417

driving population would be expected to respond faster than the estimate plus the estimate multiplied by the range percent. Table 14.4 shows the ranges in response times in a path-intrusion situation and Table 14.5 shows how the range of driver response times fluctuates in a car-following situation. The range of responses tightens considerably if the adjustment-to-baseline method of estimating response time is also considered. “Adjustment to baseline” (A2B) is a corroborative method of estimating driver response. Baseline refers to an analogous study. Therefore, the A2B method involves making adjustments to an analogous study. For a path-intrusion crash, we want to obtain studies that measure driver response to a path intrusion. The study may report perceptionreaction rather than perception-response time, or may be a daytime situation, although our case occurred at night. In such a case, we would apply an adjustment to the results of the analogous study so that the result would be applicable to the situation being investigated.

TABLE 14.5 Simplified Estimates of Full Braking Response Times in Vehicle-Following Situations Looking ahead

Looking at instrument Looking at rearview panel mirror Straight Curve/cue/ Straight Curve/cue/ Straight Curve/cue/ intersection intersection intersection 1 second 1.5 0.8 1.9 1.2 2.2 1.5 2 1.9 1.2 2.3 1.6 2.6 1.9 seconds 3 2.3 1.6 2.7 2 3 2.3 seconds 4 2.7 2 3.1 2.4 3.4 2.7 seconds 5 3.1 2.4 3.5 2.8 3.8 3.1 seconds Multiple 1 second 2 1.3 2.4 1.7 2.7 2 stimuli 2 2.4 1.7 2.8 2.1 3.1 2.4 seconds 3 2.8 2.1 3.2 2.5 3.5 2.8 seconds 4 3.2 2.5 3.6 2.9 3.9 3.2 seconds 5 3.6 2.9 4 3.3 4.3 3.6 seconds Notes: The response time varies in different situations. This chart should not be used without consideration of other variables, corroboration by other means, and consideration of individual variability. Single hazard

Muttart (2003b) analyzed analogous path-intrusion studies involving simulator and closed-course studies. Also, the analogous studies included responses to pedestrians and vehicles. Lastly, the goal was to select studies that reported data for a larger number of variables. The following six path-intrusion studies were selected: (1) Barrett et al. (1968);


418

(2) Broen and Chiang (1996); (3) an unpublished study conducted by the Southwestern Association of Technical Accident Investigators (SATAI; 1999); (4, 5) Mazzae et al., 1999a, b; and (6) Hankey et al. (1996). Barrett et al. (1968) was a simulator study that involved driving on an oval course. On the 11th lap, a pedestrian emerged from a shed and toward the path of the car. Broen and Chiang (1996) involved a simulator study in which the subjects were told to brake as fast as possible for two pedestrians who emerged into the path of the vehicle. The SATAI study involved subjects driving in a mostly vacant lot of a sports stadium and responding to a “skateboarder” (a cutout mounted to a skateboard) after emerging from around the corner of a stadium. The two studies by Mazzae et al. (1999a, b) were conducted using a simulator and a closed course and involved a vehicle emerging from the passenger side of the road and stopping with the front near the center of the subject driver’s lane. Hankey et al. (1996) conducted a simulator study that involved vehicles emerging into the path of the subject’s vehicle slightly more than 3 sec before impact. Only two studies have recorded the times necessary for drivers to respond to vehicles changing lanes (Currie, 1969; Muttart, 2004). Currie reported an average response time of 588 ms when his subjects “drove” an electric car on a track and responded to the other vehicle changing from its original lane and into the subject’s lane. The driver response times to a vehicle changing lanes (originally in a sideswipe type configuration) can be seen in Table 14.6. In car-following research, an attempt was made to select research that addressed short and long headways as well as simulator and closed-course or road results. The analogous vehicle-following response studies were conducted by Summala et al. (1998), van Winsum (1998), Kane et al. (1999), and McGehee et al. (1998). Summala et al. produced a comprehensive study relative to driver response times. They examined the influences of headway, eccentricity, and taillights on or taillights off on the response times of drivers in a car-following situation. These researchers examined the response times when looking straight ahead, at the dashboard, at the rear view mirror, and at the center console. Kane et al. measured response times in a simulator study at short headways, while comparing athletes to nonathletes. van Winsum measured driver response times on curves and McGehee et al. (1998) measured the response times of drivers to stopped cars and involved a longer headway.

TABLE 14.6 Comparison of Average Estimated Response Time Using Stepwise Equations with Average Real-Life Response Times in Helsinki and Kentucky N Path intrusions Turning Path intrusion while turning left Path intrusion while turning right After starting from

Actual average

Standard deviation

Predicted Within N Hazard N Hazard average 40% to right to left

58 8

2225

680

2443

86% 2

1747 6

2384

3

1905

563

2760

50% 2

2133 1

1448

6

3000

1125

2425

67% 1

1935 5

3213

Estimating driver response times stop: violation of red traffic control signal No turning or red light violation Natural lighting Path intrusion day Path intrusion lighted road at night Eccentricity Path intrusion visible ahead Path intrusion visible out windshield Path intrusion visible out side or rear windows Distracters Path intrusion was the only stimulus Path intrusion while turning or with distracter stimulus Following

419

41

1638

695

1752

67% 16

32 26

1742 2035

913 730

1666 2383

58% 77%

25

1543

538

1524

72%

19

1877

942

2128

67%

14

2460

872

2696

73%

36

1706

874

1739

22

2147

723

2341

534

Within 52% 1122 67% Within 30% 928 50%

12

Following with short headway Vehicles changing lanes Vehicle changing one lane day Vehicle changing one lane night Vehicle changing two lanes day Vehicle changing two lanes night Green traffic signals

12

Response to green traffic signal TOTAL

16

1124

61 870

308

10

1328

455

1305

70%

18

1213

375

1257

83%

17

1430

415

1590

82%

753

Within 53% 1400 69%

147

1119

1566

70%

16

16

1751 25

Within 39% 70% of estimate Notes: Response times varied based upon collision configurations; eccentricity, lighting, distracters, and turning also influenced response times. Furthermore, the stepwise equations were versatile enough to account for those differences. Source: Muttart, J.W. (2004). DRIVE3: a simplified method for estimating driver response, Auckland, NZ: Australasian S. Pacific Assoc. Crash Investigators 2004 Conf. Proc.


420

To obtain the amplitude of the adjustments, a series of 32 methodology and substantive variables were mathematically evaluated to determine if statistically significant linear or quadratic relationships existed between that variable and driver response times based upon the meta-analysis of 145 driver response studies (a complete listing of these references can be seen in the original research). In preliminary analysis, it was learned that the contribution of a variable to a driver’s response time is usually more accurate when considered in context with several other variables. Therefore, the adjustments were determined based upon how the individual variable was linearly associated with driver response and reduced by 50% to account for interaction effects. The 50% figure was obtained by comparing the linear relationship of the variables included in the stepwise regression equations with the slope for that same variable when considered individually.

TABLE 14.7 Rule-of-Thumb Perception-Response Time Estimates 1.5 D/2.5 N

2.5

1.6

0.75–1.5

Within range 42 27 47 43 Outside range 47 62 42 46 Percent accuracy 47.2% 30.3% 52.8% 48.3% Notes: These estimates performed poorly, were inconsistent, and offered no versatility as a method for estimating response times. No rule-of-thumb method estimated all response times within 33% of the estimate more than 53% of the time. Source: Muttgart, J.W. (2003). Evaluation of methods for estimating driver response times, Proceedings of the Sixth International Conference of Traffic Accident Investigators, Stratfordupon-Avon, England, pp. 1–29.

Therefore, the variables that were statistically significant predictors of response time if considered alone but that did not reach significance enough to be included in the stepwise regression equations can be accounted for in the A2B method. The amplitude of the adjustments from each variable that will most accurately estimate driver response time (based upon the current available research) is listed in Table 14.3. Note that the adjustment for the same variable may change in different accident scenarios (pathintrusion vs. car-following situations as an example). Therefore, just as a single response time will not likely predict response times for all situations, a single adjustment applied to all collision scenarios will not likely be effective. Please refer to the original research paper for the method for making each adjustment (Muttart, 2003b). Muttart (2003b) compared the stepwise equations, rule-of-thumb estimates, and adjustment-to-base-line estimates to the real-life responses. The rule-of-thumb estimates that were compared included the often-used “1.6 sec” figure, the 2.5 sec figure addressed in the Manual on Uniform Traffic Control Devices, a range of 0.75 to 1.5 sec, and 1.5 for daylight and 2.5 for nighttime. Rule-of-Thumb Estimates The rule-of-thumb estimates did not perform well at all (see Table 14.7). The “1.5 sec day-2.5 sec night rule” was accurate within 33% only 47% of the time. The “2.5 sec rule”


421

proposed by the MUTCD (U.S. Department of Transportation, 2001) was accurate within 33% only 30% of the time. The “straightforward” range (Olson, 1996) of 0.75 to 1.5 sec, if used improperly (for all responses), would estimate the response time within 33% only 48% of the time, and the relatively popular 1.6 sec estimate is within 33% of the actual response time only 52% of the time (Muttart, 2003b). The stepwise regression equations and the adjustment to baseline estimate were averaged to produce an overall estimate which was within 33% of the actual response time more than 67% of the time. The average margin of error for all 147 real-life responses was 436 ms. Vehicles Changing Lanes Humans are relatively good at detecting lateral movement or movement across their visual field. Therefore it is not surprising the responses to sideswipe configurations and to vehicles changing lanes were relatively fast. A vehicle changing lanes is defined as a vehicle originally traveling in the same direction of the responder but that then changes lanes across or into the responder’s path. The range of responses to a vehicle changing lanes was relatively small. Drivers who responded to a vehicle changing from the next lane during daylight responded on average in approximately 0.9 sec, while a driver responding to a vehicle that started turning from two or more lanes from the responder at night needed closer to 1.5 sec to initiate a response maneuver. Path Intrusions For the 58 real-life path-intrusion situations evaluated, the stepwise equation method estimated the response time of the real-life driver within 40% for 67% of the time (Muttart, 2004). The precision improved to 35% for the middle two thirds of drivers if the average A2B estimate from the six analogous studies was averaged with the stepwise equation estimate. It is apparent that the more sources that were considered in the analysis, the more accurate the estimate became (up to this level of analysis). Therefore, Table 14.4 and Table 14.5 should be utilized as rough guidelines only and are not appropriate for use in forensic settings without further analysis, adjustment, and substantiation that the estimate offered is appropriate for that particular situation. Response time varied due to eccentricity, natural lighting, and distracters, after red traffic signal violations, and while turning (when compared to driving straight). The average response time for a driver who started into the intersection while facing a green signal and had to respond to a red light violator was 3 sec. Five of the drivers averaged 2.7 sec and one responded in 4.7 sec. Drivers making a right turn responded faster to hazards emerging from their left than right, while drivers making left turns responded faster to hazards emerging from the right than left. More data should be collected and individual case anomalies considered before making conclusive findings relative to response time while turning right or left. However, it is apparent that drivers typically need more time to respond when attempting a turn. This finding corroborates the research by Hancock and colleagues (1990), who found that turning maneuvers were associated with greater mental workload and head movements, which would increase the potential for detection failures and, as found in this research, delayed detection, when compared to drivers traveling straight.


422

Driver response times at night were only about 0.3 sec longer during dawn, dusk, and night than during daylight, which is consistent with prior research shown in Figure 14.2. The limited difference between lighted and low-light response times may also be due to the fact that the real-life responses occurred on lighted roadways. Based upon these results and the results of others (Figure 14.2), response times may increase significantly in crash situations involving pedestrians and nonilluminated objects on unlighted roads in most cases. Green Traffic Signals and Gaps in Traffic Several studies have measured the time necessary for a driver to respond to a green traffic signal or a gap in traffic (Graham et al., 1995; Fugger et al., 2001; Lerner et al, 1996; Naylor and Graham, 1997; Wortman and Matthias, 1983). When the result from each study is averaged, it suggests that drivers will respond to a green traffic signal or a gap in traffic in approximately 1.4 sec (SD=0.5 sec). Muttart (2004) measured the response time of 16 subjects responding to a green traffic signal; the average response time was 1.1 sec (SD=0.5 sec). Therefore, most drivers can be expected to respond to a green traffic signal or a gap in traffic in 0.75 to 1.75 sec, which is very similar to the “straightforward” estimate offered by Olson (1996). Car Following When evaluating the adjusted studies by Summala and colleagues, van Winsum, and Kane and colleages (who all investigated response times at shorter headways), the average margin of error was 187 ms (Muttart, 2003b). The adjusted estimates of response times developed using McGehee et al. (1998) were typically underestimated. However, the McGehee study is still recommended as a baseline because it involved a long headway event and, when averaged in with the other adjusted studies, provided a relatively accurate estimated response time most of the time. When the stepwise equations and the A2B estimates were averaged, the precision of that overall estimate improved to within 40% for two thirds of the real-life drivers. The stepwise equation estimated the actual average response time nearly exactly. However, all of the real-life following situations to date involve headways of less than 2.5 sec. It is expected that range of responses will become larger as the complexity of the event increases. Because longer headway situations are typically more complex for a driver, it is recommended that a larger range of potential responses be used (50 to 55% until more data are collected) when evaluating the response of a driver to a slow-moving or stopped vehicle (with no warning) on a straight road. The stepwise regression equations yielded the most accurate predictors. The A2B method considers many more variables and is based directly upon an individual research study. Therefore, to reduce the effects of confounding variables in any one study, the average of several analogous studies offers a more accurate estimate of response time. Furthermore, if the A2B estimate and the stepwise equation estimate are averaged and an overall estimate developed, the accuracy of the estimate increases even further.


423

14.6 Urgency Several theorists have claimed that driver PRT is based in large part upon the urgency required. They have based their opinions upon the fact that drivers respond faster to an object that they first discover within a shorter period before “impact.” Significant confounding variables exist in this theory, however. First, if an object is presented to a driver 3 sec before impact, then the greatest PRT that can be recorded is 3 sec. Therefore, when those time-to-contact data are compared to a response to a vehicle stopped on a highway several meters ahead, the end result is that researchers may inappropriately determine that the cause was due to urgency (when it was, in fact, due to much more complicated issues). Furthermore, the research by Muttart (2001, 2003a, b, 2004) has shown that as eccentricity increased, so did response time. Therefore, if an object is thrown in front of a car 1.5 sec before impact and from the same position when the vehicle is 3.5 sec before impact, the driver will likely respond marginally faster when 3.5 sec away because the barrel has a smaller eccentricity relative to straight ahead. In a vehicle-following research study (Sivak et al., 1981), the brake lights on the lead vehicle illuminated but the vehicle did not decelerate. Therefore, the true time to contact was infinite (they are both still traveling at the same speed). However, the drivers who followed reacted very much as they would when a lead car decelerates from the same headway. Green (2000) pointed out that a vehicle traveling head-on within a driver’s lane may have a very short time to contact, but a much longer perception-response time, due to the confusion created by that situation. Furthermore, if the speed of a vehicle approaching a stopped lead vehicle is doubled, the time to contact will decrease by 200%. However, no evidence indicates that response times will change significantly if all else is held constant. In fact, research has shown that those who drive faster may perform more efficiently (Alm and Nilsson, 1991; Triggs and Harris, 1982; Sidaway et al., 1996; Muttart, 2001) as long as they are not overloaded by a second object to which to respond (see Figure 14.3). The very nature of the scenarios that we are examining involves a great deal of urgency, which begins when a driver perceives something as a hazard to himself or others, to his property, or to his time. Therefore, we are usually examining the performance of drivers when they are at the far right of the Yerkes-Dodson (performance vs. stress) curve. Performance is usually optimal when stress levels are at a moderate level, but when stress is increased or decreased from a moderate level, performance will likely be degraded. However, urgency may play a role in the response of drivers in less threatening situations such as responses to yellow traffic signals. 14.7 Transition Time (Limb Movement and Vehicle Latency) Despite the fact that many refer to the process as “perception-reaction,” or simply call it “reaction time,” at times, it is the physical reaction component of the process that receives some of the least attention. Much driver response research neglected the response process by having subjects press a button, brake only, or give a verbal response. In the research involving simplified responses, there are no cognitive components to the


424

response process, no decision to be made, and no possibilities to consider. Other studies cut the response process short by stopping the clock when the foot comes off the accelerator and still others stop the clock the moment that the foot touches the brake. Again, these scenarios do not account for mechanical latency or the time necessary for the physical response. When comparing driver responses, we must make sure we know what we are comparing. Research by van Winsum (1998) and

FIGURE 14.3 The influence of number of stimuli and speed upon the driver response time in vehiclefollowing situations. This type of relationship holds true in pathintrusion situations as well. Note that response time does not increase with increased speed when responding to a single object, but increases significantly if the driver is responding to two or more objects. Kane et al. (1999) has shown that the variability within the response phase of the perception-reaction process is affected by physical response time and driver cognition. When the data for 145 studies that reported driver response were compiled, they showed that it typically took passenger vehicle drivers 0.15 to 0.25 sec to get full braking after application of the brake. The braking latency for commercial vehicles is usually longer. It has taken a little more than 0.5 sec to move the foot from the throttle to the brake pedal. Glencros and Anderson (1976) found that it took approximately 0.4 sec for


425

subjects to move their foot from the accelerator to the brake. Burkhardt (1985) (cited in Cohen, 1987) found an approximately 0.6-sec delay between accelerator release and full braking. Research has shown that a cognitive component in the response phase could delay full braking by approximately 2 sec from the time the brake is applied if there is confusion (or indecision) during that portion of the response (Mazzae et al., 1999b; Kane et al., 1999). Bartram and colleagues (1985) indicated that there was no difference in the time to respond when subjects were asked to respond with a foot or a hand. Much of the latency differences found between steering and braking responses can be attributed to the greater movement distance required to transfer the foot from accelerator to brake when compared to no transfer of hands when steering, minus the greater cognitive requirements of steering vs. braking. The end result is that braking takes only slightly longer than a steering response, but not enough to be statistically significant (Muttart, 2003a, b). Several researchers have measured response latency. There have been two types of studies in this regard: those that measured response latency using a light stimulus (Scott et al., 1996; Hoffman, 1991; Parkes and Hooijmeijer, 2000; Burkhardt, 1985; Koppa et al., 1996) and those that measured the response latency to a road hazard, whether on the road or when driving a simulator (Broen and Chiang, 1996; Otto et al., 1980; van Winsum, 1998; McGehee et al., 2000; Barrett et al., 1968). Studies by Burkhardt (1985) and Lieberman et al. (1995) involved drivers responding to the brake lights of a lead vehicle. However, there was no “hazard” and the subjects were told to brake when the light went on. Therefore, even though a lead vehicle was present, these studies are best categorized as studies that measured drivers responding to a known light stimulus. Subjects responding to a known stimulus (usually a light on the dashboard or a command to brake) had movement times and braking latencies much faster than those responding to a hazard. Therefore, in a real-life situation, drivers are likely to need a half-second or more to move their foot from the accelerator to the brake. There is also a delay due to vehicle latency. Vehicle latency delay refers to the time period from when the brakes are first applied to the moment at which the wheels begin to leave skid marks (or to full braking). In many simulator studies, full braking is defined as 200 N brake-pedal force. Visible skid marks begin to appear before the wheels are locked (Goudie et al., 1995; Reed and Keskin, 1989). Research that reports vehicle latencies from application to lock up may slightly overestimate the vehicle latency time, but that difference is usually not more than 100 ms. Simulator studies reported very long vehicle latencies (Mazzae et al., 1999b; van Winsum, 1998; Brown, 2000; Kane et al., 1999). Kane and colleagues (1999) surmised that the reason for this is that the subjects were responding to the perceived lack of vehicle response. Because driving simulators may not offer the driver the same vehicle pitch and perceptible feeling of slowing, subjects may respond to that perception of failing to slow fast enough by reapplying the brakes or attempting a different maneuver. Muttart’s research was from intersections that were well lighted and there were no casual responses. Therefore, the estimations assume that the object is easily discernible as an immidiate hazard (usually that means daylight, well-illuminated conditions or a distinctly shaped, self-illuminated object at night). If a situation involving an object that is not easily identified as a hazard is evaluated, the equations and adjustment-to-baseline method will likely underestimate the response time. Also, the analyst must understand


426

that an object’s being within the driver’s sight line does not equate to the object’s being perceivable as a hazard in all cases. One particular case would involve vehicles on straight planes associated with relatively faster traffic speeds that are stopped without priming to the approaching drivers. The analyst must also understand that the driver is most likely not looking in the direction of the hazard initially. No significance relative to the side of the road of the hazard or the direction of steering was found. Therefore, there is no reason that this research should not apply to driver responses in countries that drive on the left. (An eccentricity difference may be present if the crash involves a right-side-drive vehicle on a right-side-drive road, or a leftside-drive vehicle driving on the left.) Furthermore, the data upon which the equations were based were multicultural and they were compared to responses in Helsinki and the U.S. Therefore, this research suggests that drivers faced with emergency response situations respond similarly in analogous situations. If the situation does not call for an emergency response maneuver, this research may not apply. 14.8 Summary An investigator should understand that estimating the response of an individual must be done with the consideration that individual variability exists. Variability does not necessarily come from driver age as some suggest, but can be better explained by driver expectancy and attention toward a particular hazard. Therefore, it would be inappropriate to claim that a single number could define a driver’s response time and the investigator should report the precision along with the estimate. Some situations are not accounted for in the research. If that is the case, then adjustments to the response time estimates should be considered. (Suggested adjustments can be found in Table 14.3.) If the scenario considered does not involve an easily identified hazard or a nonanalogous event, then additional and/or alternative methods must be employed. In two known situations the estimated response time may be overestimated or underestimated due to the interaction of two variables. These situations involve drivers traveling at high speeds and responding to multiple objects (two hazards that are spatially separated or a hazard and a distracter) or when there are long headways on straight roads (with no perceptual cues that the lead vehicle is stopped or traveling slowly). When a driver travels at high speeds and responds to multiple objects, the response time increases significantly, as can be seen in Figure 14.3. These finding are corroborated by the research of Johnson and colleagues (2004). With regard to rear-end collisions at intersections, the response time is usually relatively quick, while response times to stopped vehicles that are straight ahead on straight and level highways are usually much longer (see Figure 14.4).


427

FIGURE 14.4 The influence of headway and topography on the driver response times in vehicle-following situations. Note that response time remains constant regardless of headway when responding to lead vehicles at intersections and road curvatures. 14.9 Discussion Although driver response estimates have been offered, it is recommended that analogous studies be obtained to corroborate the response time estimate. Also, this research has focused upon easily identifiable objects. If the object is not easily identifiable, additional analysis is necessary. The driver response methods explained in this chapter have been adopted by a computer program called Driver Response in Various Environments Estimated Empirically or DRIVE3 (a software program sold by REG-TEC, LLC.). DRIVE3 offers a method of estimating driver response times using the stepwise equation and A2B methods and also offers the user the precision of their estimates as well as time and distance relationships of each vehicle or object. The stepwise equations, A2B method, driver response estimates, and adjustments are based entirely upon how drivers have responded in research and in real-life incidents. There has been no difference in the response of drivers in the emergency response incidents that did not involve crashes and those that did. Therefore, DRIVE3 and the research discussed in this chapter offer methods that had been validated against real-life drivers and based upon research involving several thousand driver responses (in real life and research).


428

References Alexander, G.J. and Lunenfeld, H. (1975). Positive Guidance in Traffice Control . Washington, D.C.: US Department of Transportation, Federal Highway Administration. Alexander, G.J. and Lunenfeld, H. (1986). Driver expectancy in highway design and traffic operations. Publication No. FHWA-TO-86–1. Federal Highway Administration, Washington, D.C. Allport, A. (1987). Selection for action: Some behavioral and neurophysiological considerations of attention and action. In H.Heuer and A.F.Sanders (Eds.). Perspectives on Perception and Action (pp. 395–419). Hillsdale, NJ: Erlbaum. Alm, H. and Nilsson, L. (1991). Changes in driver behavior as a function of hands free mobile telephones: A simulator study, Linoping, Sweden: National Technical Information Service, PB92–153188. Barrett, G., Kobayashi, M., and Fox, B.H. (1968). Feasibility of studying driver reaction to sudden pedestrian emergencies in an automobile simulator, Hum. Factors , 10, 19–26. Bartram, D., Banerji, N., Rothwell, D., and Smith, P. (1985). Task parameters affecting individual differences in pursuit and compensatory tracking performance, Ergonomics , 28(12), 1633– 1652. Batchelder, S., Rizzo, M., Vanderleest, R., and Vecera, S. (2003). Traffic scene-related change blindness in older drivers, Car Assessment 2nd Proc. 2003, 177–181. Bekkering, H. and Neggers, S.F.W. (2002). Visual search is modulated by action intentions, Psychol. Sci. , 13, 370–374. Bhise, V.D., McMahan, P.B., and Farber, E.I. (1976). Predicting Target Detection Distance with Headlights, Washington, D.C., Prepared for the Annual Meeting of the Transportation Research Board. Blackwell, H.R. and Blackwell, O.M. (1971). Visual performance data for 156 normal observers of various ages, J. Illuminating Eng. Soc. , 1(1), 3–13. Blackwell, H.R. and Blackwell, O.M. (1980). Population data for 140 normal 20–30 year olds for use in assessing some effects of lighting upon visual performance, J. Illuminating Eng. Soc. , 9(3), 158–174. Blackwell, H.R. and Blackwell, O.M. (1980). Individual responses to lighting parameter for a population of 235 observers of varying ages, J. Illuminating Eng. Soc. , 9(4), 205–232. Broen, N.L. and Chiang, D.P. (1996). Braking response times for 100 drivers in the avoidance of an unexpected obstacle as measured in a driving simulator, Proc. Hum. Factors Ergonomics Soc. 40th Annu. Meeting , 900. Brown, T.L. (2000). Modeling driver performance: the effects of rear-end collision warning algorithms, unpublished doctoral dissertation, University of Iowa, Iowa City. Bundensen, C. (1990) A theory of visual attention, Psychol Rev. , 97(4), 523–547. Carpenter, S. (2001). Sights unseen, Monitor on Psychology , April, 54–57. Cohen, A.S. (1987). The latency of simple reaction on highways: A field study, Public Health Rev. , 15, 291–308. Crawford, A. (1962). The perception of light signals: the effect of the number of irrelevant lights, Ergonomics , 417–428. Crawford, A. (1963). The perception of light signals: The effect of mixing flashing and steady irrelevant lights, Ergonomics , 287–294. Currie, L. (1969). The perception of danger in a simulated driving task, Ergonomics , 12, 841–849. Deubel, H. and Schneider, W.X. (1996). Saccade target selection and object recognition: Evidence for a common attentional mechanism, Vision Res. , 36, 1827–1837. Dewar, R.E. and Olson, P.L. (2002). Human Factors in Traffic Safety (Ch. 4). Tucson, AZ: Lawyers and Judges Publishing, Inc.


429

Dilich, M.A., Kopernik, D., and Goebelbecker, J. (2002). Evaluating driver response to a sudden emergency. (Technical paper # 2002–01–0089). Warrendale, PA: Society of Automotive Engineers. Fambro, D., Koppa, R.J., Picha, D.L., and Fitzpatrick, K. (1998). Driver perception—brake response in stopping sight distance situations. (Transportation Research Record 1628). Washington, D.C.: National Academy Press. Farber, E. (1988). Revising the DETECT seeing distance model. (SAE Technical paper #880716). Warrendale, PA: Society of Automotive Engineers. Farber, E. and Matle, C. (1988). PCDETECT: a revised version of the DETECT seeing distance model. (TRB Technical Paper #880547). Washington, D.C.: Transportation Research Board, National Academies P.2. Fisher, J. (1992). Testing the effect of road traffic signs’ information value on driver behavior, Human Factors , 34(2), 231–237. Fugger, Wobock, Randles, Stein, and Whiting (2001). Driver characteristics at signal controlled intersections. (2001–01–0045) Warrendale, PA: SAE. Fuller, R. (1984). A conceptualization of driving behaviors as threat avoidance, Ergonomics , 27(11), 1139–1155. Glencross, D.J. and Anderson, G.A. (1976). Operator response factors in the location and control of foot pedals, Ergonomics , 19, 399–408. Goudie, D.W., Bowler, J.J., Brown, C.A., Heinrichs, B.E., and Siegmund, G.P. (1995). Tire friction during locked wheel braking. (SAE Paper No. 2000–01–1314). Warrendale, PA: Society of Automotive Engineers. Graham, D.J., King, L.E., Naylor, D., and Blakley, S. (1995). Intersection design decision/reaction times for older drivers. (NTIS PB96–139332). Springfield, VA: Federal Highway Administration. Green, M. (2000). How long does it take to stop? Methodological analysis of driver perceptionbrake times, Transp. Hum. Factors , 2, 195–216. Hancock, P.A., Wulf, G., Thom, D., and Fassnacht, P. (1990). Driver workload during differing driving maneuvers, Accident Anal. Prev. , 22, 281–290. Hankey, J.M., McGehee D.V., Dingus, T.A., Mazzae, E.N., and Garrott, W.R. (1996). Initial driver avoidance behavior and reaction time to an unalerted intersection incursion, Proc. Hum. Factors Ergonomics Soc. 40th Annu. Meeting , pp. 896–899. Harrell, W.A. (1994). Effects of pedestrians’ visibility and signs on motorists’ yielding, Perceptual and Motor Skills , 78, 355–362. Hoffman, E.R. (1991). Accelerator-to-brake movement times, Ergonomics , 34, 277–287. Hoffman, E.R. and Mortimer, R.G. (1996). Scaling of relative velocity between vehicles, Accident Anal. Prev. , 128, 415–427. Hole, G.J. and Tyrell, L. (1995). The influence of perceptual “set” on the detection of motorcyclists using daytime lights, Ergonomics , 38, 1326–1341. Johnson, M.B., Voas, R.B., Lacey, J.H., McKnight, A.S., and Lange, J.E. (2004). Living dangerously: driver distraction at high speed, Traffic Injury Prev. , 5, 1–7. Kane, M.J., Pearce, K.D., Hancock, P.A., Scallen, S.F., and Heniff, C.B. (1999). Investigating differences in driver accident involvement: the influence of perceptual motor competence, competitive athletics, and gender, University of Minnesota, Human Factors Research Laboratory. Koppa, R.J., Fambro, D.B., and Zimmer, R.A. (1996). Immediate and long-term effects of glare from following vehicles on target detection in driving simulator, human performance, driving simulation, information systems, and older drivers. (Transportation Research Record 1550). Washington, D.C.: National Academy Press. Lassiter, G.D. (2002). Illusory causation in the courtroom, Current Directions in Psychological Science , 11, 204–208.


430

Lassiter, G.D., Geers, A.L., Munhall, P.J., Ploutz-Snyder, R.J., and Breitenbecher, D.L. (2002). Illusory causation why it occurs, Psychological Sci. , 13, 299–305. Lerner, N.D., Harpster, J.L., Huey, R.W., and Steinberg, G.V. (1997). Driver backing-behavior research implications for backup warning devices. In J.Overton (Ed.), Human Performance in Intelligent Transportation Systems, Information Systems, and Highway Design and Older Drivers . (Transportation Research Record No. 1573, pp. 23–29). Washington, D.C.: National Academy Press. Lerner, N.D., Huey, R.W., McGee, H., and Sullivan, S. (1996). Older driver PRT for intersection sight distance and object detection. Accident Invest. Q. , 9, 20. Liebermann, D.G., Ben-David, G., Schweitzer, N., Apter, Y., and Parush, A. (1995). A field study on braking responses during driving. I. Triggering and modulation. Ergonomics , 38, 1894– 1092. Liebowitz, H.W. and Owens, D.A. (1991). Can normal outdoor activities be carried out in civil twilight? Appl. Opt. , 30, 3501–3503. Louma, J. (1988). Driver’s eye fixations and perceptions. In Gale, A.G. et al. (Eds.), Vision in Vehicles II . Amsterdam: Elsevier Science Publishers, 231–237. Mack, A. (2003). Inattentional blindness: looking without seeing, Curr. Directions Psychol. Sci. , 12(5), 180–184. Mazzae, E.N., Barickman, E, Baldwin, G.H.S., and Forkenbrock, G. (1999a). Driver crash avoidance behavior with ABS in an intersection incursion scenario on dry versus wet pavement. (SAE Paper No. 1999–01–1288). Warrendale, PA: Society of Automotive Engineers. Mazzae, E.N., Baldwin, G.H.S., and McGehee, D.V. (1999b). Driver crash avoidance behavior with ABS in an intersection incursion scenario on the Iowa driving simulator. (SAE paper No. 1999–01–1290). Warrendale, PA; Society of Automotive Engineers. McGehee, D.V., Mazzae, E.N., and Baldwin, G.H.S. (2000). Driver reaction time in crash avoidance research: validation of a driving simulator study on a test track, Proc. Int. Ergonomics Assoc. 2000 Conf. , 3, 320–323. McGehee, D.V., Brown, T.L., Wilson, T.B., and Burns, M. (1998). Examination of driver’s collision avoidance behavior in a lead vehicle stopped scenario using a front-to-rear-end collision warning system. (Contract DTNH22–93-C-07326). Washington, D.C.: National Highway Traffic Safety Administration. Most, S.B., Simons, D.J., Scholl, B.J., Jimenez, R., Clifford, E., and Chabris, C.E (2001). How not to be seen: the contribution of similarity and selective ignoring to sustained inattentional blindness, Psychol Sci. , 12, 9–17. Mourant, R.R. and Rockwell, T.H. (1972). Strategies of visual search by novice and experienced drivers, Human Factors , 14, 325–335. Muttart, J.W. (2000). Effects of retroreflective material upon pedestrian recognition at night, Accident Reconstruction J. , 11(1), 51–57. Muttart, J.W. (2001). Evaluation of the influence of several variables upon driver perception response times, York, England: Proc. 5th Int. Conf. Instit. Traffic Accident Investigators , 116– 129. Muttart, J.W. (2002). Development and evaluation of driver response time predictors based upon experimental results. Master’s thesis, West Hartford, CT: University of Hartford. Muttart, J.W. (2003a). Development and evaluation of driver response time predictors based upon meta-analysis. (SAE paper No. 2003–01–0885). Warrendale, PA: Society of Automotive Engineers. Muttart, J.W. (2003b). Evaluation of methods for estimating driver response times, Insti. Proc. 6th Int. Conf. Instit. Traffic Accident Investigators , Stratford-upon-Avon, England. Muttart, J.W. and Bonnett, G. (2004). DRIVE3: driver response in various environments estimated empirically. Rockledge, FL: REC-TEC, LLC. Muttart, J.W. (2004). DRIVE3: a simplified method for estimating driver response, Auckland, NZ: Australasian S. Pacific Assoc. Crash Investigators 2004 Conf. Proc. , 1–10.


431

Muttart, J.W., Northrop, J., Russell, G., Reynolds, M., Muir, B., and Muttart, M. (2003). Identification distances of pedestrians and objects at night, Can. Assoc. Tech. Accident Investigators Annu. Meeting . Naylor, D.W. and Graham, J.R. (1997). Intersection design and decision-reaction time for older drivers, In J.Overton (Ed.), Human Performance in Intelligent Transportation Systems, Information Systems, and Highway Design and Older Drivers , (Transportation Res. Rec. 1573, 68–71). Washington, D.C.: National Academy Press. OECD Research Group (1970). Driver behaviour, in Synopsis of the State-of-the-Art of Driver Behaviour Research , Department of Civil Engineering, Monash University. Olson, P.L. and Sivak, M. 1983. Comparison of headlamp visibility distance and stopping distance. Perceptual Motor Skills , 57, 1177–1178. Olsen, P.L. and Sivak, M. (1984). Glare and headlighting design. (SAE 840047). Warrendale, SAE, 1985. Olson, P.L. and Sivak, M. (1986). Perception-response time to unexpected roadway hazards, Human Factors , 28(1), 91–96. Olson, P.L. (1996). Forensic Aspects of Driver Perception and Response , Tucson, AZ: Lawyers and Judges Publishing, Inc. Otto, W.M., Otto, C.L., and Overton, R.K. (1980). Response characteristics of motorcycle riders to a complex emergency situation, Washington, D.C.: Int. Motorcycle Safety Conf. Proc. , IV, pp. 1489–1503. Owens, D.A, Francis, E.L., and Liebowitz, H.W. (1989). Visibility distance with headlights: a functional approach. (Technical paper No. 890684). Warrendale, PA: Society of Automotive Engineers. Parkes, A. and Hooijmeijer, V. (2000). The influence of the use of mobile phones on driver situation awareness, Cawthorne, England: Transport Research Laboratory. Petersen, H.E. and Dugas, D.J. (1972). The relative importance of contrast and motion in visual detection, Human Factors , 14(3), 207–216. Rasanen, M., and Summala, H. (1998). Attention and expectation problems in bicycle-car collisions: an in-depth study, Accident Anal Prev. , 30, 657–666. Reed, W.S., and Keskin, A.T. (1989). Vehicular deceleration and its relationship to friction. (Technical Paper # 890736). Warrendale, PA: Society of Automotive Engineers. Resnick, R.A. (2004). Visual sensing without seeing, Psycholog. Sci. , 15, 27–32. Roper, V.J. and Howard, E.A. (1938). Seeing with motor car headlamps, Illuminating Eng. Soc. , 33, 417–438. Scott, P.A., Candler, P.D., and Li, J.C. (1996). Stature and seat position as factors affecting fractionated response times in motor vehicle drivers, Appl Ergonomics , 27, 411–416. Sidaway, B., Fairweather, M., Sekiya, H., and McNitt-Gray, J. (1996). Time-to-collision estimation is a simulated driving task, Human Factors , 38, 101–113. Simon, D.J. and Chabris, C.R (1999). Gorillas in our midst: sustained inattentional blindness for dynamics events, Perception , 28, 1059–1074. Sivak, M., Post, D.V., Olson, P.L., and Donohue, R.J. (1981). Automobile rear lights: effects of the number, mounting height, and lateral position on reaction times of following drivers, Perceptual and Motor Skills , 52(3), 795–802. Sohn, S.Y. and Stepleman, R. (1998). Meta-analysis on total braking time, Ergonomics , 41, 1129– 1140. Southwestern Association of Technical Accident Investigators (1999). Research results for subjects who responded to mock skateboarders. (Unpublished data). Summala, H. (2000). Commentary: brake reaction times and driver behavior analysis, Transp. Hum. Factors , 2, 217–226. Summala, H., Lamble, D., and Laakso, M. (1998). Driving experience and perception of the lead car’s braking when looking at in-car targets, Accident Anal Prev ., 30(4), 401–407.


432

Tabachnick, E.C. and Fidell, L.S. (1996). Using Multivariate Statistics , 3rd ed., New York: Harper Collins College Publishing. Todosiev, E.P. (1965). Velocity thresholds in car following, Columbus, OH: Ohio Department of High-ways. Triggs, T.J. and Harris, W.G. (1982). Reaction time of drivers to road stimuli, (HFR No. 12) Victoria, Australia: Monash University, Human Factors Group. Turner, J.D., Simmons, C.J., and Graham, J.R. (1997). High-visibility clothing for daytime use in work zones. (Paper No. 970430). Washington, D.C.: Transportation Research Board, van Winsum, W. (1998). Preferred time headway in car-following and individual differences in perceptual motor skills, Perceptual Motor Skills , 87, 863–873. Wortman, R.W. and Matthias, J.S. (1983). An evaluation of driver behavior at signalized intersections. (PB83 218560). Springfield, VA: FHWA. U.S. Department of Transportation (2001). Manual on Uniform Traffic Control Devices for Highways and Streets , Washington, D.C.: Federal Highway Administration. Vaughan, R. and Vaughan, P. (2001). Evaluation of the influence of several variables upon driver perception response times, Instit. Traffic Accident Investigators Int. Conf. Proc. , York, England, pp. 1–8. Yoo, H., Tsimhoni, O., Watanabe, H., Green, P., and Shah, R. (1999). Display of HUD warnings to drivers: Determining an optimal location. (Technical report number UMTRI-99–9). Ann Arbor, MI: University of Michigan, Transport Research Institute.

Appendix A: Operationalization of Terms A=Anticipation 1. Subject knows stimulus and response multiple exposures 2. Subject knows stimulus and response infrequent exposure (1 or 2) 3. Stimulus or response is unknown, multiple exposures 4. Stimulus or response is unknown, single exposure 5. Stimulus and response are unknown, single exposure Al=Alcohol (in BAC%) Age=Age (in years) C=Contrast (Luminance reflectivity) 1. Audible stimuli 2. Illuminated object (includes reflective objects) 3. High contrast (>25%) 4. Moderate contrast (10 to 25%) 5. Low contrast (6 sec 4.1 >4 sec 4.7 >2 sec 6.5 a Feet per second (1 ft/s=0.305 m/s). Source: Eubanks, J.J. and Hill, P.L., 1998, Pedestrian Accident Reconstruction and Litigation (2nd ed.). (Tucson, AZ: Lawyers & Judges Publishing Co.).

slower pedestrians use the crosswalk; however, it seems reasonable to assume that any crosswalk will occasionally be used by slow walkers.


472

Walking speeds and start-up times of 7123 pedestrians, more than half of whom were over the age of 65, were observed by Knoblauch et al. (1996) at a variety of urban intersections under a number of conditions. Older pedestrians were significantly slower than those under 65 and walked more slowly when it was snowing or when the street was covered by snow than under other weather conditions. The mean and 15th-percentile walking speeds were 1.46 and 1.21 m/s, respectively, for young (under 65) pedestrians, and 1.20 and 0.94 m/s, respectively, for older pedestrians. Mean start-up time, from the start of the WALK signal to the moment at which the pedestrian stepped off the curb and started to cross, was longer for older pedestrians (2.48 s) than for younger ones (1.93 s). 16.6 Perception-Response Time One of the human factors often cited by those investigating traffic accidents is perception-response time (PRT), that is, how quickly a driver or pedestrian can respond to an emergency situation and how much distance was between the driver and the pedestrian when the latter was first seen. This plays a role in many traffic accident litigation cases. The components of PRT are search, detection, recognition, decision, and action. PRT slows with fatigue, alcohol, distraction, darkness, and age (older drivers are slower by 15 to 25%). Unexpected situations can slow PRT by 30% or more (Olson and Farber, 2003). Responses are also slower in a more complex driving situation. A reasonable PRT for an alert driver in most situations would be of the order of 1.5 s, but it can be somewhat longer than this, depending on the situation (e.g., darkness) and the age and condition of the person involved in the accident (Olson and Farber, 2003). Although PRT among pedestrians has not been well documented, it seems reasonable to suggest that they would be quicker to respond in an emergency than would drivers because the latter are preoccupied with the driving task, which typically requires greater attention than does crossing the street. Each accident must be examined on its own merits, taking into account the various factors that might have influenced PRT. 16.7 Alcohol Use Driving under the influence of alcohol is a contributing factor in a high proportion of traffic accidents. The “impaired driver” has received a great deal of attention over the years; however, less well known is the influence of alcohol impairment on pedestrians involved in crashes. It is much higher than most people realize. For example, of the pedestrians killed in the U.S. in 1998, 31% were intoxicated, as defined by the maximum blood alcohol content (BAC) allowed for drivers, while the intoxication rate for drivers involved was 12% (BAC 0.10 or higher). Alcohol involvement was greater in accidents in which the pedestrian was walking along the road and was struck from behind. Alcohol influences perception and behavior in a number of ways. It not only slows responses of drivers and pedestrians but also impairs vision (e.g., reduced visual scanning of the environment, reduced night vision), slows decision making, and impairs judgment. In particular, impaired road users have difficulty dividing attention and in making

Pedestrian accidents in traffic

473

decisions, have problems in complex driving situations, and are more easily distracted (Moskowitz, 2002). Thus, drivers and pedestrians who are impaired may not notice each other. They also have difficulty recognizing their degree of impairment when under the influence of alcohol. With these limitations it is easy to see how drivers and pedestrians can make errors that lead to collisions. The effects of alcohol are to reduce the likelihood of detection except when a vehicle is traveling slowly (32 km/h) or the pedestrian is wearing reflectorized marking. Pedestrians with black or gray clothing can almost never be detected within the critical stopping distance by drivers traveling at 97 km/ h or greater (Olson and Farber, 2003). It should be noted that in some jurisdictions it is illegal for a drunk pedestrian to be on a roadway. Therefore, an accident is often the responsibility of the impaired pedestrian, rather than (or in addition to) the driver. However, drivers are also expected to exercise precaution when observing an obviously intoxicated person. This assumes that drivers are capable of making a judgment about degree or impairment of pedestrians. 16.8 Child Pedestrians The behavior of child pedestrians is different from that of adults. Children’s conception of safety is poorly formulated, and their mental representation of relevant behaviors, such as crossing the street, is not well developed. Road accident rate is a function of the age of young pedestrians. It is greatest for those in the 3- to 8-year range. Children aged 5 to 7 are most over-represented, even though they are less exposed to traffic than older children. Boys have more accidents than girls, in part because they also spend more time on the streets. As indicated earlier, street crossing involves observation, perception, judgment, and decision. The road is scanned, traffic is perceived, and judgments are made about the distance and movement of vehicles. On the basis of this information, a decision is made about whether to cross the road, and where to cross. Young children have difficulties with one or more of these subtasks. Children’s perception of driver behavior and cause and effect is not well developed (Sandels, 1968). Drivers are much less likely than young pedestrians to take evasive action in a potential conflict situation. In an observational study, it was found that almost all child pedestrians took action if the vehicle was closer than 20 m, and the majority did so even when the vehicle was 40 m away (Howarth, 1985). Drivers took action when a pedestrian stepped onto the road only if the vehicle was closer than 20 m. The reasons for the relatively high accident rate among young pedestrians are many. Sandels (1968) has examined the behaviors and cognitions of children and suggests that the necessary degree of maturity for safe behavior is reached between the ages of 9 and 12. Several factors contribute to child pedestrian accidents: • They have difficulty distributing their attention and are easily preoccupied or distracted in hazardous traffic situations. • They have difficulty in correctly perceiving the direction of sound and the speed of vehicles.


474

• Many believe the safest way to cross the street is to run and that it is safe to cross against the red light. • They have a poor understanding of the use of traffic control devices and crosswalks. Perception of oncoming traffic requires a field of view that may not be available to a small child, who cannot see over parked cars. Young children have limited ability to process information in peripheral vision (i.e., to see “out of the corner of the eye”) and thus need more time to react once objects in the periphery are seen. Planned search is often poor because young children have problems dividing their attention and are attracted by the most salient objects in their environment (e.g., an animal, another child). Systematic search begins about the age of 6. Children under age 6 have difficulty localizing sound, so they may hear a car coming but not know where to look for it. Finding relevant information in a complex environment is difficult for most children, partly due to limitations in speed of eye movements, attention, and memory, rather than visual acuity. A crucial judgment in traffic is whether a vehicle is moving, how fast, and at what distance. Children are generally unreliable in judgments of distance of cars. They often make the error of assuming a greater distance for smaller vehicles; thus, when they encounter small cars, smaller gaps are accepted in crossing the street. Size constancy— the ability to judge accurately the size of objects independent of their distance from the observer—develops with age, as the child learns to interpret different environmental cues to distance. Young children have poor size constancy, so they could easily overestimate the distance of vehicles. A factor in avoiding danger on the road is the speed with which a pedestrian can recognize and react to the situation. Visual reaction time decreases with age in children, by a factor of about three between the ages of 4 and 17 (Reiss, 1977). Auditory reaction times are also slower for younger children and their attention span is shorter than in adults. They have difficulty attending to more than one thing at a time. For this reason, they often run onto the road into the path of moving vehicles in pursuit of objects such as balls. A large proportion of child pedestrian accidents are the result of unsafe or illegal actions on the part of the child. The ability of children between the ages of 5 and 14 to judge the velocity of oncoming cars (as slow, medium, or fast) was studied by Salvatore (1974). Vehicles that could be heard by the subjects were more likely to be classified as fast; inaudible vehicles were judged to be slow. Judgment of vehicle speed was slightly more accurate at near (76 m) than at far (153 m) distances. Many young children are unable to make appropriate use of the information in the traffic environment to judge safe gaps in traffic. In a study of vehicle gap perception in New Zealand, Connelly and colleagues (1998) had children in three age groups (5 and 6; 8 and 9; and 11 and 12) indicate when it was safe to cross a road as vehicles approached an intersection from the right (equivalent to left in Europe and North America). The threshold for distance gaps ranged from 10 to 146 m. Vehicle speeds appeared to be ignored in these estimates because the children set similar distance thresholds at all vehicle speeds. The 8- and 9-year-olds selected the least safe gaps, and most children were safest at the lowest vehicle speeds. The mean times of arrival dropped from 2.34 s with vehicles going up to 50 km/h to 0.47 s for those speeds over 65 km/h. About three fourths of the judgments made were considered safe. When


475

asked what information they used in making judgments, 63% indicated distance, 19% indicated speed, and the remainder said both or were unable to express a decision. Zeedyk and associates (2002) observed the actions of children aged 5 and 6 years in a realistic road-crossing situation. Performance by the children was “extremely poor,” with 60% failing to stop before entering the road and fewer than 42% looking for oncoming traffic. When children did look initially, it was often in an inappropriate direction. When crossing between parked cars, only 29% looked before reaching the curb. 16.9 Elderly Pedestrians Pedestrians over the age of 70 are more likely to be involved in a severe accident than are younger ones. This is due in part to greater vulnerability of older people because of physical fragility, more easily broken bones, longer recovery times, etc. An examination of age differences indicated that pedestrians over 64 accounted for 45% of the fatalities but only 15% of the western European population (Choueiri et al., 1993). In the U.S., the older group constituted 22.4% of the pedestrian fatalities and 12.3% of the total population. Problems leading to accidents among older pedestrians include (McKnight, 1988): • Gap judgment—misjudging vehicle speeds and the distances of and intervals between approaching vehicles • Attention—stepping off the sidewalk when distracted • Visual search—watching the traffic signal instead of the traffic • Expectation—misinterpreting vehicle movement • Assuming that drivers will yield • Haste—impatiently crossing after waiting • Crossing midblock between parked cars • Slow walking speed Older pedestrians often have difficulty in assessing the speed of approaching vehicles, thus misjudging when it is safe to cross the road. Even though the vast majority of their accidents occur during the daytime, they frequently report failure to see, or to see in time to take evasive action, the vehicle that struck them. Sheppard and Pattinson (1988) found that about two thirds of the older pedestrians they surveyed saw the vehicle that struck them only when it was within 9 m of them. In addition, many

TABLE 16.5 Mean Walking Speedsa for Disabled Pedestrians and Users of Various Assistive Devices Disability/assistive device Cane/crutch Walker Wheelchair Immobilized knee Below-knee amputee Above-knee amputee

Speed 2.62 2.07 3.55 3.50 2.46 1.97


476

Hip arthritis 2.24 to 3.66 Rheumatoid arthritis (knee) 2.46 a Feet per second; 1 ft/s=0.305 m/s. Source: Perry, J., 1992, Gait Analysis (New York: McGraw Hill).

indicated that the vehicle that struck them was doing something unusual before the collision. Frequent reports related to this were: vehicle reversed into me (30%), expected driver to stop or alter course (20%), thought it was not moving (11%), and vehicle came from behind a corner, parked car, etc. (10%). It is possible that reduction in peripheral visual information processing was a contributing factor. Concerning their ability to judge speeds of approaching cars, about 30% said that they could do this “not well at all.” 16.10 Handicapped Pedestrians We tend to think of a handicapped individual as a person in a wheelchair; however, a variety of other handicaps are relevant to understanding pedestrian behavior. Vision and hearing deficits can pose difficulties in traffic. The “handicapped” pedestrian includes those with physical problems, such as restricted mobility or perception, as well as those whose mobility is temporarily reduced because they are encumbered by carrying luggage, packages, children, etc. Handicaps are not only physical in nature, but also mental. Pedestrians with low intelligence, illiteracy, dyslexia, attention deficit, and other such problems often have difficulties understanding traffic control devices. Perry (1992) reported lower speeds for physically impaired pedestrians, as would be expected (see Table 16.5). None of these reaches the average walking speed assumed for design of pedestrian walk-signal timing. 16.11 Nighttime Conditions When light reaches a surface, some of it is absorbed, and some is reflected. Reflectivity refers to the extent to which an object or environment reflects the light falling on it. Nothing in the natural world reflects all the light that it receives. The more distant an object is from a light source, the less light will fall on it. In order to understand pedestrian visibility, especially at night, the concepts of contrast and reflectivity need to be appreciated. Contrast is the difference in brightness between a target and the background against which it is viewed. Positive contrast occurs when an object is brighter than its background, and negative contrast is the opposite. Print on a white page is an example of the latter. There must be a certain contrast ratio before an object is visible. The greater the contrast is, the more likely the object is to be detected. Dark-clad pedestrians at night are more difficult to detect, except when seen against a bright surround. White clothing is highly reflective, but still only about 60 to 70% so. The reflectivity of the roadway background ranges from 2 to 10%, and that of blue denim is about 3% (Olson and Sivak, 1984). Research on the reflectivity of pedestrian clothing suggests that about 60% of it has a reflectivity value of 10% or less (Olson, 2002b). The reflectivity of pavement is typically about 6%; that of wet pavement is about half that of dry pavement,


477

especially at distances of less than 121 m. The combination of wet pavement and glare makes it very difficult to see pedestrians at night. Olson and Farber (2003) studied the typical contrast of different parts of an adult pedestrian with clothing of 6% reflectance seen with vehicle low-beam headlights at night and determined that the contrast is greatest at the waist and knees, less at the head, and least at the feet. Even though the feet would be receiving the greatest amount of illumination from low beams, the contrast with the pavement would be relatively less than the contrast of other parts of the body seen against darker pavement (darker because it is further away and receiving less light from the headlights). The upper body would be better illuminated by high beams and would be seen against the darker background of the distant pavement. If the pedestrian were wearing a light-colored top, contrast would be that much better. However, whether the pedestrian would be seen in time to stop depends on the pedestrian’s position and the vehicle’s speed. A bright roadway (e.g., illuminated by a street light, as might be the case in an urban area) behind a dark-clad pedestrian will make him more visible, but the same pedestrian could be nearly invisible in a dark environment until vehicle headlights shine on him at close range (Janoff, 2002). Factors influencing nighttime visibility are: target distance, size, contrast with background, time spent looking for the target, glare, and driver age. The evidence on roadway lighting indicates that accidents reduce when street lights are installed (Janoff, 2002). Glare from headlights of oncoming vehicles obviously makes it more difficult to see pedestrians on the street or the highway. With headlights on low beam, pedestrians will not be visible as far away as with high beam. However, with high beam, less light is cast on the bottom part of a nearby pedestrian and those wearing light pants and dark top may be less visible than expected under high beam. Rain at night reduces pedestrian visibility further. Wet pavement at night greatly increases the amount of glare in oncoming drivers’ eyes from reflections of headlights and street lights off the pavement. The detrimental effects of oncoming headlight glare are described by Olson and Farber (2003). A driver’s visibility distance to detect a 12% reflective target on the right side of the road is unaffected until the headlights are about 300 m away. As the vehicle comes closer, this distance reduces from approximately 90 to 64 m just before the vehicle is passed. With a target on the left, visibility distance begins to reduce at a distance of about 600 m and decreases fairly quickly from 50 to less than 30 m as the distance between the vehicles decreases from 300 to 30 m. It is evident that glare from headlights can make a pedestrian on the road essentially invisible in a collision until the last second or two before impact. Olson (2002b) measured pedestrian visibility under different nighttime viewing conditions. Details of his findings are seen in Table 16.6. It can be seen that the probability of detecting a pedestrian in time at night is much less when the pedestrian is on the driver’s left than when on the right, and less when wearing a dark top. When the pedestrian was unexpected (i.e., drivers were not alerted to the possibility of a pedestrian), probability of hitting the pedestrian increased dramatically under most conditions. It should be noted that the data reported here are for young drivers in a situation in which they expected a pedestrian. Older drivers and those in a typical driving situation would detect the pedestrians at a somewhat closer distance, and all drivers must be closer when the pedestrian is unexpected, as on the highway, for example.


478

TABLE 16.6 Percentage of Trials in Which Drivers Would NOT Have Been Able to Stop in Time to Avoid Hitting Pedestrian as Function of Clothing/ Position on Road Speed (km/h)

Dark/ right

Dark/left

White top/right

40 56 73 89 105

White top/left

90 >90 65 105 >90 >90 90 Source: Olson, P.L. and Farber, E., 2003, Forensic Aspects of Driver Perception and Response (2nd ed.). (Tucson, AZ: Lawyers & Judges Publishing Co.).

90

Visual acuity at night is reduced, for example, for those with average vision, from 20/20 in daylight to 20/25 or 20/30 at night; it is typically 20/40 or worse for 70-yearolds. Another factor is that most drivers overdrive their headlights at night. Fog and smoke also reduce night visibility because they reduce contrast of the roadway scene. Glare of another sort (from the sky) can reduce visibility at dusk and dawn because the bright sky viewed just after sunset or before sunrise causes light adaptation, making the eyes less sensitive to objects on the road ahead. Similarly, sun glare just before sunset or after sunrise when an individual is driving toward the sun is a major cause of reduced visibility. It can be seen that pedestrian visibility at night is greatly reduced for several reasons, including dark clothing, headlight condition (beam used, aim), glare, reduced visual information for the driver, and pedestrian and/or driver impairment from alcohol. When a nighttime pedestrian accident is investigated, it is essential to consider all of these factors, as well as driver and pedestrian age. 16.12 Weather Conditions The presence of ice patches, snow, and slush on roads and sidewalks often leads to walking difficulties for pedestrians. The most obvious problems are poor footing, increased chance of slipping, and slower walking speeds. These conditions not only make walking difficult and dangerous, but also distract pedestrians’ attention from vehicles on the roadway. The presence of snow makes it difficult to detect curbs, potholes, and debris


479

on the road, thus increasing the chances of a fall. One of the main fears among the elderly is falling, which is clearly justified when walking outside in winter conditions. Additional problems encountered by older people include greater susceptibility to glare from snow and ice, poor contrast due to glare and light conditions, and more hours of darkness. Blowing snow can also distract pedestrians’ attention from traffic or make it difficult to detect vehicles or judge vehicle speed. The observational study by Knoblauch et al. (1996) found that walking speed reduced when snow was present, thus increasing the length of time that the pedestrian is exposed to traffic when crossing the road. 16.13 Roadway Design Roadway design features such as road width, number of lanes, intersection complexity, and refuge islands have an impact upon pedestrian behavior and safety. A search of the literature on pedestrian safety (Zegeer, 1983) revealed several treatments to increase safety. Use of one-way streets will lessen the complexity of crossing for pedestrians because they need to look in only one direction for traffic. Also, drivers can devote more attention to pedestrian traffic because vehicle traffic is moving in only one direction. Left turns are easier here for drivers and safer for pedestrians. A study of 1297 intersections in 15 U.S. cities indicated fewer pedestrian accidents at intersections of one-way streets than at those involving two-way streets. Stemley (1998) indicates the advantages of one-way streets to be: • They are safer as a result of fewer conflict points at intersections (fewer turning movements). • More gaps are available for pedestrians because they need only look in one direction for gaps. • Pedestrians cannot be caught between opposing lines of traffic when crossing. • Drivers and pedestrians are more likely to see each other. • Pedestrians are less likely to become frustrated waiting to cross and to engage in dangerous behavior. • Drivers making turns can monitor pedestrian movements more easily because they do not need to watch for a gap in oncoming traffic. Oxley et al. (1997) found that older pedestrians experience difficulties when crossing two-way streets, as compared with one-way streets, because they tend to cross with closemoving traffic in the near-side lane and with oncoming traffic in the far-side lane. They do not compensate appropriately for their slower walking speeds. Greatest difficulties on two-way streets were experienced by pedestrians 70 years of age and over. These pedestrians waited at the curb longer, delayed their departures longer, and looked at the ground more while crossing than did others. They often crossed with smaller safety margins than did others. A collection of road modification techniques referred to as “traffic calming” has come into use in recent decades, especially in Europe, to slow traffic and increase pedestrian safety. For a detailed review, see O’Brien and Brindle (1999).


480

16.14 Pedestrian Traffic Control Devices Pedestrian laws and traffic control devices are poorly understood, according to a large survey conducted by Tidwell and Doyle (1995). For example, 83% of drivers did not know the difference between an ADVANCE PEDESTRIAN CROSSING sign and a PEDESTRIAN CROSSING symbol sign (which includes the crosswalk lines). This study also found that pedestrians are ignorant about many of the basic rules of the road. About half thought that jogging on the road is legal and 80% thought that wearing white clothing at night provided adequate visibility to drivers. According to a questionnaire survey of more than 4700 people by Tidwell and Doyle (1995), just under half indicated that the flashing DON’T WALK signal means to return to the curb, and a similar proportion thought that a WALK signal meant there were no turning conflicts. They also found that about half of the people in a large survey felt that the WALK signal guarantees their safety. Pedestrians often become impatient and start across the street before the WALK signal activates or start across after the DON’T WALK signal has come on. The proportion of pedestrians violating signals has been reported by Virkler (1998) to be about 16% for those entering the crosswalk late and 10% for those starting early. It may be assumed that the use of pavement markings for uncontrolled crosswalks would lead to greater pedestrian safety. However, a large U.S. study (Herms, 1972), based on 5 years of data at 400 intersections (each with one painted and one unpainted crosswalk), found that pedestrian accident rates were about double at marked, as opposed to unmarked, crosswalks. A more recent study of marked and unmarked crosswalks gathered data at 1000 marked and 1000 unmarked pedestrian crosswalks in 30 U.S. cities (Zegeer et al., 2001). Information collected included pedestrian volume, traffic volume, accident history, speed limit, number of lanes, and pavement marking patterns. Crossings with traffic calming measures and school crossings were not included, nor were those with traffic signals or stop signs. Crash rate was lower with raised medians or raised crossing islands with both types of crosswalks, but painted medians did not benefit pedestrians. The factors having no effect on pedestrian crash rate were area (residential, central business district, midblock), traffic operations (one- vs. two-way traffic) and pavement marking pattern. On two-way roads, no differences were found between marked and unmarked crossings. Similarly, on multilane roads with an ADT of 12,000 or less, the presence of markings made no difference in crash rate. However, higher crash rates were found at marked crossings on multilane roads with no raised medians and an ADT of more than 12,000. Even with raised medians, locations with more than 15,000 vehicles per day had higher crash rates with marked crosswalks. Use of marked crosswalks may induce more pedestrians to cross there instead of at a signalized crossing. Marked crossings at multilane sites are especially prone to multiple-threat type collisions—when one driver stops for the pedestrian and another driver approaching in the same direction does not stop and hits the pedestrian. In these cases, usually pedestrian and driver fail to see each other in time. In the Zegeer et al. study, this type of crash constituted 17.6% of all pedestrian collisions in marked crosswalks, but none of these crashes occurred in unmarked crosswalks. The former may lead to a false sense of security and reduced vigilance by pedestrians. Drivers failed to yield to pedestrians in a


481

high proportion of accidents (41.5% of the time at marked and 31.7% at unmarked crosswalks). Dart-out collisions were much more likely at unmarked crossings. Pedestrians failed to yield 5.8% of the time in marked crossings and 34.2% in unmarked crossings. Pedestrians over 65 were over-represented in nearly all situations. Whether drivers slow at crosswalks may depend on the action of pedestrians. Knoblauch and Raymond (2000) examined this issue by staging pedestrian activity at six uncontrolled crosswalks before and after the installation of markings. Approaching vehicles slowed slightly when a pedestrian was not present and when the pedestrian was not looking at the driver, but not when the pedestrian did look at the driver. The interpretation was that the driver assumed the pedestrian was not paying attention when not looking and thus thought greater care was needed. However, none of the vehicles actually stopped in the presence of the pedestrian. 16.15 Work Zones Pedestrian movement in work zones can lead to safety problems. In 2000 in the U.S., 1026 fatalities occurred in work zones, including maintenance and utilities areas. Pedestrians frequently need to be guided through the work zone, must walk over rough surfaces, and be alert to the movement of workers and machines. Ideally, pedestrian traffic should be physically separated from the work area. Vertical curbs, barriers, or chain-link fences can be used to accomplish this. Construction and maintenance workers are particularly vulnerable to roadway traffic, especially in high-speed areas. Although those concerned with traffic movement attempt to maintain roadway speeds close to the speed limit, it is often necessary to slow traffic down; however, many drivers fail to slow or to slow in time to avoid a collision. Each year many workers are injured or killed by roadway traffic. For example, about 30 workers are killed each year in the U.S. by vehicles. Safety benefits can be derived from the use of highly visible clothing. Major work zones require a detailed plan with traffic safety as a high priority. This includes safety to workers and pedestrians as well as motorists. 16.16 In-Line Skates, Skateboards, and Scooters The number of certain nonstandard “pedestrians” has increased greatly over the past decade. Use of inline skates, skateboards, and nonmotorized scooters has become a very popular recreational activity. A large proportion of in-line skating accidents are due partly to inexperience. How in-line skating should be accommodated by the transportation system, including infrastructure design and regulatory and operational principles, is not entirely clear. Skaters, skateboarders, and scooter users appear to be somewhere between pedestrians and bicyclists. They may be considered “assisted pedestrians.” One issue is whether they should be allowed to operate on the road, sidewalk, specific facilities such as bicycle paths, or some combination of these. The number of injuries associated with the use of in-line skates and skateboards has increased substantially. Because the manipulation and control of skates and skateboards


482

is different from that of bicycles and may require considerable practice for some users, especially young children, a number of hazards are associated with their use. The Centers for Disease Control estimates annual in-line skating injuries in the U.S., based on emergency room visits, to be more than 100,000. It is estimated that there are about 29 million skaters in the U.S., thus yielding an injury rate per skater of 0.356%, up from 0.266% in 1993. The most common injuries are to the wrist (24.2%). Primary causes are falls due to spontaneous loss of control (41%) and hitting a stationary object (40%). Key causes of injury are skating out of control (67%), hazardous road condition (53%), and fatigue (37%). Only a small proportion of those receiving such an injury wore a wrist guard at the time of the fall. The likelihood of sustaining a wrist injury when not wearing a wrist guard was determined to be 10.4 times greater than when wearing one. Similar numbers for other parts of the body, with and without relevant gear, are: elbows, 9.5 times, and knees, 2.2 times (Schieber et al., 1996). Experienced skaters commonly reach speeds of up to 27 km/h. A study by Mulder and Hutten (2002) examined in-line skating injuries (excluding traffic injuries) in several European countries and reported that about 65,000 people were treated in emergency departments in 1996. The most common accident mechanism was falling (53 to 96%, depending on the country). The most common injuries by far were to the wrist. The hazards associated with in-line skating include difficulty controlling movement and stopping; the need for considerable minimal lateral operating space, which can endanger others such as pedestrians; and excessive speed. Also, surface conditions (e.g., steep grades, pebbles, speed, potholes, sidewalk grates, bumps, railroad tracks) can impede safe operation. Downgrades play a role in safe skating, with beginners able to handle grades up to 3%; intermediates, up to 5%; and experienced skaters, up to 10%. The maximum distance of a downgrade is recommended to be 100 m. As estimated by the U.S. Consumer Product Safety Commission, in the U.S. in 2000, 50,000 emergency room visits required treatment of skateboarders, an increase from 24,000 in 1994; in 2001, 84,000 scooter riders required treatment. The vast majority of skateboarding injuries occur to those under the age of 15; about 30% of scooter injuries happen to those under 8. The most common injuries are to the head and face. It is recommended that those under 8 use scooters only under direct adult supervision and that scooters not be ridden in streets, at night, and or on surfaces with water, sand, gravel, or dirt present. Although scooter deaths are rare, the majority (about 80%) involve a motor vehicle and the rest involve falls from the scooter. Young children have difficulty in judging their own skill and strength as well as pedestrian and vehicle traffic activity. In addition, the relatively high center of gravity of young children makes balance and control more difficult. Those using skateboards, in-line skates, and scooters are strongly advised to use protective equipment, including helmets, to avoid traveling on rough or wet surfaces, and not to ride at night. 16.17 Conclusion This brief review of accidents among pedestrians shows that these road users are very vulnerable to injury and death. Older pedestrians are more vulnerable than younger ones,


483

due to their slower response and decision times, slower walking speeds, and increased physical frailty. Child pedestrians are also highly vulnerable in the traffic environment, due in part to their risky behavior and their difficulty in dealing safely with vehicle traffic. A recent safety problem has arisen with the increasing use of skateboards and inline skates. Many who use them are young and inexperienced and more likely than adult road users to take risks in traffic. Although it is tempting for the accident investigator to think, “Oh, the poor pedestrian,” when considering pedestrian accidents, it should be kept in mind that pedestrians and drivers are about equally at fault in such collisions. The former often fail to attend properly to the movement of vehicles (assuming that they have the right of way when they do not or that drivers will notice them and stop if necessary) and often take chances when crossing the road. This is particularly the case at night when pedestrians are much less visible than in daylight because of less conspicuity due to low contrast and the effects of glare on drivers’ vision. Pedestrians typically believe that they are more visible to drivers than they really are. On the other hand, drivers frequently do not watch for pedestrians or fail to yield to them when turning. In determining whether a driver had sufficient time to take evasive action (typically, braking) in a pedestrian collision, the speed of pedestrian movement and the reasonable perception-response time on the part of the driver must be considered. Reduced attention and distraction of drivers (especially the use of cell phones) contribute to a large proportion of collisions and are particularly a problem at intersections where vehicles are turning while pedestrians cross in the same direction of travel. Obstruction of vision from within the vehicle can be a problem here. Children pose a particular hazard in many traffic situations because their behavior is quite unpredictable and they often pay little attention to traffic. Older pedestrians may be disadvantaged by reduced walking speed and the inability to react quickly in an emergency. In the investigation of pedestrian accidents, it is important to determine the actions of pedestrian as well as driver. How reasonable were the pedestrian’s behavior and the response of the driver? Did one party fail to yield right of way? Was the pedestrian visible in time for the driver to respond (especially at night)? Was the pedestrian or the driver impaired? Often this question is not asked of pedestrians who are struck. Special consideration may be required in assessing the capabilities (e.g., speed of movement, attention, vision) of children and elderly pedestrians. Finally, it is essential to examine each accident on its own merits because no two pedestrian crashes are identical. References Abdulsattar, H.N. and McCoy, P.T., 1999a, Pedestrian blind-zone areas at intersections. Paper presented at 75th Annual Meeting of the Transportation Research Board, Washington, D.C. Abdulsattar, H.N. and McCoy, P.T., 1999b, Pedestrians’ right of way at signalized intersections. Paper presented at 75th Annual Meeting of the Transportation Research Board., Washington, D.C. Bowman, B.L. and Vecellio, R.L., 1994, Pedestrian walking speeds and conflicts at urban median locations. Transp. Res. Rec. , 1438, 67–73.


484

Caird, J.K. and Hancock, P., 2002, Left-turn and gap acceptance crashes. In R.E.Dewar and P.L.Olson (Eds.) Human Factors and Traffic Safety . (Tucson, AZ: Lawyers & Judges Publishing Co.), 613–652. Choueiri, E.E., Lamm, R.L., Choueiri, G.M., and Choueiri, B.M., 1993, Pedestrian accidents: a 15year survey from the United States and Europe, ITE J. , 63(7), 36–42. Connelly, M.L., Conaglen, H.M., Parsonson, B.S., and Isler, R.B., 1998, Child pedestrian crossing gap thresholds. Accident Anal. Prev. , 30(4), 443–453. Dahlstedt, S. Walking speeds and walking habits of elderly people, National Swedish Road and Traffic Research Institute, Stockholm, Sweden, undated. Eubanks, J.J. and Hill, P.L., 1998, Pedestrian Accident Reconstruction and Litigation . (2nd ed.) (Tucson, AZ: Lawyers & Judges Publishing Co.). Habib, P.A., 1980, Pedestrian safety: the hazards of left-turning vehicles, ITE J. , 50, 33–37. Hauer, E., 1988, The safety of older people at intersections. In Special Report #281, Transportation in an Aging Society (Washington, D.C., Transportation Research Board), 194–252. Herms, B.F., 1972, Pedestrian crosswalk study: accidents in painted and unpainted crosswalks, Highway Res. Rec. , 406, 1–13. Howarth, I., 1985, Interactions between drivers and pedestrians: some new approaches to pedestrian safety. In L.Evans and R.Schwing (Eds.) Human Behavior and Traffic Safety (New York: Plenum Press), 171–178. Hunter, W.W., Stutts, J.C., Pein, WE., and Cox, C.L., 1995, Pedestrian and bicycle crash types of the early 1990s. Report No. FHWARD95163, (Washington, D.C.: Federal Highway Administration). Janoff, M.S., 2002, Visibility under roadway lighting. In R.E.Dewar and P.L.Olson (Eds.) Human Factors and Traffic Safety . (Tucson, AZ: Lawyers & Judges Publishing Co.), 459–491. Johnson, C.D., 1997, Pedestrian fatalities on interstate highways; characteristics and countermeasures. Transp. Res. Rec. , 1578, 23–30. Knoblauch, R.L., Pietrucha, M.T., and Nitzburg M., 1996, Field studies of pedestrian walking speed and start-up time, Transp. Res. Rec. , 1538, 27–38. Knoblauch, R.L. and Raymond, P.D., 2000, The effect of crosswalk markings on vehicle speeds in Maryland, Virginia and Arizona. (Washington, D.C., Federal Highway Administration). Lord, D., 1996, An analysis of pedestrian conflicts with left-turning traffic, paper presented at 75th Annual Meeting of the Transportation Research Board., Washington, D.C. McKnight, A.J., 1988, Driver and pedestrian training, In Transportation in an Aging Society , Vol. 2, Special Report #218, Transportation Research Board, Washington, D.C.: 101–133. Moskowitz, H. 2002. Alcohol and drugs. In R.E.Dewar and P.L.Olson (Eds.) Human Factors and Traffic Safety . (Tucson, AZ: Lawyers & Judges Publishing Co.), 177–207. Mulder, S. and Hutten, A. (2002). Injuries associated with inline skating in the European region. Accident Anal Prev. , 34, 65–70. National Highway Traffic Safety Administration, 1999, Traffic Safety Facts 1998: Pedestrians . Washington, D.C. O’Brien, A.P. and Brindle, R.E., 1999, Traffic calming applications. In J.Pline (Ed.) Traffic Engineering Handbook , 5th ed. (Washington, D.C.: Institute of Transportation Engineers), 219–256. Olson, P.L., 2002a, Driver perception-response time. In R.E.Dewar and P.L.Olson (Eds.) Human Factors and Traffic Safety . (Tucson, AZ: Lawyers & Judges Publishing Co.), 43–76. Olson, P.L., 2002b, Visibility with motor vehicle headlamps. In R.E.Dewar and P.L.Olson (Eds.) Human Factors and Traffic Safety . (Tucson, AZ: Lawyers & Judges Publishing Co.), 341–379. Olson, P.L. and Farber, E., 2003, Forensic Aspects of Driver Perception and Response , 2nd ed. (Tucson, AZ: Lawyers & Judges Publishing Co). Olson, P.L. and Sivak, M., 1984, Visibility problems in nighttime driving, In G.A.Peters and B.J.Peters (Eds.), Automotive Engineering and Litigation , (New York: Garland Law Publishing), pp. 383–405.


485

Oxley, J., 2000, Age differences in road crossing behavior. Unpublished Ph.D. Thesis, Department of Psychology, Monash University, Melbourne, Australia. Oxley, J., Fildes, B. Ihsen, E, Charlton, J., and Day, R., 1997, Differences in traffic judgments between young and old pedestrians. Accident Anal Prev. , 29, 839–847. Perry, J., 1992, Gait Analysis (New York: McGraw Hill). Preusser, W, Leaf, K., DeBartolo, R., Blomberg, R.D., and Levy, M., 1984, The effect of rightturn-on-red on pedestrian and bicyclist accidents, J. Saf. Res. , 13, 45–55. Preusser, D.F., Wells, J.K., Williams, A.F., and Weinstein, H.B., 2000, Pedestrian crashes in Washington, D.C. and Baltimore. Accident Anal. Prev. , 34, 703–710. Reiss, M., 1977, Knowledge and perceptions of young pedestrians, Paper presented at the 56th Annual Meeting of the Transportation Research Board., Washington, D.C., January. Salvatore, S., 1974, The ability of elementary and secondary school children to sense oncoming vehicle velocity. J. Saf. Res ., 6(3), 118–125. Sandels, S., 1968, Children in Traffic , (London: Paul Elek). Schieber, R.A., Branche-Dorsey, C.M., and Ryan, G.W., 1994, Comparison of in-line skating injuries with roller skating and skateboard injuries. JAMA , 271(23), 1856–1858. Sheppard, D. and Pattinson, M., 1988, Interviews with elderly pedestrians involved in road accidents. Transportation and Road Research Laboratory Research Report #98, (Crowthorne, U.K.). Shieber, R.A., Branche-Dorsey, R.A., Ryan, C.M., Rutherford, G.W., Stevens, J.A., and O’Neil, J., 1996, Risk factors for injuries from in-line skating and the effectiveness of safety gear, New Engl. J. Med. , 335(22), 1630–1635. Shinar, D., 1984, Actual versus estimated nighttime pedestrian visibility, Ergonomics , 27, 863– 871. Snyder, M.B., 1972, Traffic engineering for pedestrian safety: some new data and solutions, Highway Res. Rec. , 406, 21–27. Stemley, J.J., 1998, One-way streets provide superior safety and convenience. ITE J. , 68(8), 47– 50. Stutts, J.C., Hunter, W.H., and Pein, WE., 1996, Pedestrian-vehicle crash types: an update. Transp. Res. Rec. 1538, 68–74. Tidwell, J.E. and Doyle, D., 1995, Driver and pedestrian comprehension of pedestrian law and traffic control devices. Transp. Res. Rec. , 1502, 119–128. van der Molen, H., 1981, Child pedestrian exposure, accidents and behavior. Accident Anal. Prev. , 13, 193–224. Virkler, M.R., 1998, Pedestrian delay effects on signal delay. Transp. Res. Rec. 1638, 88–91. Zeedyk, S.M., Wallace, L., and Spry, L., 2002, Stop, look and think? What children really do when crossing the road. Accident Anal. Prev. , 34, 43–50. Zegeer, C.V., 1983, Feasibility of Roadway Counter-measures for Pedestrian Accident Experience, Pedestrian Impact Injury and Assessment, P-121, (Warrendale, PA: Society of Automotive Engineers), 39–49. Zegeer, C.V., Stewart, J.R., Huang, H., and Langerwey, P. (2001). Safety effects of marked vs. unmarked crosswalks at uncontrolled locations. Transp. Res. Rec. , 1773, 56–68.

Further Information Additional helpful information is available at the following web sites: http://www.walkinginfo.org/insight/features_articles/userguide.htm http://www.walkinginfo.org/rd/devices.htmtfcrosl http://www.walkinginfo.org/rd/for_ped.htm#sidewalk http://www.walkinginfo.org/rd/index.htm

17 Commercial Motor Vehicle Collisions Dennis Wylie D.Wylie Associates 0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

Commercial motor vehicles (mostly trucks and busses, in North American terminology) are typically larger, heavier, and more complicated to operate than cars and other personal vehicles. Commercial drivers are subject to the same human errors that car drivers are but typically drive more, and many drive at night and on irregular schedules that sometimes result in fatigue and impaired performance. Commercial motor vehicle operations are held to a higher standard of care because they involve more extensive damage potential than personal vehicles do. Litigations arising from commercial motor vehicle collisions typically involve larger damage claims and a larger set of questions than car crashes, often leading to requirements for detailed forensic study by human factors experts familiar with the specialized aspects of commercial motor vehicle collision analysis. 17.1 Introduction to Human Factors Principles and Relevance to Standard of Care in Commercial Motor Vehicle Operations Human factors/ergonomics is a broad, long-established, internationally recognized scientific discipline and systems design profession. Human factors and ergonomics professionals improve safety, comfort, and productivity by developing better operator selection criteria, licensing systems, training, operating procedures, and equipment designs. They do this by taking into account detailed scientific knowledge of human biomechanical, physiological, psychological, and anthropometrical characteristics. This section introduces some of the human factors principles involved in commercial motor vehicle (CMV) operations and relates them to the standard of care expected of drivers and motor carriers. Each of these principles will be expanded in following sections. Skill and Knowledge Requirements Achieving appropriate standard of care requires that a driver have adequate skill and knowledge to operate the CMV that he intends to drive safely. Minimum skill and knowledge requirements for operating various classes of CMVs have been studied and quantified by human factors specialists, and the capabilities of drivers can be measured to


488

determine if they meet required levels. The forensic human factors specialist may be called upon to evaluate a CMV driver’s qualifications. Personnel Selection and Training Standard human factors principles of personnel selection and training apply to the selection and training of drivers by motor carriers. In addition to the intrinsic motivation of motor carriers to achieve safety and efficiency, government regulations specifically assign duties to motor carriers regarding the selection and training of drivers. Government regulations also require specific licensing procedures for CMV operators to ensure that they meet minimum skill and knowledge requirements. A forensic human factors specialist may be asked to evaluate a motor carrier’s personnel selection and training policies, in general, and specifically with regard to a driver involved in a CMV crash. Attention, Perception, Decision Making, and Motor Response (Muscular Reaction) Driver performance in each of these areas is critical to safe operation. Drivers and motor carriers have a duty to ensure that a driver does not operate a CMV if he is impaired by fatigue, illness, or any other factor. Failures in one or more of these functional areas may be proximate causes of a CMV crash. Identifying such causes is often a central issue in forensic human factors analysis of CMV crashes. Fatigue and Circadian Rhythms Many CMV operators work long hours, work through the night, and/or have irregular work-rest schedules. Consequently, fatigue and circadian rhythms are important factors in CMV operations because they can impair attention, perception, decision making, and motor response. Satisfactory standard of care requires that motor carriers and CMV operators understand the causes and effects of fatigue and circadian rhythms, and employ effective countermeasures, chiefly suitable work-rest scheduling (Mackie and Wylie, 1991). Government hours-of-service regulations must be observed. The forensic human factors specialist may be required to analyze the work-rest history of a driver regarding compliance with regulations and for likely effects on the driver’s capabilities in connection with a specific CMV crash. 17.2 Objective and Scope of the Chapter The objective of this chapter is to provide an overview of some important aspects of forensic human factors analysis of commercial motor vehicle collisions. It is intended to be useful to human factors/ ergonomics and safety professionals, defense and plaintiffs’ lawyers, judges, forensic and paralegal professionals, accident investigators, students, government officials, consultants, motor carriers, and commercial drivers. The scope is

Commercial motor vehicle collisions

489

intended to be broad; therefore, the amount of detail is necessarily limited. However, numerous references are cited that provide additional detail. 17.3 Foundations Characteristics of Commercial Motor Vehicles Related to Operator Requirements Various classes of commercial vehicles pose varying skill and knowledge requirements, all generally exceeding requirements for operating ordinary automobiles. The classified licensing system used in the United States (the commercial driver’s license system) is based on knowledge and skills tests suitable to the type of vehicle that the driver operates or will operate. Human factors specialists conducted research to identify vehicle groups that had substantially different skill and knowledge requirements (Mackie et al., 1989). These groups form the basis of the CDL licensing system and are described in 49 CFR §383.91. They are summarized below. Commercial Motor Vehicle Groups Group A vehicles are combination vehicles comprising a large truck towing a large trailer or a truck tractor towing a large semitrailer. A truck tractor is a vehicle that is not directly loaded with cargo; instead, the front of a semitrailer (a trailer with a rear axle but no front axle) is placed on the rear of the truck tractor, which bears half the weight of the semitrailer and provides motive power for the tractor-trailer combination. Group A vehicles have the greatest skill and knowledge requirements, and holders of the Group A license can operate Group B and Group C vehicles as well. Group B vehicles are large, straight trucks and buses. (These are referred to as “straight” vehicles because they do not “bend” in the middle, as Group A combination vehicles do.) A driver with a Group B license may pull a small trailer behind the straight truck or bus and may operate Group C vehicles. Group C commercial motor vehicles are vehicles that do not belong to Groups A or B, but are designed to transport 16 or more passengers or are placarded for hazardous materials. Operational Requirements The three commercial motor vehicle groups were identified because they differ substantially in the skill and knowledge necessary to meet most of the 15 operational requirements that the investigators found to be inherent in commercial motor vehicle operation (Mackie et al., 1989). These operational requirements are: • Determining vehicle and safety system conditions • Basic maneuvering skills including eye-hand-foot coordination and spatial visualization • Coupling procedures • Visual search requirements • Communication requirements


490

• Speed control • Space management • Special handling characteristics (e.g., off-tracking, rearward amplification, high center of gravity, surge) • Hazard perception • Night driving requirements • Driving in adverse weather (rain, snow, ice, wind) • Driving in hot weather • Driving in mountainous terrain • Emergency maneuvers and skid control • Cargo-related factors (weight distribution, securement, stability, hazardous materials, passengers) Capabilities and Limitations of Humans as Motor Vehicle Operators Human factors/ergonomics and safety professionals have studied the capabilities and limitations of humans as operators of most kinds of motor vehicles. Human factors considerations play a part in the forensic evaluation of accidents involving any kind of motor vehicle; these include: • Conspicuity • Driver behaviors (e.g., risk taking, car following, gap acceptance, overtaking) • Driver expectation • Fatigue • Impairment • Lighting • Medical condition • Perception-reaction time • Recognition and attention errors • Road characteristics • Signs and signals • Traffic conditions • Viewing sight distances and visibility Because these considerations are the subject of the chapter on traffic collisions, it is recommended that the reader review that chapter. Commercial Driver Licensing The Commercial Driver’s License (CDL) The U.S. standard of licensing requires skill and knowledge testing appropriate to the specific class of commercial vehicle to be operated. The U.S. Department of Transportation (DOT) was directed to develop this modern, uniform, scientifically valid system by the Commercial Motor Vehicle Safety Act of 1986. The system was developed by human factors specialists, with the aid of subject matter experts (federal and state


491

regulators, motor carriers, and drivers) (Mackie et al, 1989). The CDL system is defined in the Federal Motor Carrier Safety Regulations at 49 CFR Part 383. The Model Driver’s Manual The Model Driver’s Manual (Wylie and Schultz, 1989) was developed as the knowledge base for CDL testing. The manual reflects, as completely as practical, the full knowledge domain associated with safe operation of commercial motor vehicles. The sections of the manual, like the written knowledge tests, were composed so that they are tailored to the applicant’s particular class of vehicle and, in some cases, the type of cargo to be transported. The major sections of the manual include: • Introduction: brief descriptions of all CDL tests, their purpose, and how the tests that an applicant must take depend on the type of vehicle and type of cargo transported. • Driving safely: a compendium of topics including vehicle inspection, basic control, speed management, space management, night driving, adverse weather driving, mountain driving, emergen-cies, skid control, staying fit to drive, etc. This part of the manual should be read by all applicants. The general knowledge test is based on this material. • Transporting cargo safely: this part of the manual should also be read by all applicants and is covered by the general knowledge test. • Transporting passengers safely: this part should be read by all applicants who will require a passenger endorsement to their CDL. The passenger transport test is based on this material. • Air brakes: this part is to be read by all applicants whose vehicles have air brakes. The air brakes test is based on this material. • Combination vehicles: this part should be read by all applicants who wish to drive a Class A combination vehicle and by those who wish an endorsement to operate with double or triple trailers. The combination vehicles test and the doubles/triples test are based on this material. • Hazardous materials: this part should be read by all applicants who wish a hazardous materials endorsement to their CDL. The hazardous materials test is based on the information in this section. The Model Driver’s Manual defines minimum standard of care for operating commercial motor vehicles in the U.S. It has been adopted, published, and made available to the public by every state. Most states include information concerning state-specific commercial vehicle regulations as well. CDL Knowledge Tests A battery of CDL knowledge tests was developed that reflects the general knowledge required of all commercial drivers and specialized knowledge required of operators of particular classes of vehicles or vehicles hauling particular kinds of cargo. The knowledge tests to be taken by a CDL applicant directly reflect the type of vehicle that he operates or proposes to operate. They include: • General knowledge test of safe driving principles (taken by all applicants)


492

• Air brakes test (taken by all applicants with a vehicle having air brakes) • Combination vehicles test (taken by all applicants wishing to drive large combination vehicles) • Tanker test (taken by all applicants who wish an endorsement to their CDL that permits them to drive tank vehicles) • Doubles/triples test (taken by all applicants who wish an endorsement to their CDL permitting them to pull double or triple trailers) • Passenger transport test (taken by all applicants who require an endorsement to their CDL to carry passengers) • Hazardous materials test (taken by all applicants who require a hazardous materials endorsement to their CDL) Two roughly equivalent forms of each of these tests were developed and evaluated. Each meets professional standards for test reliability, content validity, and comprehensiveness in its coverage of the knowledge domain. Details are given by Mackie and colleagues (1989). CDL Skill Tests The CDL licensing system employs three types of driver skill tests: • Vehicle inspection test, to determine whether the driver has an adequate understanding of how to ascertain the condition of key operational and safety systems of the vehicle • Basic control skills test, to determine whether the driver has the fundamental psychomotor and perceptual skills necessary to control and maneuver heavy vehicles • Road test, to determine whether the driver is capable of safely driving the vehicle in a variety of road environments and traffic conditions These tests were designed to be adaptable to different vehicle sizes and configurations. Each meets professional standards for reliability and validity and each measures an important, yet relatively independent, area of required driver skill. Details are given by Mackie and associates (1989). Driver Fatigue and Government Hours-of-Service Regulations Background In the late 1930s, concern about truck driver fatigue led to development of the first hoursof-service regulations of U.S. commercial drivers. These regulations have remained largely unchanged to the present time. For drivers in interstate commerce, they permit no driving after 10 hours of cumulative driving and/or 15 hours of cumulative on-duty time, until an 8-consecutive-hour off-duty break is taken. (A common misconception is that driving is limited to 10 hours in a 24-hour period; in fact, it is limited to 10 hours in an 18-hour period. Therefore, driving up to 16 hours in a 24-hour period is permitted. The 18-hour cycle results in a backward-rotating work-rest schedule.) Furthermore, the hours-of-service regulations permit no driving after 60 hours cumulative on-duty time in any 7-day period, or 70 hours cumulative on duty-time in any


493

8-day period (49 CFR §395). An excellent and well-documented review of the history of these regulations and their strengths and weaknesses is given in the Federal Motor Carrier Safety Administration’s “Notice of Proposed Rulemaking for Driver’s Hours of Service” (FMCSA, 2000). Driver Fatigue Remains a Problem Hours-of-service regulations have not eliminated the problem of fatigue. A survey of 511 long-haul tractor-trailer drivers around the U.S. revealed that 57% had been drowsy one or more times at the wheel in the preceding month, and that 28% had dozed or fallen asleep at the wheel one or more times in the preceding month (Abrams et al., 1997). Of those surveyed, 58% said they “sometimes” or “always” get less sleep to keep up with their schedules. The U.S. Department of Transportation held a Truck and Bus Safety Summit that brought together more than 200 people, representing the many organizations involved in motor carrier operations, regulation, research, and safety, to prioritize the safety issues facing the industry. After 2.5 days, the representatives completed their discussions and voted to determine the top issues affecting the safety of motor carriers. The top-priority issue was driver fatigue (U.S. DOT, 1995). Driver fatigue has continued to be a principal concern in highway safety to this day. The Federal Motor Carrier Safety Administration is currently drafting new hours-of-service regulations in light of modern scientific knowledge. Fatigue impairs vigilance, attention, perception, decision making, and muscular reaction—processes that are crucial to safe driving. These objective, measurable fatigue effects are primarily cognitive, or mental, and can occur without any marked degree of prior physical exertion. Driving fatigue is caused by long and/or irregular hours of work and too few hours of sleep or poor-quality sleep. Sleep debt is defined as the difference between one’s sleep requirement (7 to 8 hours daily for most people) and the amount of sleep actually received. In their literature review, Wylie et al. (1996) reported that sleep debt of as little as a few hours can adversely affect vigilance and that commercial drivers, especially those on irregular schedules, can build up such debts in 2 days. The Expert Panel on Driver Fatigue and Sleepiness made statements applicable to this issue: Habitually restricting sleep by 1 or 2 hours a night can lead to chronic sleepiness…. Sleepiness causes auto crashes because it impairs performance and can ultimately lead to the inability to resist falling asleep at the wheel…. Factors recognized as increasing the risk of drowsy driving and related crashes include: Sleep loss Driving patterns, including driving between midnight and 6 a.m.; driving a substantial number of miles each year and/or a substantial number of hours each day;…driving for longer times without taking a break. (NHTSA/NCSDR, 1998)


494

In addition, fatigue, loss of alertness, and drowsiness are strongly affected by circadian rhythms (the human body’s “clock”) and other time-of-day effects. Lack of Sleep and Time-of-Day Effects The importance of driver fatigue led the U.S. Department of Transportation and Transport Canada to commission their largest, most comprehensive over-the-road study of driver fatigue and alertness. My associates and I designed, executed, and documented a study involving 80 U.S. and Canadian tractor-trailer drivers in an operational setting of real-life, cargo-hauling trips totaling more than 200,000 miles and 4000 hours of driving. We monitored the drivers and trucks continuously by electronic instrumentation. We found that the drivers obtained less sleep than is required for alertness on the job and that the strongest and most consistent source of variation in driver fatigue and alertness was time of day, due in part to the effects of 24-hour biological rhythms, known as circadian rhythms (Wylie et al., 1996; Mitler et al., 1997). Drowsiness, as judged from video images of drivers’ faces, was markedly greater during night driving than during daytime driving. Peak drowsiness occurred during the 8 hours from late evening until dawn. Circadian rhythms are biological rhythms that repeat approximately every 24 hours. An internal clock, or pacemaker, located in the brain generates circadian rhythms. These rhythms promote sleep during the night and alertness in the day. When sleep periods must be taken in the daytime, circadian rhythms act to reduce quality and quantity of sleep. These rhythms have an adverse influence on the performance of tasks that must be done between midnight and dawn because of the dual effects of depressing performance at night and interfering with sleep in the day. Nighttime drivers are especially susceptible because of the stimulus deprivation associated with driving during those hours. The visual field is restricted mostly to a narrow cone of headlight illumination, traffic is at a minimum, and the likelihood of boredom and monotony, long known to be factors that hinder vigilance (Wylie et al., 1985), is maximized. Accident data also reveal these time-of-day effects. For example, the Notice of Proposed Rulemaking for Driver’s Hours of Service (FMCSA, 2000) shows in Chart 3 the relative risk of truck driver fatigue by time of day, derived by University of Michigan Transportation Research Institute scientists. These relative risk estimates are based on accident data from 1991 to 1996. The shape of the distribution of relative risk over the hours of the day is remarkably similar to the distribution by time of day of the 1,989 video samples judged to show a drowsy truck driver in the Driver Fatigue and Alertness Study (Wylie et al., 1996). The two curves correlate well, r=0.88, reinforcing our conclusions regarding the powerful influence of time of day. The daily variation of drowsiness and fatigue risk is illustrated in Figure 17.1 by a mathematical model indicated by the solid line (Wylie, 2000) and drowsiness data from actual truck drivers indicated by the dotted line (Wylie et al., 1996). It can be seen that time-of-day effects lead to greatly increased incidence of drowsiness by midnight.


495

Cumulative Fatigue and Irregular Schedules In 1990, the National Transportation Safety Board completed a study of 182 heavy-truck accidents that were fatal to the driver (NTSB, 1990). The primary purpose in investigating fatal-to-the-driver heavy-truck accidents was to assess the role of alcohol and other drugs in these accidents. The study found, however, that the most frequently cited probable cause was fatigue. The NTSB stated that it believed that the 31% incidence of fatigue in fatal-to-the-truckdriver accidents found in the 1990 study represented a valid estimate of the portion of fatal-to-the-driver heavy truck accidents that are fatigue-related. Therefore, the NTSB initiated another study—of drivers who survived truck crashes— to examine the role of specific factors, such as drivers’ patterns of duty and sleep, in fatigue-related heavy truck accidents. They found that irregular duty and sleep patterns are a significant risk factor. About 67% of the drivers with irregular patterns had fatiguerelated accidents (43 of 64), whereas about 38% of drivers with regular patterns had fatigue-related accidents (9 of 24) (NTSB, 1995). A driver’s duty hours were classified as irregular if the start times of two consecutive duty periods varied by 2 or more hours at least twice during the previous 96 hours (4 days). A driver’s sleep hours were classified as irregular if the start times of two consecutive rest periods varied by 2 or more hours at least twice during the previous 96 hours.

FIGURE 17.1 Daily variation of drowsiness and fatigue risk. Regarding irregular schedules, the Expert Panel on Hours of Service Recommendations appointed by the U.S. Department of Transportation (Transportation Research Institute, 1998) stated what it thought was the primary critical issue of hours-of-service rules for commercial drivers:


496

1. 24-Hour Cycle—For most of the history of humankind, the natural light cycle governed activity. Although firelight has long been available, it was not until the advent of lamps and eventually electric lighting that daylight was significantly extended. This change did not occur until very late in the evolution of man, and, from a biological and physiological standpoint, we are very much the same as we were tens of thousands of years ago. Basically, we are on a 24- to 25-hour cycle, and, independent of light cycles, our bodies still appear to respond accordingly. Therefore, whatever hours-of-service regulations are adopted should conform to this cycle that is apparently “normal” from an inherent constitutional standpoint. This biological predisposition cannot be changed by bureaucratic regulation or even economic necessity (Mitler et al., 1988; Wylie et al, 1996; Folkard, 1997) (p. 5). Irregular work-rest schedules cause an accumulation of sleep loss and fatigue as days and nights go by. This point was addressed by the Expert Panel (Transportation Research Institute, 1998): Sleep loss, extended hours of service, and “unnatural” schedules (e.g., night driving) cannot be compensated for by a single night’s rest. Sleep loss is substantial. The FHWA-Transport Canada study that recorded sleep of 80 drivers over a week of driving found average sleep lengths of just under 4 hours to 5.5 hours (Wylie et al., 1996).This is substantially less than the normal length of 7.5 hours to 8 hours (Gallup Organization, 1995), but typical of industries where 12-hour shifts are worked (Smiley and Heslegrave, 1997). Such partial sleep deprivation results in cumulative sleep debt that requires extended time to recover. Recovery time periods must take into consideration the necessity for overcoming cumulative fatigue resulting from such schedules and must include sufficient nighttime sleep. Compared to daytime sleep, nighttime sleep is more efficient, that is, a higher proportion of the time spent in bed is actually spent in sleep (Lavie, 1989). When sleep time occurs during the day, fewer hours of sleep occur (Akerstedt, 1995) (p. 12). Adherence to the current U.S. hours-of-service regulations does not prevent cumulative fatigue for drivers with irregular schedules. The Expert Panel addressed the current policy: Overall, the current policy, Policy Option A, fails on every criterion identified by the Panel as important for a reasonable hours-of-service policy. The 18-hour cycle ignores real physiological factors and violates everything that is known about the powerful effects of circadian rhythms (Mackie and Miller, 1978; Mitler et al, 1988; Home and Reyner, 1995; Gillberg et al., 1995; Wylie et al., 1996; Folkard, 1997). Furthermore, allowing only 8 hours off-duty for sleep, with no additional time for other activities, translates into continuous sleep deprivation that accumulates


497

over time (Wylie et al., 1996; Mitler et al., 1997; Smiley and Heslegrave, 1997). The combination of sleep loss and driving during circadian troughs is synergistic, exacerbating the effects of both (Lavie, 1986; Mitler et al., 1988; Rosekind et al., 1997). Furthermore, over a 24-hour cycle, as much as 16 hours on-duty time is allowed, all of which may be spent driving. Such a schedule inevitably results in fatigue and impaired performance (p. 19). Fatigue Can Be Equivalent to Alcohol Intoxication Williamson and Feyer (2000) compared the relative effects of fatigue and alcohol on performance. Performance of 39 transport industry people on more than a half-dozen tests related to safe driving (e.g., reaction time, vigilance) was measured on two occasions. On one occasion, the performance tests were given periodically over many hours as the people consumed given quantities of alcohol. Breath analyzer measurements were used to confirm their blood-alcohol concentration (BAC). On another occasion, the same people were given the same performance tests repeatedly during a long period of being awake starting at about 6:00 a.m. and lasting well past midnight. The test results were then analyzed to find out how many hours of wakefulness were equivalent to various BAC levels. These researchers found that “at a BAC of 0.10%, equivalence occurred after between 17.74 and 19.65 hours of wakefulness which falls in the late evening to early hours of the morning, corresponding in this study to between 2328 [11:28 p.m.] and 0123 [1:23 a.m.]” (Williamson and Feyer, p. 653). This level of BAC is more than twice the 0.04% U.S. federal limit for commercial drivers (49 CFR §382.201). The legal limit was set because the impairment of alcohol intoxication at levels higher than 0.04% BAC is associated with unacceptable probability of injury and death (Transportation Research Board, 1987). 17.4 Human Factors Forensic Analysis of CMV Collisions The goals of the human factors forensic expert are to gain an understanding of the collision from a technical human factors point of view and then to explain it in terms understandable to attorneys, judge, and jury. Pursuing this goal usually requires study of the actions and inactions of all involved motorists, pedestrians, etc., as well as the physical setting of the accident. It can be expected, however, that major emphasis will be placed on the particulars of the CMV driver’s performance. This will involve forensic analysis of the driving environment and the driver’s apparent actions in the proximate time period leading up to and including the accident, as related by witnesses, indicated by analyses of physical evidence, and inferred from reconstructions based on engineering and physics. It is likely that analyses and reconstructions by experts from other fields will be considered, in addition to any performed by the human factors expert. It is the latter, however, who must characterize the driver’s performance with regard to numerous specific human factors elements of driving. If the expert concludes that CMV driver error was a cause of the accident, an important question will then be whether driver fatigue might have been responsible.


498

There is no breath analyzer or blood test that can reveal the presence of a critical level of fatigue in an accident-involved driver. However, a forensic driver-fatigue expert can often conclude to a reasonable degree of scientific certainty whether fatigue was a cause of an accident by conducting a human factors analysis of the driver’s work-rest history leading up to the accident. Ideally, a commercial driver’s work-rest history would be adequately documented in the driver’s record of duty status required by the Federal Motor Carrier Safety Regulations (49 CFR §395.8), corroborated by regular work hours and personal routine. Unfortunately, this is not always the case. The driver’s work hours may be very irregular; his log may be days in arrears or may be missing altogether. Even if present, the log may be inaccurate or falsified, so it must be tested against other sources. The forensic driver-fatigue expert examines many records and documents, uses specific tools and techniques to reconstruct the driver’s probable work-rest history, and describes the implications of that history in terms of causal factors of fatigue. If the expert judges that the CMV driver’s actions or inactions were a cause of the accident, the expert may also be requested to analyze whether the motor carrier exercised proper standard of care in performing its duties of selecting, qualifying, training, motivating, disciplining, dispatching, and monitoring drivers in the interest of safety. These forensic human factors analyses often permit the expert to identify any specific deficiencies of CMV driver performance and to conclude to a reasonable degree of scientific certainty whether fatigue was or was not a cause of the accident. The expert may also opine as to the motor carrier’s standard of care. These findings and expert opinions must then be organized and communicated effectively. Driver-error analysis, work-rest history analysis, motor carrier standard of care, and effective communication of opinions will be considered in turn in the following subsections. Analysis of Driver Errors Proximate to the Accident Analysis of driver errors proximate to the accident requires study and assimilation of many different sources of information. Some of the sources that may be available include the police accident report; police accident reconstruction; statements taken by police from witnesses at the scene of the accident; photographs of the scene taken by the police, insurance adjusters, and others; accident reconstruction experts’ reports; and deposition transcripts. The depositions of the CMV driver, CMV passengers, people in other vehicles, and any other witnesses must be studied carefully. Similarly, answers to interrogatories and responses to requests for document production should be examined carefully for their relevance to the human factors aspects of the case. The expert should also review the complaint and other pleadings in the case to gain a better understanding of who is suing whom and what is being alleged. This is likely to give the expert a better understanding of the case, which helps him to be a more effective expert witness. If the expert enters the legal process late in its course, much of this information will already be available, and he will be faced with an intensive reading assignment. If the expert is retained earlier in the matter, the workload will be more spread out, and the expert can consult with and advise his attorney client with respect to important human factors questions that might be posed by plaintiffs’ and defense attorneys.


499

In most cases, the expert should visit the accident site to view it directly; to record salient human factors features through photographic, video, or other means; and to take appropriate measurements as needed, such as luminance levels and sound pressure levels. It is difficult to provide a formula or recipe for how all this information should be put together to form an expert opinion. One or more human errors can contribute to an accident in many ways. Some categories of human performance and error have been mentioned in this chapter, and the reader can review additional details regarding traffic collisions in the chapter on traffic collisions. Ultimately, however, proper synthesis requires the education, training, experience, and inquiring mind of a qualified human factors forensic expert. Analysis of Commercial Driver Work-Rest History When there is the question of driver fatigue, an analysis of the driver’s work-rest history is necessary. In the U.S., the Code of Federal Regulations (49 CFR §395) requires drivers engaging in interstate commerce to follow specific rules in making the prescribed driver’s daily records of duty status, commonly called “logs.” (A driver need not cross state lines to be involved in interstate commerce; if the driver transports cargo that has crossed state lines, he is engaged in interstate commerce.) Most U.S. states have similar requirements. The principal reason for these logging requirements is to assist drivers and motor carriers in complying with federal and state hours of service regulations and to permit law-enforcement officials to verify compliance. By reviewing drivers’ logs, the forensic expert can evaluate compliance with the U.S. 10-hour driving limit, 15-hour on-duty limit, and the 70/80 hour cumulative on-duty rules (or corresponding state or other jurisdictional hours-of-service regulations). It is very desirable to have at least 8 days of consecutive logs ending on the day of the accident. It is desirable to have logs for prior weeks as well because these can help to evaluate work-rest patterns and, in some cases, to evaluate place names and travel times between geographic locations that have been visited repeatedly. (Deciphering handwritten place names is no trivial matter in evaluating logbooks; a driver is often asked at deposition to clarify the place names in logbooks, which is very helpful in analyzing the logged data.) The expert should analyze available logbooks to determine compliance with applicable hours-of-service rules and to gain an understanding of the work-rest cycles recorded by the driver during the days leading up to the accident. The expert must be thoroughly familiar with the applicable logging rules and the hours-of-service rules (e.g., for U.S. interstate transportation, 49 CFR §395 “Hours of Service of Drivers,” plus the U.S. DOT “Interpretations” published in the Federal Register on April 4, 1997, plus the subsequent DOT interpretations). The expert must be diligent in doing the required arithmetic; it is sometimes helpful to use logbook audit software (for example, that available from J.J. Keller and Associates) to double-check this work. Analysis of logbooks is usually not sufficient, however. The driver’s logs may be missing important data and may be inaccurate with respect to the places and/or the times of day of changes in duty status, by reason of simple human error, inattention to detail, or willful falsification intended to conceal violations of hours of service rules. (Many drivers have an economic incentive to drive beyond the limits because they get paid by the mile; motor carriers obviously have a similar economic incentive, and drivers and


500

motor carriers are often under pressure from shippers and receivers, particularly with the prevalence of current “just-in-time delivery” inventory practices (U.S. DOT, 1995).) Therefore, the forensic expert must aim to reconstruct the driver’s trip itinerary and work-rest schedule to evaluate fatigue and compliance with the hours of service regulations. The reconstructed itinerary should take into account the data in the driver’s logbooks, as well as supporting documents. The U.S. DOT defines in its interpretations that “supporting documents” are the records of the motor carrier maintained in the ordinary course of business and used by the motor carrier to verify the information recorded on the driver’s record of duty status in accordance with 49 CFR 395.8(k)(1), which also requires that they be retained for a period of 6 months after their receipt. Examples are: Bills of lading Carrier pros Freight bills Dispatch records Driver call-in records Gate record receipts Weight/scale tickets Fuel receipts Fuel billing statements Toll receipts International registration plan receipts International fuel tax agreement receipts Trip permits Port-of-entry receipts Cash advance receipts Delivery receipts Lumper receipts Interchange and inspection reports Lessor settlement sheets Over/short and damage reports Agricultural inspection reports CVSA reports Accident reports Telephone billing statements Credit card receipts Driver fax reports On-board computer reports Border-crossing reports Custom declarations Traffic citations Overweight/oversize reports and citations Other documents directly related to the motor carrier’s operation that are retained by the motor carrier in connection with the operation of its transportation business A good approach to assimilating the various available bits of information regarding geographic locations at particular times is to use trip-planning software, such as


501

Microsoft Streets and Trips or PC Miler. A first-estimate itinerary can be obtained by entering the geographic place names into the application program in assumed chronological order, assigning durations for periods logged as “not driving,” and assigning assumed average speeds for various types of roads, according to the capabilities of the particular software. (If the software will not account for variations in speed limit by state, the analysis may be done in segments, wherein each segment comprises a state or states with identical speed limits.) The software then calculates an itinerary, and that itinerary is compared to the driver’s logs and the supporting data. The calculated itinerary may be found to be a good fit to the driver’s logbook or it may be very different. In the latter case, the expert must identify likely reasons for discrepancies between the logs and the calculated itinerary; make adjustments to the assumed times, geographic locations, and average speeds; and calculate a new itinerary. In what is typically an iterative process, the expert (hopefully) converges on a calculated itinerary that reasonably explains the most credible bits of data available concerning the work-rest schedule of the driver. The “best fit” itinerary can then be analyzed with respect to hours-of-service compliance and likely fatigue effects. It can also be used to judge probable speed limit violations. (The U.S. DOT, in an interpretation of 49 CFR §392.6, stated that average speed greater than 5 mph below the speed limit over a many-hour period is considered incapable of being made in compliance with the speed limit and hours of service limitation.) The fatigue evaluation must take into account the time of day of the accident, the duration of the preceding period of driving and of wakefulness, the duration and time of day of the last principal sleep, and so on, backward in time, taking account of the duration, time of day, and regularity/ irregularity of principal rest and on-duty periods. As in the case of evaluating driver errors, it is difficult to provide a formula or recipe for how all this information should be put together to form an expert opinion. Ultimately, proper synthesis requires the education, training, experience, and inquiring mind of a human factors forensic expert. Analysis of Motor Carrier Standard of Care U.S. motor carriers are held responsible by the Federal Motor Carrier Safety Regulations for selecting, qualifying, training, motivating, disciplining, dispatching, and monitoring drivers in the interest of safety. Motor carriers must ensure that a driver is qualified to drive a commercial motor vehicle, has sufficient hours of service available, and is not ill or fatigued. The regulations state that whenever a duty is prescribed for or a prohibition imposed upon the driver, it is the duty of the motor carrier to require observance of such duty or prohibition (49 CFR §390.11). “Motor carrier” is defined by U.S. federal regulations as a for-hire motor carrier or a private motor carrier. This term is broadly defined to include a motor carrier’s agents, officers, and representatives as well as employees responsible for hiring, supervising, training, assigning, or dispatching drivers, and employees concerned with the installation, inspection, and maintenance of motor vehicle equipment and/or accessories (49 CFR §390.5). Motor carriers must perform initial qualification of drivers and follow-up qualification. Initial qualification starts with a written application for employment that


502

identifies the motor carrier and the applicant; describes the applicant’s experience operating motor vehicles; lists motor vehicle accidents in the last 3 years; lists convictions for violations of motor vehicle laws in the last 3 years; contains statements regarding denial, revocation, or suspension of driver’s licenses; and lists previous employers during the last 3 years. The motor carrier must then check into the applicant’s driving record and employment record. Responses to these inquiries from state driverlicensing agencies and previous employers are to be kept in the driver’s qualification file. The motor carrier is required to administer a road test to drivers that will operate double/triple trailers or tank vehicles to determine if they can safely operate the motor vehicle. For other commercial vehicles, the regulations now permit, but do not require, the motor carrier to accept the commercial driver’s license as evidence of minimum skill. A certificate of road test issued by another motor carrier during the preceding 3 years is also acceptable. If a road test is given, a record of the road test form must be included in the driver’s qualification file. The applicant must have a currently valid DOT medical examiner’s certificate of physical examination, which must be carried by the driver while operating a commercial motor vehicle, with a copy kept in the driver qualification file. Follow-up qualification of a driver by the motor carrier includes assuring that the medical examiner’s certificate is renewed at least every 2 years. At least every 12 months, the motor carrier must check the driver’s driving record by requesting information from each state in which the driver held a license during the year. The motor carrier must evaluate the driving record, judge the driver’s suitability, and keep a written record of the review in the driver’s qualification file. The motor carrier must also require that each driver submit a list annually of all violations of traffic safety laws for which he was convicted. The motor carrier must evaluate this record with respect to the suitability of the driver, and the list must be kept in the driver’s qualification file. Motor carriers know (or should know) that in a legal action, they are likely to be asked to produce numerous documents and answer many questions of interest to human factors forensic analysts for plaintiff and defense. For example, with reference to the driver, the following documents and items may be produced: • Driver’s Record of Duty Status logs (49 CFR §395.8) for 1 month preceding the accident • Records of driver daily report-for-duty times, release-from-duty times, and total on-duty time for the 7 days preceding the accident in question, if driver was exempt from logbook requirements of 49 CFR §395.8 because of the 100 air-mile radius exemption per 49 CFR §395.1(e) • Records of disciplinary action • The truck driver’s qualification file (49 CFR §391.51), including • Employee’s application • List of truck driver’s previous employers for 10 years preceding the date of the application • Reasons for leaving these employments • Medical examiner’s certificate • A note showing when and who reviewed the driver’s record with him for each year of employment (49 CFR § 391.25)


503

• A list of certificates showing all violations of motor vehicle laws and ordinances (49 CFR § 391.27) • Responses from state agencies and employers to employer’s inquiries about the truck driver’s employment and driving records (49 CFR §391.23) • Certificate of road test (49 CFR §391.31(e)) • Records of drug and alcohol tests • An accident register listing all DOT recordable preventable accidents (Note: the employer is required to keep the driver’s file in its principal place of business as long as the driver is employed and for 3 years thereafter.) • Copy of all commercial driver’s licenses of the truck driver • All writings giving notification to the employer of the truck driver’s convictions or suspensions for violating a state or local law relating to motor vehicle traffic control (49 CFR §383.31) • All records of the driver’s alcohol tests with confirmed reading of 0.02% or greater, confirmed positive controlled substance test results, documentation and refusal to take alcohol and/or drug tests, instrument calibration documentation, driver evaluation by a substance abuse professional, and calendar year summaries for the last 5 years • All records of alcohol tests with less than 0.02% blood-alcohol reading and negative drug tests • All copies of alcohol test forms, controlled substance standard custody forms, documents related to the refusal of any driver to submit to testing, documents supplied by the driver to dispute test results, and signed acknowledgments of required training documents • Copies of educational materials explaining drug and alcohol testing regulations submitted to drivers • Copies of the employer’s policies and procedures relating to alcohol and drug testing • Copies of the driver’s signed receipt for the preceding materials • Copies of all company manuals governing commercial motor vehicle safety, maintenance, fleet safety programs, and driver’s standards (in lieu of presenting a copy, the responding party may present a list and date of publication of any and all manuals on hand) With regard to the defendant motor carrier’s policies and programs, the following documents and items may be requested: • Defendant motor carrier’s safety program, including studies and tests to determine the safety of the single, double, or triple trailer configurations used by the defendant motor carrier; actions taken to assure that drivers are not violating the federal regulations relating to the maximum hours of work by drivers and that accurate logs are submitted by the drivers and accepted by the defendant motor carrier for filing; as well as the continued testing of drivers for competency and substance abuse • Driver standards • Dispatcher standards • Driver compensation and incentives • Driver disciplinary policy


504

The motor carrier may be asked the following: • State the policy with respect to the retention and destruction of drivers’ logs and trip receipts and explain any differences between that policy and the manner in which the defendant driver’s logs and trip expenses for the trip in question were treated. • What steps, if any, the company has taken to assure the accuracy of the logs submitted by its drivers as required by the Federal Motor Carrier Safety Regulations, 40 CFR §390.11 and 390.13. • State the policy with respect to testing drivers for substance abuse and explain any difference between that policy and the manner in which the defendant driver was treated. • Was postaccident testing for alcohol and controlled substances conducted in accordance with 49 CFR §382.303? If not, why not? • State the policy with respect to the operational speeds for the company’s commercial motor vehicles and explain exactly how compliance is enforced. • State the policy with respect to the use of CB radios by the company’s drivers and explain any difference between that policy and manner in which the defendant driver was treated. • State the policy with respect to the use of radar detectors by the company’s drivers and explain any differences between that policy and the manner in which the defendant driver was treated. The answers to these questions and the documents produced may permit the forensic expert to evaluate motor carrier standard of care. Organization and Communication of Findings and Expert Opinions The human factors forensic expert must express his or her findings and expert opinions plainly and in a well-substantiated manner. The expert should use theories and techniques that are tested, peer reviewed, published, and generally accepted within the human factors community. It is also important to identify the known or potential rate of error in the methods used. Various cases will involve varying degrees of challenge to the expert, based upon Daubert and other factors that give the trial judge a gatekeeper role in evaluating expert testimony (see the chapter on general testimony issues). Often, the expert’s client attorney may request his findings and opinions in the form of a letter report. The format of letter reports can vary considerably, but the contents likely would include: • Salutation/introduction • Materials reviewed (often included as an attachment if the list is long) • Short accident summary • Summary of the human factors analysis of proximate driver errors • Summary of the work—rest history analysis (if applicable) • Findings regarding motor carrier standard of care (if applicable) • Explicit statement of expert opinions • Complementary close • Reference list • Figures and tables


505

• Attachments such as maps and listings of “best fit” trip itineraries Sometimes, the expert’s client attorney may only want a simple outline or perhaps just an oral report. Regardless, the expert must be prepared to give clear, well-founded summaries of findings and opinions tailored for the average technical level of the intended audience: attorneys, judge, and jury. Charts and other graphical methods can be helpful. For example, in explaining an accident in which a commercial driver failed to detect a yellow traffic signal light immediately, decided it was too late to stop, and broadsided another motorist, a chart like that shown in Figure 17.2 might be helpful to explain the split-second timing involved and the alternative outcomes, based on whether the driver attempts to stop or not. A complex chart like this example should not be expected to stand on its own; the expert should use it as a visual aid for explanation of the underlying concepts. Such charts can be prepared in Excel and other widely used programs. Similarly, it might help to explain an irregular work-rest schedule to use a spreadsheet chart similar to that shown in Figure 17.3. 17.5 Examples Example 1 At 12:20 a.m., a commercial driver on cruise control drove his tractor-trailer at high speed into a well-marked and illuminated mile-long approach to a construction zone on an interstate highway, at the end of which a car was stopped by a flagman. The truck driver reacted only in the last fraction of a second before impact, resulting in fatal injuries to the driver of the car. The highway was straight, level, and dry, and the sky was clear, with good visibility. The human factors forensic expert retained by the plaintiff examined the driver’s logs and found that the driver had had a very irregular work and rest schedule in the 4 days prior to the accident, which he judged led to cumulative fatigue. The cumulative fatigue was made sharply worse during the day of the trip leading to the accident by a longdelayed start, the relentless pace of the trip once it was underway, and circadian rhythms and other time-of-day effects occurring at the late hour of the crash. The truck driver had risen at 5 a.m. on the day of the accident, but had had to wait for his truck to be serviced, and did not leave on his trip until between 2:30 and 3:00 p.m. By the time he got to the crash site, he had been up for 19 hours 20 minutes. The human factors expert reconstructed the driver’s itinerary, revealing that he kept up a relentless pace during approximately 10 hours of driving, averaging about 5 mph above the speed limits in the states through which he passed, which must have involved much higher speeds at some points. By his own testimony at deposition, the truck driver only stopped for a few minutes two or three times during the 10-hour trip. The last time he stopped, he said it was to go to bed for 8 hours because he was tired, but he traveled on because he did not see a parking spot. However, he admitted that there was another truck stop between that place and the accident site, at which he did not stop because he preferred another place farther on.


506

FIGURE 17.2 Charts and other graphical methods can help the expert to explain technical points in court.

FIGURE 17.3 Example of a spreadsheet chart. The plaintiff’s human factors forensic expert opined that the truck driver’s failure to slow down and stop at the construction zone was due to fatigue, drowsiness, and a lack of alertness resulting from an irregular work-rest schedule, the long (19+ hours) period


507

between starting his day and crashing, the relentless pace of his high-speed driving for almost 10 hours with no significant breaks, and driving past midnight, when time-of-day effects approach maximum likelihood of causing drowsiness. The case was settled out of court. Example 2 At 6:30 p.m., a tractor-trailer driver on a secondary road approached an intersection with a divided highway. The highway had a wide median; traffic on the secondary road was controlled by two stop signs requiring that drivers stop at the first set of lanes and then stop again in the median, prior to crossing the lanes going in the other direction. The truck driver failed to stop at the first stop sign and crossed into the path of an oncoming car, which collided with the truck, resulting in fatal injuries to a passenger in the car. The plaintiff alleged that the crash was the result of extreme driver fatigue and sought punitive damages. The plaintiff’s expert pointed to a payroll timesheet filled out by the driver as indicating excessive and illegal hours of work, leading to fatigue. The defendants retained a human factors forensic expert, who studied the driver’s work-rest history and his driving performance proximate to the accident. The work-rest history analysis revealed that in the two work weeks preceding the week of the accident, the truck driver had a regular duty schedule, with two nights and one day off at the end of each week. The expert concluded that no cumulative fatigue would be expected. During the week of the accident, the truck driver continued the pattern seen in the prior two weeks, which consisted of a medium-length drive to another city on the first day of the work week, followed by daytime local deliveries in that area and sleeping in the same motel each night. The accident occurred on the second day of local, short-haul deliveries, so the expert judged that cumulative fatigue was improbable. Similarly, acute fatigue was judged to be improbable. The driver’s task composition that day had a favorable mixture of activities, driving for a total of 4.75 hours and spending 7.75 hours making deliveries and taking breaks. The total elapsed time was well short of the 15-hour limit and well within the range of normal performance. Circadian rhythms would not be expected to be a factor at 6:30 p.m., and at the time of the accident, the driver had been driving less than 15 minutes since his last delivery, which was of about 30 minutes’ duration. Therefore, the expert judged that the energizing effects of the physical activities associated with the delivery likely carried over during this brief period of driving, making loss of alertness at the time of the accident even less likely. The defense expert used Microsoft Streets and Trips software to calculate an itinerary based on the places to which the driver had made deliveries and found that the calculated itinerary agreed well with the driver’s DOT logs. The payroll timesheet that the motor carrier required the driver to fill out showed time added to the beginning and/or end of the work days recorded in the DOT logs. Because the DOT logs accounted for all the travel that the driver accomplished, the forensic expert concluded that the driver was recording hours on the company payroll timesheet that he did not actually work. This fictitious work earned the driver more money but obviously did not result in driver fatigue. The truck driver testified at his deposition that, at the time of the accident, he was trying to find an on-ramp to the interstate highway in that area. In investigating driver


508

performance, the defense expert visited the accident site and retraced the route of the truck driver. He found that the secondary road approaching the divided highway passed through a forested area. Very near the intersection, a driver emerged from the trees and the view opened up, revealing the interstate highway some distance beyond the intersecting divided highway. An off-ramp and an overpass were visible sharply to the left and there were glimpses of the interstate moderately to the left; another overpass and on-ramp signs were straight ahead across the intersection. The expert judged that the complexity of this visual scene and its obvious relevance to the driver’s goal of getting onto the interstate stimulated substantial visual scanning. To interpret most visual cues properly, humans must use their sharp central vision by fixing the gaze upon each visual feature, sequentially. The more cues available, the longer this takes and the more likely it is that one will be missed. The driver testified at deposition that he did see the second stop sign and planned to stop there; unfortunately, he did not detect the first stop sign. The expert explained that driver behavior is influenced by expectations built up from driving experience and that this driver probably expected to see only one stop sign at the intersection. Having detected the expected stop sign (which he now knows was the second of two stop signs), he may have resumed his visual search for an on-ramp to the interstate highway. There was no question that the truck driver committed a fatal error. Extensive analyses of traffic accident data indicate that 45 to 56% of crashes involve “recognition failure.” The defense expert opined that the recognition failure in this case was more likely than not caused by distraction associated with looking for the interstate highway on-ramp. In his opinion, to a reasonable degree of scientific certainty, truck driver fatigue was not a cause of this accident. The case was settled before it got to court.

Primary factor

17.6 Checklist of Main Factors to Consider Considerations

Preceding driver errors Vehicle safety inspection Cargo securement Vehicle operating practices Proximate driver errors Attention Perception Detection Decision making Motor response Cumulative driver Driver’s logs fatigue Supporting documents Itinerary reconstruction Irregularity of rest periods Duration of rest periods Place of sleep (sleeper berth/truck moving, sleeper berth/parked, motel, home, etc.) Irregularity of work periods Duration of work periods Circadian rhythms and other time-of-day effects Sleep debt Duration and frequency of recovery periods (days/nights off)


Proximate driver fatigue

509

Compliance with hours-of-service regulations Sleep disorders Circadian rhythms Daylight/dark Time elapsed since awakening from last principal sleep period Time and duration of naps since awakening from last principal sleep period Time and duration of breaks since awakening from last principal sleep period Time, duration, and effort of cargo loading/unloading since awakening from last principal sleep period Duration of proximate driving period Average speed Monotony

Defining Terms Bus—North American term for a vehicle designed primarily to transport passengers. CDL—Commercial driver’s license. A U.S. driver’s license associated with specific classes of commercial motor vehicles and granted after successfully completing specific tests of knowledge and skill. CFR—U.S. Code of Federal Regulations. Circadian rhythms—Biological rhythms that repeat approximately every 24 hours. An internal clock, or pacemaker, that is located in the brain generates circadian rhythms. These rhythms promote sleeping during the night and alertness in the day. CMV—Commercial motor vehicle—a precisely defined class of vehicle subject to specific government rules and regulations. Cumulative fatigue—A state of fatigue caused by accumulated sleep debt over a period of several days. A work-rest schedule that results in daily sleep debt will cause an increasing state of cumulative fatigue as days go by, requiring a recovery period of days and nights off duty to restore performance. Fatigue—Driver fatigue has been defined in many ways. Wylie and colleagues (1996) defined it as a state of the driver manifested by increased lapses of attention; increased information-processing and decision-making time; increased reaction time to critical events; more variable and less effective control responses; decreased motivation to sustain performance; decreased psycho-physiological arousal (e.g., body temperature, brain waves, heart action); increased subjective feelings of drowsiness or tiredness; decreased vigilance (e.g., watchfulness); and decreased alertness (e.g., readiness to meet danger). Motor carrier—An entity that provides transportation of property or passengers and is responsible for dispatching drivers of commercial motor vehicles for that purpose, regardless of whether the drivers are employees of the motor carrier or independent contractors. An owner-operator is a motor carrier as well as a driver. Motor carriers are subject to government regulation.


510

Placard, hazardous materials—A precisely defined type of sign required on commercial motor vehicles to indicate the presence and nature of hazardous materials carried as cargo. Proximate driver fatigue—The state of fatigue proximate to a motor vehicle collision. Proximate fatigue is a combination of cumulative fatigue reflecting work-rest history in previous days and the factors influencing driver fatigue within 24 hours of the accident. Recovery period—A “weekend” or other extended off-duty period that may restore performance after a “work week” resulting in cumulative fatigue. Straight truck—North American term for a heavy cargo-bearing vehicle that is not articulated (i.e., that does not “bend” in the middle). Tractor-trailer—North American term for a heavy vehicle comprising a truck tractor, a semitrailer, and possibly one or more full trailers. Truck tractor—North American term for a heavy vehicle that bears the weight of the front part of a semitrailer and provides motive power for the articulated tractor-trailer combination. Vigilance—Sustained attention for low-probability events under monotonous conditions. There has been a strong interest in vigilance research since the studies of performance decrements of ships’ lookouts and airborne radar operators during World War II. Vigilance is required for the safe operation of motor vehicles, as well as many other critical tasks. References Abrams, C, Shultz, T, and Wylie, C.D., 1997. Commercial motor vehicle driver fatigue, alertness, and countermeasures survey. Decision Research, Santa Barbara, CA. Akerstedt, T. and Home, J., 1995. Work hours, sleepiness and accidents. J. Sleep Res. , 4(suppl. 2), 1–83. FMCSA (Federal Motor Carrier Safety Administration), 2000. Hours of service of drivers; driver rest and sleep for safe operations. Fed. Regist , 65(85), 22540–25611. Folkard, S., 1997. Black times: temporal determinants of transport safety. Accident Anal. Prev. , 29(4), 417–430. Gallup Organization, 1995. Sleep in America. The Gallup Organization, Princeton, NJ. Gillberg, M., Kecklund, G., and Akerstedt, T., 1995. Sleepiness and performance of professional drivers in a truck simulator—comparisons between day and night driving. J. Sleep Res. , 5, 12– 15. Home, J.A. and Reyner, L.A., 1995. Sleep-related vehicle accidents. Br. J. , 310, 565–567. Lavie, P., 1986. Ultrashort sleep—waking schedule, III. “Gates” and “forbidden zones” for sleep. Electro-encephalogr. Clin. Neurophysiol. , 63, 414–425. Lavie, P., 1989. To nap, perchance to sleep—ultradian aspects of napping. In Dinges, D. and Broughton, R. (Eds.) Sleep and Alertness: Chronobiological, Behavioral and Medical Aspects of Napping . Raven: New York. Mackie, R.R. and Miller, J.C., 1978. Effects of hours of service, regularity of schedules, and cargo loading on truck and bus driver fatigue. Technical Report No. 1765. Human Factors Research, Inc., Santa Barbara, CA. Mackie, R.R. and Wylie, C.D., 1991. Countermeasures to loss of alertness in motor vehicle drivers: a taxonomy and evaluation. Proc. Hum. Factors Soc. 35th Annu. Meeting , 2, 1149–1153.


511

Mackie, R.R., Wylie, C.D., Shultz, T, Engel, R., Townsend, M., Lammlein, S.E., and Johnson, S., 1989. Development of a recommended testing program for commercial motor vehicle operators: the CDL system. Essex Corporation, Santa Barbara, CA. Mitler, M.M., Miller, J.C., Lipsitz, J.J., Walsh, J.K., and Wylie, C.D., 1997. The sleep of long-haul truck drivers. New Engl. J. Med. , 337, 755–761. Mitler, M.M., Carskadon, M.A., Czeiler, C.A., Dement, W.C., Dinges, D.F., and Graeber, R.C., 1988. Catastrophes, sleep, and public policy: consensus report. Sleep , 11(1), 100–109. National Highway Traffic Safety Administration/National Center on Sleep Disorders Research Expert Panel on Driver Fatigue and Sleepiness, 1998. Drowsy driving and automobile crashes. Report No. DOT HS 808 707. (Washington, D.C.). NTSB (National Transportation Safety Board), 1990. Fatigue, alcohol, other drugs, and medical factors in fatal-to-the-driver heavy truck crashes. Safety Study NTSB/SS-90/01. Washington, D.C. NTSB (National Transportation Safety Board), 1995. Factors that affect fatigue in heavy truck accidents. Volume 1: analysis. Safety Study NTSB/SS-95/01. Washington, D.C. Rosekind, M.R., Neri, D.F., and Dinges, D.F., 1997. From laboratory to flightdeck: promoting operational alertness. Fatigue and Duty Time Limitations—an International Review . (London: The Royal Aeronautical Society), 7.1–7.14. Smiley, A. and Heslegrave, R., 1997. A 36-hour recovery period for truck drivers: synopsis of current scientific knowledge. Report No. TP 13035E. Transport Canada, Montreal. Transportation Research Board, 1987. Zero alcohol and other options: limits for truck and bus drivers. Washington, D.C. Transportation Research Institute, 1998. Potential hours-of-service regulations for commercial drivers: report of the Expert Panel on Review of the Federal Highway Administration Candidate Options for Hours of Service Regulations. University of Michigan, Ann Arbor. U.S. DOT, 1995. 1995. Truck and bus safety summit: report of proceedings. U.S. Department of Transportation, Washington, D.C. Williamson, A.M. and Feyer, A.-M., 2000. Moderate sleep deprivation produces impairments in cognitive and motor performance equivalent to legally prescribed levels of alcohol intoxication. Occup. Environ. Med. , 57, 649–655. Wylie, C.D., 2000. Hours of service, driver fatigue, and crash risk: an operations research model based on human factors data. Technical report prepared for Federal Motor Carrier Safety Administration. D. Wylie Associates, Santa Barbara, CA. Wylie, C.D. and Shultz, T, 1989. Model Driver’s Manual for Commercial Vehicle Driver Licensing . Alexandria, VA: Essex Corporation. Wylie, C.D., Mackie, R.R., and Smith, M.J., 1985. Comparative effects of 19 stressors on task performance: major results of the operator survey. Proc. Hum. Factors Soc. 29th Annu. Meeting , 1, 457–466. Wylie, C.D., Shultz, T, Miller, J.C., Mitler, M.M., and Mackie, R.R., 1996. Commercial motor vehicle driver fatigue and alertness study: project report. Federal Highway Administration Office of Motor Carrier Safety Report No. FHWA-MC-97–002. NTIS Accession No. MIC 10000009. Essex Corporation, Santa Barbara, CA.

18 Human Factors Issues in Motorcycle Collisions Peter A.Hancock University of Central Florida Tal Oron-Gilad University of Central Florida David R.Thom Collision and Injury Dynamics, Inc. 0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

18.1 Introduction In the mind of the general public, riding a motorcycle involves certain inherent dangers. Riding differs from driving in some of its basic characteristics. For example, the motorcycle rider sits “outside” in the real world, not “cocooned” inside an enclosed space as with virtually all other road vehicles. In addition, motorcycles with a constant or increasing velocity (i.e., those not slowing down) will remain upright with no immediate control inputs from the rider. However, for some phases of operation, active balance is required. This contrasts with virtually all other road vehicles, which remain fundamentally stable, independent of the driver’s control actions. These nuances of motorcycle operation are poorly understood by most of the traveling public, who rarely ride powered vehicles with less than four wheels. Because of the large differences between motorcycles and regular enclosed vehicles, the rider faces a spectrum of environmental and collision hazards generally not encountered by most car drivers. At the same time, the motorcycle rider involved in collisions is usually less culpable than the other drivers involved. In consequence, there is a circular reinforcement in which unfamiliarity is aligned with the perception of danger that subsequently discourages ridership. Our knowledge of the relative fatality rates confirms that the perception of danger in ridership is not without substance. Many societal stereotypes and associations are made with motorcycle riding through a general cultural perspective. As with all such stereotypes, each is true and false to a greater or lesser degree. This confluence of circumstances has meant that the study of motorcycle riders and their specific issues has been greatly underserved, even within the transportation research community, which itself is poorly supported as an effort to battle society’s death and injury (Hancock and de Ridder, 2003; Evans, 1991; TRB, 1990). In this chapter, we seek to redress this

Human factors issues in motorcycle collisions

513

imbalance somewhat by our focus on specific motorcycle issues and problems. This approach is selective because a full elaboration of all related issues would require a text of its own. Also, we have approached the respective problems primarily from a human factors perspective, which is one that, as well as stating problems, enjoins us to suggest solutions. 18.2 Organization of the Chapter In the first section of this chapter we seek to describe the scope of the problem of motorcycle-related crashes and examine some informative data on high fatality raterelated factors such as age, licensing, helmet usage, and alcohol consumption. In the second section, we examine motorcycle riders’ attitudes toward other road users and their contingent risk assessment. Because riding is different from driving, a spectrum of different demands is imposed on the motorcyclists. The motorcycle rider’s understandings of other vehicle drivers’ perceptual and attentional limitations are also important in establishing what riders anticipate to be more or less hazardous travel environments. The third section discusses the role of other drivers in violating the motorcyclist’s right of way and tries to elucidate the intrinsic human structural and functional limitations that lead to such driver errors. This exposition includes examining issues such as the conspicuity of the motorcycle, failure to estimate time to collision due to the physical appearance of the motorcycle, and failures related to the sustained performance nature of the driving task. The fourth section is concerned with human factors approaches to collision mitigation. These include a brief examination of means for implementing intelligent transportation systems (ITSs) for motorcyclists. The concluding section presents a summary of our observations, including a general checklist and reference listing of further works as well as a call for a systematic program of privately and federally funded research to serve this most at-risk community of road users. 18.3 Scope and Scale of the Problem In 1999, motorcycles made up less than 2% of all registered vehicles in the United States and accounted for only 0.4% of all vehicle miles traveled (NHTSA, Traffic Safety Facts, 2000). However, per vehicle mile traveled, motorcyclists were approximately 18 times as likely as passenger car occupants to die in a traffic crash and 3 times as likely to be injured. Similarly in Australia, motorcycles represent less than 1% of the traffic stream yet, per 100 million km traveled, motorcyclists were approximately 20 times as likely as passenger car occupants to be involved in a severe or fatal crash (Federal Office of Road Safety, 1997). Motorcycle fatalities in the U.S. have increased more than 50% over the past 5 years, as shown in Figure 18.1, and the percentage of motorcycle deaths in 2002 increased to 7.6% of all motor vehicle fatalities.


514

FIGURE 18.1 Motorcyclists killed by year in the U.S. (From Fatality Analysis Reporting System, wwwfars.nhtsa.dot.gov, 2002.)

FIGURE 18.2 Percentages of motorcyclists killed in 2002 in the U.S. by type of event. (From Fatality Analysis Reporting System, wwwfars.nhtsa.dot.gov, 2002.) The distribution of motorcycle-collision typology is shown in Figure 18.2. Collisions between motorcycles and other motor vehicles account for approximately 50% of all fatal motorcycle crashes. Collisions with fixed objects account for another 28% and the rest of the fatalities are attributed to collisions with nonfixed objects or other factors (e.g., loss of stability). The initial point of impact in fatal crashes involving motorcycles and other motor vehicles was predominantly the front, which accounted for approximately 38% of the total number of fatal crashes (NHTSA, FARS, 2003). This collision configuration occurs, for example, when a left-turning vehicle violates the right of way of the


515

motorcycle. These collisions are usually followed by the ejection of the rider from the motorcycle, leading to severe injury or death (see Peek-Asa and Kraus, 1996). Age and Gender Traditionally, casualty statistics showed that young motorcyclists (under 30 years of age) were more likely than older motorcyclists to be killed or seriously injured (Rutter and Quine, 1996). The high fatality rate among young motorcyclists has been frequently attributed to inexperience. However, Rutter and Quine (1996) found that it was the pattern of behavior associated with youthful individuals, such as risk-taking, that resulted in the high collision rate, not the level of experience per se. Because neither experience nor risk-taking behavior necessarily correlates with age, the factors can be separated. However, it must be remembered that risk taking and inexperience often do accompany youth; this argues for early training, especially focused on risk assessment, to improve rider safety. U.S. data available from the most recent 4-year period (NHTSA, FARS, 1999–2002) shows an interesting trend in motorcycle fatalities. Such fatalities increased radically for riders who were over 40 and the largest rate of increase in fatalities was among those over 50 (26% in 1 year). Fatality rate remained relatively high among young riders (under 30) but is now even higher among older riders (over 45), as shown in Figure 18.3. This trend should not be ignored and, indeed, older users’ limitations have become of great concern in the recent years for all vehicles (see Hakamies-Blomqvist, 1996). Such trends clearly indicate that more research is especially needed on older motorcycle riders. In general, as we grow older a whole spectrum of cognitive abilities declines (see Fozard et al, 1994; Birren and Schaie, 2001), as do muscle strength, coordination of force and grip, and general freedom of neck and limb movements (see Groeger, 2000). Although this is a general tendency, it need not necessarily be true about any single individual, especially one who is fit and active. However, it is clear that younger individuals can tolerate crashes of any specific severity more successfully than their older peers (Evans, 1988; 1991). These changes in fatality rates maybe reflective of the change in rider demographics because, with the price of some motorcycles, only those with significant disposable income can afford the larger machines. Interestingly, the definition of older users varies according to the vehicle used. Older drivers are categorized as beyond 65 years of age, but for motorcycle riders this threshold is at 40 years of age (see, for example, NHTSA, FARS, 2003). This may well reflect the traditional perception of the user population. With respect to gender differences, the collision statistics continue to confirm that motorcycle riding remains a predominantly male activity: 90% of motorcyclists killed in 2002 were male. Of the females who died in motorcycle crashes, 66% were passengers, while 99% of the males who died in motorcycle crashes were the controlling rider (NHTSA, FARS, 2003). Currently, the riding population is largely undefined due to lack of research in this area. However, evidently there are large demographic differences among the states contingent on demands, wealth, and weather. The denominator for all current motorcycle crash statistics is largely unknown. In 2000, the Motorcycle Safety Foundation and the National Highway Traffic Safety Administration (NHTSA) cosponsored the National Agenda for Motorcycle Safety (NAMS) (available at


516

www.ahainc.com/nams). This document consists of a national committee of experts’ prioritized needs for motorcycle safety for the coming decade; one of its top recommendations was to study the population at risk as part of an extensive crash study (similar to the one conducted by Hurt and his colleagues in 1981 (Hurt et al., 1981).

FIGURE 18.3 Number of motorcyclists killed, by age group, by year in the U.S. (From Fatality Analysis Reporting System, wwwfars.nhtsa.dot.gov, 2002.) Alcohol Involvement Alcohol involvement remains a significant factor in single-vehicle motorcycle fatal crashes (Williams and Hoffman, 1979a); 41% of the motorcyclists who died in singlevehicle crashes in 2000 in the U.S. were intoxicated (NHTSA, FARS, 2003). In 2000 in the U.S., the percentage of alcohol involvement in fatal crashes for motorcyclists was approximately 50% higher than for drivers of passenger vehicles, as shown in Figure 18.4 (NHTSA, 2003). Motorcycle riders had the highest intoxication rates, with blood-alcohol concentration (BAC) of 0.10 g/dl or greater. Also, motorcycle riders killed at night were nearly four times as likely to be intoxicated as those killed during the day (43% vs. 12%, respectively) (NHTSA, FARS, 2003). Some calls to decrease the legal BAC level for motorcyclists to a level more consistent with riding coordination and balance requirements have recently been made (Sun et al., 1998). However, legislation might not be as effective in reducing crash severity and fatalities because of a generally low compliance rate. In an effort to explore these questions, NHTSA enlisted a group of researchers in the fields of alcohol and motorcycle safety. This group analyzed numerous ways of exploring the relationship among motor-


517

FIGURE 18.4 Proportion of alcoholintoxicated (BAC=0.10+) motorcycle drivers and other drivers involved in fatal accidents in the U.S. (From NHTSA, Alcohol involvement in fatal crashes, 1998, 2001.) cycle riding, BAC, and impairment. These include field surveys, riding experiments, and simulator studies (Becker et al., 2003; NHTSA, 2004). Motorcycle Size and Rider Licensing Internationally, graduated motorcycle licenses restrict a novice or learner or a young rider to a relatively small motorcycle, typically 250 cc. In New Zealand, for example, Langley and colleagues (2000) found that even though about a third of novice riders and restricted license holders did not comply with this engine-size restriction, evidence did not indicate that these individuals were more likely to be involved in injurious or fatal crashes. Other graduated license systems found in the U.S. do not limit the size of the motorcycle, only the conditions and times of use. One problem with specifying limitations using engine displacement is that it does not necessarily reflect the actual performance of the motorcycle. It is possible that measures such as the ratio between power and weight could serve as better predictors of the involvement in high-risk crashes (Langley et al., 2000). However, the assumption that the ratio of power to weight is crash related has not yet been established. Current FARS data show larger motorcycles are increasingly involved in fatal crashes; however, the lack of exposure data limits any conclusion from being drawn. When studied in detail by Hurt and colleagues, larger-displacement motorcycles were actually under-represented in crash data (Hurt et al., 1981). Epidemiologically, it has been established that nearly one out of seven motorcycle riders involved in a fatal crash in 2000 in the U.S. was operating with an invalid license at the time of the collision


518

(NHTSA, FARS, 2000). In 2002, 24% of all fatally injured motorcycle riders did not have valid licenses. Furthermore, in 2002 nearly 50% of the young motorcycle riders (15to 20-year-olds) in the U.S. involved in fatal crashes were unlicensed or driving with an invalid license (NHTSA, TSF, 2002). Thus, licensing per se is not likely to have a significant impact on crash reduction if the rule is not enforced. Helmet Usage and Passive Safety Head injury is the leading cause of death in motorcycle crashes (Branas and Knudsen, 2001; Hurt et al., 1981). European as well as American reports attribute great significance in reducing damage to the compulsory use of a helmet, which, in spite of the increase in the number of motorcycles, helps in preventing a parallel increase in fatalities and severe injuries—in particular, incidence of brain and head injuries (Gopalakrishna et al., 1998; Sarkar et al., 1995). Helmets are estimated to be 29% effective in preventing fatal injuries to motorcyclists, which means that 29% of the injured riders would have had a fatal injury without wearing the helmet (NHTSA, FARS, 2000). However, the current helmet design does not provide complete protection against injury. A combination of helmet and protective clothing creates a passive safety envelope. European authorities have set a goal to lower the number of fatalities from 45,000 in 1997 to 25,000 in 2010 within the European Community by enhancing such passive safety measures (ETSC, 1997). Helmet usage in the U.S. varies because it is not now compulsory in all U.S. states. In the recent past, several U.S. states have weakened their helmet laws and overall helmet use decreased after the laws were changed (NHTSA, 2001). Some sources claim that this is the cause of an increase in motorcycle fatalities (although not all studies provide conclusive agreement; see Branas and Knudsen, 2001). A major concern is in the increased cost (the financial burden) of fatalities and injuries that are directly related to unhelmeted riders (e.g., Rowland et al., 1996; Lawrence et al., 2002). 18.4 Motorcycle Collisions from the Inside Out The foregoing has framed the issues with which we now deal. We can approach the forensic issues of motorcycle human factors from two directions. The first is from the perspective of the motorcyclist, essentially from the inside out. The second is from the point of view of other road users, or from the outside in. Almost half of all motorcycle fatalities occur without involvement of other road users (FARS, 2002) and therefore understanding both directions is important. Motorcycle Riding Demands and Capacity (Workload) The Motorcycle Task Analysis Report (MSF/NPSRI, 1974) has analyzed the demands involved in riding and driving and has indicated clearly that significant differences exist, with motorcycle riding much more complex than driving a car (see Hancock, 1995). The relative degree of physical and cognitive workload, then, is one of the factors that distinguish motorcyclists from other drivers.


519

Under ordinary driving situations, car and truck drivers have considerable spare attentional capacity, which they can allocate to other tasks. Modern engineering changes have caused cars and trucks to have reduced direct links between road geometry and driving performance. This is very much less the case with motorcycles. Motorcycle riders are by definition forced explicitly to adjust their driving speed and their body position to the demands of the driven road and the changes in the environmental context. Figure 18.5 illustrates these attentional workload differences between motorcyclists and car drivers. As shown in Figure 18.5(a), car-driving demands remain relatively constant throughout the drive unless there are unexpected circumstances. With the motorcycle rider, the contextual demands of the drive are highly correlated with riding demands, thus allowing fewer situations in which the intrinsic attentional capacity is not occupied by the riding. Distractions from this primary task of riding can then easily cause problems. Distraction can be caused by external influences such as high-density traffic, monitoring other displays, poor visibility, and adverse weather, or by internal sources of distraction such as personal concerns, anxiety, fatigue, lack of familiarity with the area, and route confusion. The lack of spare capacity of motorcycle riders is a major concern. As well as restricting adding additional displays to the motorcycle, the high level of chronic load means less attention can be paid to unusual or threatening conditions. This is a strong mandate for training so that riders can “get out in front” of the situation and anticipate rather than react to problems. One of the great problems of powered travel is that normal circumstances can rapidly change into

FIGURE 18.5 Attention capacity and driving demands for motorcycle drivers and other vehicle drivers. conditions requiring emergency response. As shown in Figure 18.5(b), driver and rider respond to emergencies by focusing all of their attention on the situation: the so-called “narrowing of attention” effect (see Hancock and Weaver, 2004). Unfortunately, the higher baseline of demand placed on the rider means that he encounters limitations earlier and more frequently than a driver does. One consequence is that forms of advance such


520

as technical adaptive aiding will help the rider more than the driver. Many intelligent transportation systems (ITS) implementations should be made on motorcycles; however, as is evident, because of market forces, they are not. Adaptive aiding strategies, techniques, and interfaces for assistive technologies such as collision warning and collision avoidance would benefit most from facing the higher-level challenge that motorcycles impose. The primary mandate of all advanced vehicle technologies is improved safety, and the best method currently to achieve this is to augment the rider’s natural capabilities. Braking is the most commonly used avoidance response. However, in emergency situations, motorcyclists do not apply the front brake fully or at all and rely predominantly, if not solely, on the rear brake (see Hurt et al., 1981, 1984a; Thom et al., 1985; McLean et al., 1979). The failure to use the front brakes in emergency may be due to the difficulty in reaching the front control rapidly (Mortimer, 2002). Among others, Mortimer (2002) has advocated the use of integrated brake systems, in which the front and the rear brakes have a combined control. Studies have shown no greater tendency to lock the wheels and skid in the integrated brake mode than in the more traditional separated configuration. The current controls provided on most motorcycles require actual training and subsequent practice to use and are fundamentally different from the single automobile brake control. Separated vs. integrated control systems is a very important topic in the human factors aspects of motorcycle design. Several brands and models of motorcycles have been offered with interconnected brakes and/or automatic brake systems (ABS). The effect of such equipment has been discussed from the European perspective but, as yet, not in U.S. field studies (Ecker et al., 2001). Risks of Motorcycling The illusion of control is one behavioral bias likely to be relevant to the cause of many unexplained single-motorcycle crashes. The illusion of control leads to perceptions that “the more control I have, the less likely I am to suffer an adverse event.” The more familiar we are with some activity, the more routine it gets. The more routine it gets, the more confident we become. When one’s degree of competence is correctly assessed, the situation is the best one can expect. However, overconfidence leads to inattention and carelessness. At the same time, the general impression that motorcycle riders are risk takers simply because they choose to ride motorcycles is largely incorrect. Although motorcycles do have a unique appeal, there are many different forms of riding. Individual rider motivations must be considered to understand the spectrum of motorcycle activity. General risk statements about the whole of the motorcycle-riding population are simply inappropriate. Two fundamental motives for riding a motorcycle apply to all forms of transportation. The first is the extrinsic motive of getting from origin to destination. The second is an intrinsic motivation in respect to the form of transport. Although many individuals engage in driving a car for the purpose of pleasure, motorcycle riding is probably more engaged with intrinsic motivation than any other major form of ground transportation. Any statement about risk and motorcycle riding must take these two fundamental motives into consideration. Many motorcycle riders are motivated by the intrinsic appeal, while others are drawn to motorcycling simply because it is one of the most efficient forms of


521

urban travel. Therefore, any statement about risk must be considered in light of previous rider choices and even circumstances that change daily, such as the weather. When riding the motorcycle is the motivation, risk taking is related to the envelope of the performance of the specific motorcycle (e.g., fast motorcycles can be ridden slowly). Those primarily motivated by riding a higher-performance machine may be interested in pushing the performance of themselves and the machine to their respective limits. Unfortunately, some individuals adopt this approach in everyday riding situations; for example, 50% of single-motorcycle crashes occur while the rider is negotiating a curve (NHTSA, 2002). This type of riding represents a danger to the rider and to surrounding road users as well. The vast majority of individual riders fall into the combined category in which the goal is explicit travel needs combined with the pleasure of motorcycle use. Often this is a communal activity in which the pleasure of the travel is enhanced by the presence of other individuals with similarly equipped machines. In such cases, the motivation toward risk taking and control of risk-taking behavior is a complex social property. Comparison of capabilities of different machines and different riders may well establish status within the group and create pressure to “push the envelope.” It is also important to note that risktaking behavior varies with age, gender, and experience. It may well be that excessive risk taking is unacceptable in groups of older riders, but may be the primary rationale for younger riders to get together. Unfortunately, little research has been conducted in this area and the issue remains unresolved. 18.5 Motorcycle Collisions from the Outside In To examine motorcycling from a driver’s perspective, consider a scenario in which a collision is imminent. The driver at the intersection is an older individual who has never ridden a motorcycle and never had a close friend or immediate family member who has. He is driving an early-model station wagon but with no obvious obstructions to vision. This driver is at an intersection waiting to turn left across traffic. He has not been waiting an inordinate interval and is not pressed for time. The motorcyclist is riding a newly acquired motorcycle without a fairing. It has active daytime running lights. The rider is experienced, having owned several previous motorcycles. Neither individual has ever had a major accident. It is noon and the intersection is in a suburban area but adjacent to a mini-mall and a gas station. The car driver waits for another car to pass and pulls directly across the pathway of the oncoming motorcyclist. Despite a last second attempt at avoidance, the motorcyclist strikes the front right side of the automobile and is pitched across the hood and onto the pavement beyond. The rider is wearing a helmet and protective clothing and escapes without significant injury. The motorcycle and the side of the car are severely damaged. The car driver sits in shock and later reports to the investigating officer that he never saw the motorcycle at any time (Hurt et al., 1981). This modal account could be modified to fit any number of actual incidents and consequently fosters the hope that common patterns of evidence can be derived from multiple accidents with similar antecedents. Unfortunately, as we penetrate deeper into this scenario, we begin to encounter the idiosyncratic—that is, individually unique—


522

characteristics that begin to separate this event from others. The rider is actually in the process of bringing a brand-new vehicle home; this is only the second time that he has ridden it in traffic. The older driver is one of a very few individuals who have had lens replacement using dual-focus implants. The intersection had been resurfaced 2 weeks ago and painting of road markings has left multiple marks upon the surface. The driver had his grandchild’s puppy in the back of the vehicle and was returning from a visit to the veterinarian surgeon. All of a sudden, an apparently bland and generic account of a “typical” left-turn collision is clothed in the detail, which is only evident in real life. For scientists and researchers, it is overwhelmingly tempting to focus on the issues with which they are familiar and on the aspects of accidents that seem amenable to mitigation by established and familiar methods and techniques. In analyzing this accident, and indeed all accident configurations, the focus must be directed first to the human component—always considered in conjunction with the vehicle and the environmental circumstances. It is standard practice to elevate the configuration of collisions to primary consideration (e.g., rear-end, side-impact). However, when we reorient our approach to consider the human as the center of concern, we cannot forget that each individual is different. It is our hope and expectation that some common capabilities shared between all drivers and riders offer some answers to accident avoidance. However, we propose here that it may well be in the differences between individuals that at least part of the answer lies. Although we are used to seeing research on accident “black spots” and now accident “black times” (Folkard, 1997), and even “accident-prone” individuals, we have still not given enough recognition that the unique aspects of some people might involve them in a specific accident event at a specific time and place. The Left-Turn Problem A person-centered accident science is a long way off (Hancock, 2004). Therefore, on the path from the present to the future, consider a predominant motorcycle collision configuration. Left-turning vehicles have always been a problem for motorcyclists. The principal issue in this situation is the failure of the left-turning driver to see the approaching motorcycle and/or to judge adequately the time available to clear the intersection (Hancock et al., 1986; Hurt et al., 1984b; Williams and Hoffmann, 1979b; Wulf et al., 1989b). Such collisions frequently involve motorcycles because they have a small frontal surface area and therefore possess smaller stimulus strength than passenger cars or trucks. An additional aspect is the difficulty in determining the approaching speed of motorcycles. Failure to observe and recognize an oncoming motorcyclist can be divided into two general forms that can be characterized as structural and functional fallacies. In the year 2002 in the U.S., 35% of the motorcycle riders (459 fatalities) died in collisions when another vehicle violated their right of way by turning left while the motorcyclist was traveling straight, passing, or overtaking (NHTSA, 2003). Although alcohol intoxication has been implicated in many motorcycle fatalities (Preusser et al., 1995), for the ran-traffic-control and left-turn-oncoming-crash types of accidents, BAC had little involvement. The majority of collisions occurred between 12 p.m. and 8 p.m. in urban


523

settings, mainly on dry pavement and when no adverse weather condition existed (Caird and Hancock, 2002; NHTSA, FARS, 2000). The influence of factors that cause drivers to interact with one another as a result of vehicle size is less clear. However, it is possible that motorcycles are perceived as less threatening because of their size and that this can cause smaller time-to-collision distances. Also, smaller cars and motorcycles are more likely to be hidden temporarily by trucks or large vehicles, which prevent greater sight distances from the driver of the car. Hurt et al. (1981) investigated personal or family motorcycle familiarity and found that individuals who have had riding experience are less liable to cause an accident with a motorcycle when they are the automobile driver than those unfamiliar with motorcycles. In postcollision interviews, the driver of the offending automobile most often claims not to have seen the motorcyclist prior to the collision. Although this might be due to postcrash shock and denial of responsibility, or at present unidentified other factors, it is possible that the collision may have been avoided had the driver been able to detect the motorcyclist earlier. In order to derive systematic and ultimately beneficial collision countermeasures, an understanding of the limitations of the information processing is an important foundation. Failure to observe and recognize oncoming motorcyclists can be divided into two general categories: structural and functional limits, described briefly next. “Structural limits” refers to the physical constituency of the sensory systems used to assimilate information. Here we consider only the limits to the visual systems, but the same principles apply to all other sensory modes. In the visual system, failure to detect a signal might be due to the masking that occurs from line-of-sight obstructions. A largescale study of motorcycle crashes by Hurt and his colleagues (1981) indicated that more than 30% of all accidents in which line-of-site information was available involved some form of physical obstruction between the automobile and the approaching motorcycle. A second form of structural limit is the simple failure to look at the direction of the motorcycle (Hancock et al., 1990). Given the normal range of the functional visual field, where peripheral vision is particularly sensitive to motion cues, such failures would seem unlikely; however, in stressful situations, an effective narrowing of the attentional field can occur (Hancock and Dirkin, 1983). Structural failure can also result from the intrinsic restriction or physical damage to the perceiver’s visual system. From the data available on collisions, the role of the foveal blind spot and scotomic failures cannot be immediately distinguished. However, results concerning the age of the offending drivers, where the groups of individuals 65+ are over-represented, argue for more thorough investigation of the structural limits to the visual system and more thorough testing for visual impairment (e.g., macular degradation). Motorcyclists present the smallest frontal surface area of any powered vehicle; thus, they probably represent the population at greatest risk from such intrinsic structural limitations. Functional limits also act to reduce the detection of the motorcyclist. Driving is a sustained attention task and is therefore fatiguing by its very nature. In the context of comprehension of the psychophysics of vigilance, two main factors influence detection efficiency: the physical characteristics of the signal and the temporal and spatial uncertainty associated with its appearance. Mackie and O’Hanlon (1977), among others, indicated that driving is the most common task that requires sustained attention. In vigilance studies, the primary factor influencing detection efficiency is the physical


524

characteristics of the signal to be observed. Principal among these primary factors is signal conspicuity. Another critical factor in sustained attention is the event rate. In general, the lower the signal-to-noise ratio is, the poorer is the detection efficiency (Davies and Parasuraman, 1982); also, as the total number of events per unit time increases, the efficiency of the detection declines (Warm et al., 1976). Motorcycles in the U.S. are approximately 2% of the registered vehicles, but they represent only 0.4% of the traffic density (FARS, 2000). This low probability of appearance is compounded by the fact that, in traffic-congested urban areas, the high traffic volume ensures a high overall event rate. In natural driving conditions, the appearances of other vehicles are not simply neutral background events, so they require the attentional capacity from the driver. These observations suggest that the motorcycle rider is at particularly high risk. With respect to the driver, the number of locations from which a motorcyclist can appear is limited. In approximately 40% of collisions, the motorcycle comes from almost directly in front of the driver. This may suggest that the spatial location is of limited importance, but the temporal uncertainty of appearance becomes of particular concern. Temporal uncertainty is directly related to critical event rate, which is the relative frequency of motorcycles compared to other road users. As the occurrence of motorcycles in traffic becomes more infrequent, the response time to their appearance increases (Warm and Jerison, 1984). It is therefore appropriate that motorcycle safety organizations such as the National Association of State Motorcycle Safety Administrators (SMSA) are trying to promote motorcycle awareness via bumper stickers (see Figure 18.6). Peek-Asa and Kraus (1996) studied the specific injury outcomes of approaching-turn collisions. They differentiate between collisions in which the turning vehicle strikes the approaching vehicle and those in which the approaching vehicle strikes the turning vehicle. They also distinguish between two groups of cases: when the motorcycle is the left-turning vehicle and when the car is the left-turning vehicle, thus creating a 2×2 matrix, as shown in Figure 18.7. In approaching-turn collisions, the car was much more frequently the left-turning vehicle. When the car was turning left, the motorcycle was the striking vehicle in more than 70% of the left-turn collisions. When the motorcycle struck the car, the rider of the motorcycle was much more likely to lose control during collision avoidance. When the left-turning car struck the motorcycle, the motorcycle rider was less likely to be ejected than in any other crash configuration. Approaching-turn collisions in which the car was turning left caused more injuries than those in which the motorcycle was turning left. The average injury severity score (ISS) of a collision was 16.34 compared to 11.26 when the motorcycle turned left. Injuries were also more severe when the motorcycle struck the car (ISS 16.7) than when the motorcycle was struck by the car (ISS 14.5). Riders in approaching-turn collisions frequently suffer lower-extremity and abdominal injuries and have head, chest, and facial injuries less frequently. The average ISS, percent of fatally injured, average days in the hospital, and average number of injuries are greater for riders in left-turn collisions than for riders in other crash types (except head-on collisions). Excessive speed is more often documented in these types of crashes. This is a critical concern and may be incorrectly reported. Drivers of the violating vehicle often state that they never saw the motorcycle prior to the collision (Hurt et al., 1981).


525

FIGURE 18.6 Increasing other drivers’ awareness of motorcycles.

FIGURE 18.7 The left-turn matrix. (Percentages taken from Peek-Asa, C.


526

and Kraus, J.F., Accident Anal Prev., 28(5), 561–569, 1996.) “Time to contact” is a phrase that describes conditions in which a person estimates when a moving object will collide with another object in space. These objects may be animate or inanimate. Time-to-contact studies describe situations in which vehicles collide with each other as well as those in which vehicles collide with stationary objects. This area has been the subject of intense research in an effort to understand collision events (see Caird and Hancock, 1994; Carel, 1961; Cavallo and Laurent, 1988; Ellingstad and Heimstra, 1969; Gibson and Crooks, 1938; Groeger and Cavallo, 1991; Hoffmann and Mortimer, 1994; Kaiser and Mowafy, 1993; Knowles and Carel, 1958; Lee, 1976; Manser and Hancock, 1996; McLeod and Ross, 1983; Schiff and Arnone, 1995; Schiff and Detwiler, 1979; Schiff and Oldak, 1990; Schiff et al., 1992; Tresilian, 1991, 1999). A review of this phenomenon is available (Hancock and Manser, 1997) and the specific relationship between this capability and motorcycle-vehicle collisions has been addressed by Hancock and colleagues (1991a). Conspicuity and Perceiving Unfamiliar Vehicles Conspicuity can be defined as the degree to which an object is distinguishable against its background—that is, its visual prominence due to its physical characteristics. Wulf and colleagues (1989a) have listed numerous studies that have manipulated the characteristics of the motorcycle and its rider in attempts to enhance Conspicuity. Running headlights during daytime or wearing fluorescent garments have been shown to increase the detection rate of motorcyclists (Bothwell, 1971). During nighttime, additional running lights in varying patterns, as well as illuminated leg shields, have been found to make motorcycles somewhat more conspicuous. However, the validity of most methods used is rather questionable. There has been little compelling evidence that these measures actually increase detectability in real traffic situations (see also Wulf et al., 1989b). In their landmark study, Hurt and colleagues (1981) found that motorcycles with daytime running lights (DRLs) were under-represented in their crash sample. In a number of countries in Europe and also in several states in the U.S., DRLs for motorcycles are compulsory. In addition, several countries have mandated DRLs for all vehicles. Although the conspicuousness of the motorcycle maybe improved by using DRLs, it is possible that this improvement decreases if other vehicles use DRLs. Little research to support both claims has been conducted and the two most expanded studies in this area are now 10 years old (Elvik, 1993; Hansen, 1994). Bothwell (1971) has approached Conspicuity as a biological problem of survival. Motorcycles are “extensions” of the human body, and as they proliferate, they must evolve rapidly in order to be successful. Living creatures use two survival mechanisms in regard to detectability. The cryptic (hiding) creatures merge and camouflage themselves. These animals have evolved complex concealment methods to save themselves from their predators. Often these methods are based upon using color; the animals are camouflaged in order to blend in with their surroundings. This is exactly the opposite of what motorcycles need.


527

The aposematic (warning) creatures have characteristics that indicate to potential predators that they are not for eating, or are dangerous or unpleasant; they use color and appearance as a warning that they are poisonous or harmful. Clearly, it is the latter group that motorcyclists need to emulate. A motor vehicle can be regarded as an aposematic creature—one to be avoided at all costs by other road users. Studies of car color for maximum detectability show that certain colors on average show up better against any background than others. Flame-orange is such a color and therefore should be used more often. As traffic density and speed increase, less time is available for identification, so color usage (or nonusage) becomes a more serious problem. The eye is limited in its capacity to sense color in motion. Tone contrast or brightness contrast between a vehicle and its background is the primary identification factor—particularly so in peripheral vision; thus, identification of an overtaking car and registering crossing vehicles at T-junctions are uses of this function. For example, in the fog, black, half-tone grays, and similar colors with tone value lower than the road will effectively “vanish,” so tone values lighter than the road are of greatest benefit. As someone has trenchantly noted, “A gray moving car becomes a half-tone gray metal coffin.” Two-tone colors may also simulate camouflage effects. A stimulus might impinge upon an individual’s senses but might not be recognized as relevant to the situation. Such failure related to stimulus identification can result in misidentification of a motorcycle or incorrect judgment of its speed. Although incorrect judgment of speed can occur for cars and for motorcycles, some evidence suggests that judgments of car speed are more accurate than judgments of motorcycle speed (Stroud et al., 1980; Wulf et al., 1989a). Furthermore, the gap sizes accepted for motorcycles have been found to be smaller for motorcycles than for automobiles or trucks (Hancock et al., 1991b). A number of studies have shown that illuminated highlights increase motorcycle conspicuity so that drivers are more likely to notice them, but there is little evidence that crash incidence has declined (Vredebburgh and Cohen, 1995; Cercarelli et al., 1992; Sparks et al., 1993, Williams and Hoffman, 1979b; Wulf et al., 1989a). Other suggestions to increase motorcycle and rider conspicuity include increasing the frontal surface of the motorcycle, bright clothing, and added elements to the motorcycle (William and Hoffman, 1979b; Hurt et al., 1981). Why are motorcycles overinvolved with other vehicles when the central issue is motion in depth? The total front surface area, that is, the physical size of the respective appearing vehicle, is substantially different for cars and motorcycles. At a common distance, the area they occlude may be similar (Eberts and MacMillan, 1985). The fundamental difference lies in the shape of the respective vehicles. According to strict ecological theory, the “time to contact” or, more precisely, the “time to passage” (Manser and Hancock, 1996) of the two vehicles should be coincident; from this basic perceptual capacity, we should see no differences in the collision rate if front surface area is exactly the same. However, we do not see this equivalence and motorcycles are far more likely to be turned across (see Hancock et al., 199la). Why is this? During daytime, the individual perceiving the motion in depth can see the morphology, or shape, of the oncoming vehicle. This shape is essentially square or rectangular for almost all passenger vehicles and this unitary dimensionality often connects to a manmade object. Indeed, as one goes about the world, it is evident that


528

nature abhors straight lines, but human designers revel in them. Because of this characteristic, most vehicles are quickly recognized as human creations. In contrast, the motorcycle and its rider are represented as a fractal shape and, as such, are more likely to be perceptually “camouflaged” against a predominantly fractal background. The color manipulations used to promote conspicuity are a variant of this confluence. By presenting fluorescent orange and yellow that do not naturally occur, one seeks to elevate the salience of the object above its background. However, as Gilbert and Sullivan opine, “when everyone is somebody, then no one’s anybody”; efforts to improve conspicuity can backfire, such as when workers dressed in fluorescent orange garments stand against a fluorescent orange vehicle. Consequently, the fractal argument can be reversed in environments that exclude nature and emphasize manmade objects. The notion of conspicuity and the degree to which a vehicle resembles the background against which it is perceived have much face validity. Obviously, conspicuity is a complex issue, especially in time-varying dynamic environments such as those involved in driving. Well-meaning interventions might well be counterproductive and issues such as modular running lights should be implemented with caution. Some psychological manipulations of the physical characteristics of front profiles can increase perceived size but, again, increasing conspicuity for any other particular vehicle decreases the proportional conspicuity of all other objects in the perceived field of view. 18.6 Human Factors Approaches to Crash Mitigation Ground transport systems were historically developed almost entirely to accommodate trucks and passenger vehicles. In the past years, the needs of other vulnerable road users (pedestrians and cyclists) have been formally recognized and are gradually being incorporated into road and transport systems. This oversight is now changing. For example, the AASHTO Strategic Highway Safety Plan’s current phase 3 specifically addresses numerous operator-related factors. These include distracted and fatigued drivers as well as motorcycles specifically. The earlier phases of this project are available from the AASHTO web site (AASHTO, 2003). Historically, there has been virtually no recognition of motorcyclists as road users with special needs. Instead, motorcycle riders tend to have been subsumed under broader categories of motorists. Authorities have failed to recognize motorcycles as a separate and distinct class of road users. Part of the reason for this failure lies in the low exposure rate of motorcycles compared to other road users. Other reasons for under-representation of motorcycles in future transport system planning and, especially, advanced ITSs are perhaps partly related to technical difficulties in providing motorcyclists with additional information. Consequently, the impact of the implementation of ITSs on the future of motorcycle riding and motorcycle safety is a cause for great concern. The widespread application and development of ITSs gives little, if no, current acknowledgment to motorcycle riders. It appears that many transportation management centers (TMCs) have relatively little reference to motorcyclists. True, they are served as part of the general traveling public, but there has been essentially no plan for their unique user needs or capabilities. For


529

example, if motorcycle access to the freeway system is not regulated, the traffic controller has little reason to view motorcycles as entities to be actively “managed.” This might be a seen as a potential advantage for motorcycle riders. In several large freeway ramp-regulated conurbations, motorcyclists have preferential access to highoccupancy vehicle (HOV) “diamond ramps” that provide immediate entry to the freeway system. Like drivers of carpools, buses, and emergency and law enforcement vehicles, motorcycle riders can enter the freeway system without waiting for ramp-metered access. In some jurisdictions, they can use HOV lanes. Consequently, in many current organizational structures, motorcycles are relatively “invisible” to the system. This has benefits as well as drawbacks. If motorcycles are a source of little concern, it is small wonder that vehicles like large trucks occupy a primary focus in TMC operations—especially because they are so visible and influential in the traffic stream. In some jurisdictions, motorcycles are able to “split traffic” and thus are less liable to be of concern in stalled traffic flow. This efficient traffic congestion benefit of motorcycles is illegal in many states, even though it has never been established as a safety problem (Hurt et al., 1981). Unfortunately, motorcycles do affect TMCs in the most critical condition of all—road traffic accidents—especially because motorcycles are often vulnerable to the unexpected maneuvers of other drivers. Thus, when it comes to collisions, motorcycles suddenly enlarge to become of equal concern to the TMC in coordinating emergency response and managing traffic flow around the incident. However, given the problems of access in congested traffic incidents, it may well be that motorcycles hold a unique potential as first responders to collision events. Already such unique capabilities are evident in the response of police officers on motorcycles, who frequently reach crash scenes first and can then take charge of on-scene coordination via ITS communication links. This role is especially important in urban areas that suffer from unique restrictions such as extended tunnels (e.g., Boston) or prolonged bridges (e.g., San Diego, San Francisco). This observation is particularly true for other parts of the world such as Europe, Japan, and some Latin American countries whose traffic problems and constraints on physical access exceed even those in the U.S. Because many motorcycle accidents involve only the single vehicle, one critical question concerns how modern machine-mounted technologies can help the injured or disabled rider. For example, if the motorcycle recorded a sudden stop or a horizontal position, or if information showed that the rider left the saddle in a non-normal manner, an emergency beacon could be activated. Activation of this emergency beacon could be tracked in central facilities such as the TMC (of the immediate future) or perhaps direct a GPS-specified link to emergency services. The warning beacon could also activate emergency systems onboard the motorcycle to attract the attention of individuals in the area. This form of communication is absolutely vital to the motorcyclist given the nature of many single-vehicle, run-off-the-road incidents. First indicated by Hancock (1995), this possible development was discussed in NAMS (2000). Although the majority of accidents occur in urban areas, the current majority of fatal accidents occur in rural areas. This may be in part due to the additional time taken for response in rural environments, which cuts directly into the “golden hour” of survival follow an accident. With emergency transponder beacons directly attached to GPS locator systems, the response time of emergency services in finding and rendering aid to injured


530

motorcyclists should significantly improve the motorcyclists’ chances of survival. However, we do not need to think of motorcycles solely as the victim in accident conditions. Frequently, a motorcycle officer is the first to an accident scene, especially when such accidents result in severe congestion. Using the motorcycle’s unique capabilities in navigating very crowded roadways, the officer then becomes the point coordinator for response. With the digital capabilities envisaged in this system, the response coordination role is significantly enhanced, as is the ability to render immediate aid. 18.7 Summary and Conclusions Motorcycle collisions and the over-representation of riders in fatal collisions represent a significant problem in modern transportation. Like market-driven industries, however, governmental agencies seek to place their safety improvement resources in areas that have the greatest return for investment. Under this strategy, the motorcycle-riding segment of the traveling public is neglected purely because of numbers. In consequence, few specific technologies are developed for motorcycle use. Our solution to this conundrum is technology transfer from all ITS innovations and the dissemination of understanding that all collisions are human-centered events (Hancock, 2004). The general perception of motorcycles as dangerous vehicles remains contentious. It is clear from archival data that in motorcycle-vehicle collisions, the principal responsibility in more than 50% of the cases lies with other road users failing to recognize, adapt to, and avoid motorcyclists. After a collision has occurred, there is a further perception that injury from motorcycles is more severe than that from other vehicles involved in similar collision conditions. Here, again, an epidemiological fallacy raises its head. The flaw lies in the use of the term “similar conditions.” Unfortunately, collisions are highly nonlinear events; thus, we never get similar conditions sufficient to make generalizations about equivalent risk. It is the case that motorcycles are sufficiently qualitatively different from other powered road users that trying to compare crash events across automobiles and motorcycles is fundamentally misleading. Regardless of the denominator, the relative lack of occupant protection in motorcycles means more injuries and deaths. Consequently, the perceived risks of motorcycle riding are founded upon eclectic views of different segments of society. Such views of risk, for example, differ radically as to whether someone consistently rides a motorcycle or not. They also differ as to whether an individual has ever ridden a motorcycle or not. Individuals who have had riding experience are less liable to be involved in a collision with a motorcycle when they are the automobile driver. In essence, they become “tuned” to the potential of motorcycle appearance and behave accordingly. This is reflected in bumper stickers such as “look twice save a life—start seeing motorcycles.” Our efforts can be well directed at riders so that they also can understand sources of risk. Accurate perceptions of the potential risks imposed by offending vehicle drivers (e.g., the left-turning driver) are a vital component of the rider’s defense. How to tune such perceptions might be as important in safety as high-level movement skills. Some have argued that skill-based models of rider behavior cannot account for collision likelihood and risk taking. We claim that skilled behavior is valuable in collision


531

avoidance and should be encouraged. Thus, high-level training, especially in avoidance maneuvers, is a valuable endeavor. However, it is clear that much more work is needed on this issue. Although some have examined the efficacy of rider-training programs, the use of skills in highly unusual situations and in imminent collision avoidance is certainly a topic that needs to be a strong research focus (Hancock and de Ridder, 2003). This is especially true in light of contemporary efforts to develop artificial, technology-based collision-avoidance systems (see Hancock, 1993). The future of our road transportation systems is uncertain. In North America, it has been evident since the final decades of the 20th century that we cannot simply “build” ourselves out of the problem. In virtually every major global conurbation, we have simply run out of space to do so. The offering of the technology community was to use advanced computational systems, located on the vehicle and in remote management centers, to use the existing space much more efficiently. The promise of ITSs is in part real because there are opportunities to make much better use of existing roadways. Also, an important effort has been made to explore alternative transportation options (such as light rail and personal rapid transit [PRT]), although these encounter significant political opposition from those entrenched in the current automobile industry. In all this, the motorcyclist has largely been passed over. Few have bothered to comment on how motorcycles occupy much less space and could therefore be a much more efficient transport medium in congested spaces. Motorcycles are only evident when they emerge, problematically, in crash statistics and on the road, where such collisions cause significant disruptions to traffic flow and upset the careful computer models of transit time. Initially, ITSs promised to improve safety significantly by reducing vehicle-collision frequency. That promise has yet to be fulfilled, although it is still in its early days in terms of implementation. Eventually, all countries relying on such powered transportation will run out of space altogether. At some point, ever-larger vehicles cannot be packed onto finite freeway space. On the present trajectory, an “auto-freeze” or a state of global gridlock will be reached. It is here that motorcycles may come to the fore. If we can improve perception and reality of their safety, motorcycles can represent a viable personal rapid transit alternative. To accomplish this farsighted aim, we need to invest now in motorcycle ITS systems and in thorough understanding of motorcycle-collision events. Part of this endeavor must come from the forensic and legal community. 18.8 Checklist • Riders • Age, licensing, helmet usage, and alcohol consumption • Brakes and brake control (separated vs. integrated control systems) • Collision mitigation • Distraction • Intelligent transportation systems • Riding demands and workload • Rider’s attitude and risk taking • Training issues


532

• Other vehicle drivers • Blind spot and scotomic failures • Conspicuity of the motorcycle • Failure to estimate time to collision • Functional limitations that lead to driver errors • Intersection • Right-of-way violations • Structural limitations that lead to driver errors • Sustained attention and event rate • The left turn • Time to contact

References AASHTO (2003). American Association of State Highway and Transportation Officials. (1997) Phase 1, a comprehensive plan to substantially reduce vehicle-related fatalities and injuries on the nation’s highways, http://safety.transportation.org/ Becker, L.R., McKnight, A.S., Nelkin, V.S., and Piper, D.L. (2003). Drinking, riding, and prevention: a focus group study (DOT HS 809 490), Washington, D.C.: National Highway Traffic Safety Administration. Birren, J.E. and Schaie, K.W. (Eds.) (2001). Handbook of the Psychology of Aging , 5th ed., Academic Press, San Diego, CA. Bothwell, P.W. (1983). Motor cycle conspicuity: a photographic study of motorcyclists clothing in traffic situations, unpublished technical report. Bothwell: Bedford on Avon, England. Bothwell P.W. (1971). Seeing on road safety-colour and motor vehicle safety, collected unpublished papers. Bothwell, Bedford on Avon, England. Branas, C.C. and Knudson, M.M. (2001). Helmet laws and motorcycle rider death rates. Accident Anal. Prev. , 33(5), 641–648. Caird, J.K. and Hancock, PA. (2002). Contributing factors to left turn and gap acceptance crashes. In: R.E.Dewar and P.Olson (Eds.), Human Factors in Traffic Safety , (613–652). Lawyers and Judges Publishing: Tucson, AZ. Caird, J.K. and Hancock, P.A. (1994). The perception of arrival time for different on-coming vehicles at an intersection. Ecological Psychol , 2, 295–324. Carel, W.L. (1961).Visual factors in the contact analog. Publication No. R61ELC60.34, Ithaca, NY: General Electric Company Advanced Electronics Center. Cavallo, V. and Laurent, M. (1988). Visual information and skill level in time-to-collision estimation. Perception , 17, 623–632. Cercarelli, L.R., Arnold, P.K., Rosman, D.L., Sleet, D., and Thornett, M.L. (1992). Travel exposure and choice of comparison crashes for examining motorcycle conspicuity by analysis of crash data. Accident Analysis and Prevention , 24(4), 363–368. Davies, D.R. and Parasuraman, R. (1982). The Psychology of Vigilance . London: Academic Press. Eberts, R.E. and MacMillan, A.G. (1985). Misperceptions of small cars. In: R.E.Eberts and C.G.Eberts (Eds.), Trends in Ergonomics/Human Factors II , (33–39). North-Holland: Amsterdam. Ecker H., Wassermann, J., Ruspekhofer, R., Hauer, G., and Winkelbauer, M. (2001). Brake reaction times of motorcycle riders. Proc. Int. Motorcycle Safety Conf. , March 1–4, 2001, Orlando, FL.


533

Ellingstad, V.S. and Heimstra, N.W. (1969). Velocity-time estimation as a function of target speed and concealment extent. Hum. Factors , 11, 305–312. Elvik, R. (1993). The effects on accidents of compulsory use of daytime running lights for cars in Norway. Accident Anal. Prev. , 25(4), 383–398. ETSC (European Transportation Safety Council). (1997). Safety Monitor , http://www.etsc.eu/ Evans, L. (1991). Traffic Safety and the Driver . Van Nostrand Reinhold: New York. Evans, L. (1988). Older driver involvement in fatal and severe traffic crashes. J. Gerontol: Soc. Sci. , 43, 186–193. Federal Office of Road Safety (1997). Vehicle type and the risk of traveling on the road. Monograph 17. Canberra: Federal Office of Road Safety, Australia. Folkard, S.O.S. (1997). Black times: temporal determinants of transport safety. Accident Analysis and Prevention , 29, 417–430. Fozard, J.L., Vercruyssen, M., Reynolds, S.L., Hancock, P.A., and Quilter, R.E. (1994). Age differences and changes in reaction time: the Baltimore Longitudinal Study of Aging. J. Gerontol: Psychological Sci. , 49, 179–189. Gibson, J.J. and Crooks, L.E. (1938). A theoretical field analysis of automobile driving. Am. J. Psychol , 51, 453–471. Gopalakrishna, G., Peek-Asa, C, and Kraus, J.F. (1998). Epidemiologic features of facial injuries among motorcycles. Annals of Emergency Medicine , 32(4), 425–430. Groeger, J.A. and Cavallo, V, (1991). Judgments of time to collision and time to coincidence. In A.G. Gale, I.D.Brown, C.M.Haslegrave, I.Moorehead, and S.Taylor (Eds.), Vision in Vehicles III , (27–34). Amsterdam: North-Holland. Groeger, J.A. (2000). Understanding Driving: Applying Cognitive Psychology to a Complex Everyday Task , Routledge: Psychology Press. Haddon, W., Jr. (1970). On the escape of tigers: an ecologic note. Am. J. Public Health , 60(12), 2229–2234. Hakamies-Blomqvist L. (1996). Research on older drivers: a review. IATSS Res. , 20, 91–101. Hancock, P.A. (2004). The tale of a two-faced tiger. Ergonomics Design , in press. Hancock, P.A. (1995). Motorcycles in intelligent transport systems. Technical report for the Motorcycle Safety Foundation, Irvine, CA. Hancock, P.A. (1993). Evaluating in-vehicle collision avoidance warning systems for IVHS. In: E.J.Haug (Ed.). Concurrent Engineering: Tools and Technologies for Mechanical System Design , Berlin: Springer-Verlag. PP. 947–958. Hancock, P.A., Caird, J.K., and Johnson, S.B. (1991a). The left-turn. In: Proc. Int. Ergonomics Assoc. , Paris, France, July. Hancock, P.A., Caird, J.K., Shekhar, S., and Vercruyssen, M. (1991b). Factors influencing drivers’ left-turn decisions. Proc. Hum. Factors Soc. , 35, 1139–1143. Hancock, P.A. and de Ridder, S.N. (2003). Behavioural accident avoidance science: understanding response in collision incipient conditions. Ergonomics , 46(12), 1111–1135. Hancock, P.A. and Dirkin, G.R. (1983). Stressor induced attentional narrowing: implications for design and operation of person-machine systems. Proc. Hum. Factors Assoc. Can. , 16, 19–21. Hancock, P.A., Hurt, H.H., Jr., Ouellet, J.V., and Thom, D.R. (1986). Failures of driver sustained attention in the etiology of motorcycle-automobile collision. Proc. Annu. Meeting Hum. Factors Assoc. Can. , Toronto, Canada. Hancock, P.A. and Manser, M.P. (1997). Time to contact: more than tau alone. Ecological Psychol. , 9(4), 265–297. Hancock, P.A., Rahimi, M., Wulf, G., and Briggs, R. (1988). Analyzing the behavior of left-turning drivers. In: Proc. Int. Tech. Conf. Advanced Safety Vehicles , Gothenburg, Sweden. Hancock, P.A. and Weaver, J.L. (2004). Temporal distortions under extreme stress. Theor. Issues Ergonomic Sci. , in press. Hancock, P.A., Wulf, G., Thom. D., and Fassnacht, P. (1990). Driver workload during differing driving maneuvers. Accident Anal. Prev. , 22, 281–290.


534

Hansen, L.K. (1994). Daytime running lights (DRL): experience with compulsory use in Denmark. Danish Council of Road Safety Research, Copenhagen, Denmark. Hoffmann, E.R. and Mortimer, R. (1994). Driver estimation of time to collision. Accident Anal. Prev. , 26, 511–520. Hurt, H.H., Jr. and DuPont, C.J. (1977). Human factors in motorcycle accidents. Society of Automotive Engineers, No. 770103. Hurt, H.H., Jr., Ouellet, J.V., and Thom, D.R. (1981). Motorcycle accident cause factors and identification of countermeasures (DOT HS 805 862). Washington, D.C.: National Highway Traffic Safety Administration. Hurt, H.H. Jr., Thom, D.R., and Hancock, P.A. (1984a). The effect of hand position on motorcycle brake response time. Proc. Hum. Factors Soc. , 28, 791–794. Hurt, H.H., Jr., Hancock, P.A., and Thom, D.R. (1984b). Motorcycle-automobile collision prevention through increased motorcyclist frontal conspicuity. Proc. Hum. Factors Soc. , 28, 795–798. Kaiser, M.K. and Mowafy, L. (1993). Optical specification of time-to-passage: observers sensitivity to global tau. J. Exp. Psychol: Hum. Percept. Performance , 19, 1028–1040. Knowles, W.B. and Carel, W.L. (1958). Estimating time-to-collision. Am. Psychol , 13, 405–406. Langley, J., Mullin, B., Jackson, R., and Norton, R. (2000). Motorcycle engine size and risk of moderate to fatal injury from a motorcycle crash. Accident Anal. Prev. , 32(5), 659–663. Lawrence, B.A., Max, W, and Miller, T.R. (2002). National Highway Traffic Safety Administration—costs of injuries resulting from motorcycle crashes: a literature review, NHTSA DOT HS 809 242. Lee, D.N. (1976). A theory of visual control of braking based on information about time-tocollision. Perception , 5, 437–459. Mackie, R.R. and O’Hanlon, J.F. (1977). A study of the combined effects of extended driving and heat stress on driver arousal and performance. In R.R.Mackie (Ed.), Vigilance: Theory, Operational Performance and Physiological Correlates . New York: Plenum Press. Manser, M.R and Hancock, P.A. (1996). The influence of approach angle on estimates of time-tocollision. Ecological Psychol. , 8, 71–99. Maycock, G. and Lockwood, C.R. (1993). The accident liability of British car drivers. Transp. Rev. , 13, 231–245. McLean, A.J., Brewer, N.D., Hall, C.T., Sandow, B.L., and Tamblyn, P.J. (1979). Adelaide indepth accident study 1975–1979. Part 4: motorcycle accidents. Adelaide: University of Adelaide. Research report for Australian Road Research Board and Department of Transport, Australia. McLeod, R.W. and Ross, H.E. (1983). Optic flow and cognitive factors in time-to-collision estimates. Perception , 12, 417–423. Mortimer, R.G. (2002). Motorcyclists’ brake operation, motorcycle brake controls and a case study: the need for human factors engineering. Proc. Hum. Factors Ergonomics Soc. , 46, CD listing. Motorcycle Safety Foundation, National Public Service Research Institute. (1974). Motorcycle task analysis. Irvine, CA. NAMS (National Agenda for Motorcycle Safety) (2000), http://www.nhtsa.dot.gov/people/injury/pedbimot/motorcycle/00-NHT-212motorcycle/index.html NHTSA. Fatality Analysis Reporting System (FARS) Web-Based Encyclopedia, wwwfars.nhtsa.dot.gov , 2003. NHTSA Traffic Safety Facts (2000). http://www.nhtsa.dot.gov/, report number DOT-HS-809–326. NHTSA Traffic Safety Facts (2002). http://www.nhtsa.dot.gov/, report number DOT-HS-809–619. NHTSA, Methodology for Determining Motorcycle Operator Crash Risk and Alcohol Impairment (2004). Volume I: Summary Report on Alternative Approaches with Priorities for Research, Volume II: Literature Review Report, Volume III: Expert Panel Workshop Report. Pacific Institute for Research and Evaluation, DOT-NHTSA (in press).


535

Ouellet, J.V. (1990). Appropriate and inappropriate strategies for injury reduction in motorcycle. Society of Automotive Engineers Congress, SAE Paper 900747. Peek-Asa, C. and Kraus, J.F. (1996). Injuries sustained by motorcycle riders in the approaching turn crash configuration. Accident Anal. Prev. , 28(5), 561–569. Preusser, D.F., Williams, A.F., and Ulmer, R.G. (1995). Analysis of fatal motorcycle crashes: crash typing. Accident Anal Prev. , 27(6), 845–851. Rowland, J., Rivara, E., Salzberg, P., Soderberg, R., Maier, R., and Koepsell, T. (1996). Motorcycle helmet use and injury outcome and hospitalization costs from crashes in Washington State. Am. J. Public Health , 86, 41–45. Rutter, D. and Quine, L. (1996). Age and experience in motorcycling safety. Accident Anal. Prev. , 28, 15–21. Sarkar, S., Peek-Asa, C., and Kraus, J.F. (1995). Fatal injuries in motorcycle riders according to helmet use. J. Trauma , 38(2), 242–245 Schiff, W. and Arnone, W. (1995). Perceiving and driving: where parallel roads meet. In P.A.Hancock, J.M.Flach, J.K.Caird, and K.J.Vicente (Eds.), Local Applications to the Ecology of Human-Machine Systems . Hillsdale, NJ: Erlbaum, 1–36. Schiff, W. and Detwiler, M. (1979). Information used in judging impending collisions. Perception , 8, 647–658. Schiff, W. and Oldak, R. (1990). Accuracy of judging time to arrival: effects of modularity, trajectory and sex. J. Exp. Psychol: Hum. Percept. Performance , 16, 303–316. Schiff, W, Oldak, R., and Shah, V. (1992). Aging persons’ estimates of vehicular motion. Psychol. Aging , 7, 518–525. Sparks, G.A., Neudorf, R.D., Smith, A.E., Wapman, K.R., and Zador, P.L. (1993). The effect of daytime running lights on crashes between two vehicles in Saskatchewan: a study of a government fleet. Accident Analysis and Prevention , 25, 619–625. Stroud, P.G., Kirkby, C., and Fulton, E.J. (1980). Motorcycle conspicuity. Proc. Intl. Motorcycle Conf. , Motorcycle Safety Foundation, IV, 1705–1722. Sun, S.W, Kahn D.M., and Swan, K.G. (1998). Lowering the legal blood alcohol level for motorcyclists, Accident Anal Prev. , 30(1), 133–136. Thom, D.R., Arao, H.G., and Hancock, P.A. (1985). Hand position and motorcycle front brake response time. Proc. Hum. Factors Soc. , 29, 278–281. TRB (Transportation Research Board). (1990). Safety research for a changing highway environment. Special report #229, National Research Council, Washington, D.C. Tresilian, J.R. (1999). An analysis of recent empirical challenges to an account of time-to-collision perception. Percept. Psychophys. , 61, 515–528. Tresilian, J. (1991). Empirical and theoretical issues in the perception of time to contact. J. Exp. Psychol.: Hum. Percept. Performance , 17, 865–876. Vredebburgh, A.G. and Cohen, H.H. (1995). Enhanced motorcycle visibility through use of motorcycle conspicuity enhancement system. Proc. Hum. Factors Ergonomics Soc. , 39, 1048– 1052. Warm, J.S. and Jerison, H.J. (1984). The psychophysics of vigilance. In J.S.Warm (Ed.), Sustained Attention in Human Performance (15–60). Chichester: Wiley. Warm, J.S., Wait, R.G., and Loeb, M. (1976). Head restraint enhances visual monitoring performance. Percept. Psychophys. , 20, 299–304. Williams, M.J. and Hoffmann, E.R. (1979a). Alcohol use and motorcycle accidents. Accident Anal. Prev. , 11(3), 199–207. Williams, M.J. and Hoffmann, E.R. (1979b). Motorcycle conspicuity and traffic accidents. Accident Anal. Prev. , 11(3), 209–224. Winsum. V.W. (1996). From adaptive control to adaptive driver behaviour. The Traffic Research Centre VSC, Groningen.


536

Wulf, G., Hancock, P.A., and Rahimi, M. (1989a). Motorcycle conspicuity: an evaluation and synthesis of influential factors. J. Safety Res. , 20, 153–176. Wulf, G., Hancock, P.A., and Rahimi, M. (1989b). Some causes of automobile—motorcycle collisions. Proc. Hum. Factors Soc. , 33, 910–914.

IV Physical and Cognitive Factors

19 Perceptual-Cognitive and Biomechanical Factors in Pedestrian Falls H.Harvey Cohen Error Analysis, Inc. Cindy A.LaRue Error Analysis, Inc. 0–415–28870–3/05/$0.00+$ 1.50 © 2005 by CRC Press

19.1 Introduction: Human Factors Principles and Relevance to Standard of Care Walking seems like such a simple, natural task that the body performs nearly without thinking. However, this is far from the truth. Breaking down biomechanical processes has shown that walking is actually a series of imbalances. Studies of perceptual and cognitive issues show pedestrians gather information about the walking environment using the processes of the human information model. These processes interact and determine how the individual gets from one point to another in a given environment. In general, pedestrians believe that walking surfaces are flat and uniform, unless they notice a particular discrepancy. Any unnoticed or unexpected variation in the pedestrian’s path can easily lead to a fall and potential injury. Oftentimes, the injury is severe enough to warrant filing a lawsuit to try to recover damages from a party deemed responsible for the dangerous condition. In this case, a human factors/ ergonomic (HF/E) professional can be used to render an opinion as to whether a situation is unsafe and, if so, if it led to the incident in question. When determining the cause of a pedestrian fall, the HF/E professional looks at many human factors issues, including perception, reaction, expectation, visual cues, distractions, attention, motor skills, abilities, limitations, and decision-making. The expert formulates his opinion based on the facts of the case and any applicable law or standard and then must support this theory in a court of law. At issue generally is whether the party responsible for the premises on which the fall occurred met the standard of care. In most cases, this requires the owner/manager of the property to use a level of care of a prudent adult in order to prevent an unreasonable risk of harm. It is up to the HF/E forensic professional to determine if this standard was met.


540

19.2 Objective and Scope The objective of this chapter is to discuss the perceptual-cognitive and biomechanical factors involved in pedestrian falls. This chapter also describes the role of forensic HF/E professionals and how they use their scientific background in order to formulate expert opinion. 19.3 Discussion of Principal Issues A discussion of pedestrian falls must begin with an understanding of how people walk and what they notice about the environment in which they are walking. In order to discuss the principal issues involved in pedestrian fall incidents, it is important first to discuss the biomechanical processes involved in human locomotion. Biomechanics of Walking Walking across a flat surface would seem to be one of the least complicated activities performed by a human being on a daily basis. It is generally an almost mindless task requiring little, if any extraneous thought. Obstacles such as steps, curbs, doorways, or ramps receive a rudimentary glance and, although they do register in the brain, require little formal adjustment by the body. On closer examination of walking, the biomechanical and physical parameters of human locomotion present a very different picture. From a biomechanical viewpoint, walking can be seen as a complicated process consisting of a series of continuous losses and recoveries of balance. Center of Gravity On level ground, walking can be considered as the forward translation of the body’s center of mass. The center of mass, also called the center of gravity (COG), is a theoretical point that acts as if the entire mass of an object were concentrated at that point. The location of the COG in relation to the base of the object determines the object’s stability. On a human body, the feet are the supports that create the base area under the body. The size of the base area depends on how far apart the feet are located; the larger the base area is, the more stable the person is. When one is walking, the body is constantly adjusting in order to keep the COG over the base area, thus ensuring stability. If the COG leaves the base area, the body shifts, attempting to re-establish equilibrium. If equilibrium cannot be re-established, a fall results. When upright, the COG of the human adult body is located within the body at the level of the second sacral vertebra, about 2 in. (5 cm) behind the imaginary line joining the hip joints, in the midline of the body. For a male, this is about 55% of the distance from the feet to the top of the head, at about belt level. For women, it is just above belt level, about 57% of the distance from feet to head. The exact COG location varies

Perceptual-cognitive and biomechanical factors in pedestrian falls

541

according to body type and age and moves up, down, or sideways in the course of movement and position changes. The accumulation of fat and loss of soft tissue tone are the most common reasons for the altering of the COG; tissue accumulating above the waist raises the COG, while accumulations below the belt lower it.

FIGURE 19.1 Gait cycle. The Gait Cycle The normal walking cycle comprises two phases: stance phase, when the foot is in contact with the ground, and swing phase, when the foot is moving forward through the air. The stance phase makes up approximately 60% of the walking cycle. This phase is divided into three distinct parts: heel-strike; foot flat (support); and push-off. The swing phase, which begins with the push-off and ends with the heelstrike, makes up the remaining 40% of the gait cycle. Figure 19.1 illustrates the up-down displacement of the COG during walking. As shown, each individual step begins with the trailing foot supporting all the weight. Momentum and a push from the supporting foot cause the trunk of the body to move forward. This brings the COG past the forward edge of the supporting leg. Because the COG is now outside the base area, stability is decreasing. Additionally, gravity is now acting on the body, pulling it forward and down. The free leg swings forward and is placed on the ground. The two legs now supply a much wider base area, thereby putting the COG back inside the base of support. This restores stability to the body and saves it from a forward fall. The legs alternate in their function, each in turn pushing the COG out of the support base and then swinging forward to restore it (Rasch and Burke, 1978). Between the two steps, at heel-strike, at the moment at which the weight is transferred from one foot to the other, the support base is bounded by the toes on the posterior foot and the edge of the heel on the anterior foot. This is the point of greatest instability, even though the feet are at maximum separation, creating a very large support base. The instability is due to the extremely small contact area between the supports and the floor (the toes of the back foot and the edge of the heel of the leading foot) and also the angle that they create with the floor. The majority of slip and fall incidents occur at this point.


542

Forces of Movement In order for there to be motion of the body, an external force or forces must act through the body’s COG. For a person standing upright, the only forces acting on the body are weight pushing down on the ground and the force of the ground pushing up on the body (Newton’s third law of motion). This upward force is called the ground reaction force (GRF). Because the body does not move up and down, the sum of the two forces acting on the body must be equal to zero. Because the forces act in opposite directions, they must be equal and opposite. If the forces acting through the COG were unequal, then motion would result. The weight of the body does not change, so the only way to achieve motion in the vertical plane is to vary the GRF. This results in an external force on the body that is greater than zero, causing motion. Although the GRF is the force that causes movement, it can only oppose a force applied to it. To change the GRF, the direction of application or its magnitude must change. If the GRF points straight up, opposing the weight of the body, it must be greater than the weight to cause upward motion. If the GRF is less than the weight and still points straight up, the body will move down. Directional motion, away from the vertical, results when the GRF is applied at an angle. In order to move forward, backward, or sideways, the body must be able to create a force for the GRF to oppose. To move forward, it must push back against the ground and create the opposing GRF that will push the body forward. If the foot does not slip, Newton’s third law of motion would predict that the foot will experience a force equal and opposite to the force it is applying, resulting in forward motion (Watkins, 1983). In order for the foot not to slip at push-off and heel-strike, something must hold it against the ground as the force is applied. The frictional force creates a fixed point against which the body can push. Without the frictional force between the foot and the floor, the foot would slide over the floor and not be able to create a force (push) against the ground. Without a force against the ground, there would be no opposing GRF and no directional movement would be possible. It would still be possible to have movement in the vertical plane (i.e., a jump), but unless an external force were applied to the body while it was in the air, the COG would remain over the same spot on the floor. During the gait cycle, lack of frictional force causes the pedestrian to slip and, potentially, fall. At push-off, momentum and gravity are pulling the body forward as the COG moves in front of the support base. The trailing leg pushes off to begin the swing phase in an attempt to restore the support base under the COG. If the frictional force is insufficient to allow the foot to push off, the rear foot slips backward. However, the COG continues forward, usually resulting in a forward fall or stumble. At the other end of the gait cycle is heel-strike. The instability at this point is due to the very small contact area between the edge of the heel and the floor and the strike angle. At this point, the COG has already been restored to its position over the support base, and insufficient frictional force at heel-strike results in a split-like slip and fall, or a backward fall as the body tries to compensate for the slide. Combining all these factors, walking is accomplished by alternate pushes from the legs backward against the ground. Assuming that frictional force is sufficient to hold the foot firmly against the ground, these pushes create the GRF, which pushes back against the legs, causing forward motion. For any translational motion, the body must be able to create a force for the GRF to oppose; this oppositional force creates the motion. The


543

vertical forces (GRFs) are greatest just after heel-strike, when the body weight is loaded onto the leading foot, and again at push-off of the trailing foot. Between these two maxima is a force minimum, when the COG is at its highest point in the cycle and is directly over the supporting foot. The magnitude of the forces exerted by the foot depends primarily on body weight, step speed, and the speed of direction of movement of the COG. Energy of Walking The body tries to minimize energy expenditure of all its tasks, and human locomotion is no exception. For walking, this means minimizing the displacement of the COG. Rasch and Burke (1978) describe the body as having adapted five separate but coordinated mechanisms to flatten the curves in the vertical and horizontal planes, therefore minimizint the expenditure of energy: • Pelvic rotation. During walking, a theoretical line connecting the right and left hip bones swings counterclockwise and clockwise around its center point as the right and left feet are planted. This ability to rotate through the pelvis greatly reduces the vertical displacement of the COG. • Pelvic tilt. During normal walking, the pelvis tilts forward and down approximately 5° on the side of the non-weight-bearing leg. This tilt minimizes the amount of vertical displacement of the COG during the swing phase of the gait cycle. • Knee flexion. The knee of the support leg is bent (flexed) approximately 15° at the time at which the COG passes over it. This flexion prevents the COG from the need to rise. If the knee were locked, the COG would rise substantially during the stance phase. • Foot and knee mechanisms. At heel-strike, the foot is flexed up and the knee joint is fully extended. At this time, the COG is at its lowest point. As the body weight passes onto the forefoot and the heel is raised, the knee flexes so that the COG does not need to rise excessively. • Lateral displacement of the pelvis and shoulder. The hips and shoulders move up and down in the vertical plane; in other words, they do not remain level throughout the gait. The movement of the shoulders opposes the movement of the hips, which causes the COG to oscillate slightly from right to left. These five mechanisms maximize the efficiency of the body during walking and minimize the movement of the COG. This results in the least possible energy expenditure by the body. This conservation is demonstrated when one compares the ease with which one walks to the effort required to run a similar distance. Age has an effect on the energy of locomotion. At about age 65, walking appears to become more restrained. Studies in older adults show significant reduction in walking speed, a proportionately longer stance phase, and a shorter swing phase. This effectively shortens the stride length, which decreases the angle at heel-strike and decreases the potential for the foot to slip. In addition, older people of both sexes do not lift their feet as high during the swing phase, which reduces the amplitude of the oscillations of nearly all body parts, thereby maximizing energy conservation (Rasch and Burke, 1978). It does increase the likelihood of a trip, however, because even small displacements have the


544

potential to interrupt the swing phase. In fact, the majority of locomotion falls among the elderly are initiated by a trip. To conclude this section, it should be pointed out that walking is a much more complicated activity than previously imagined and is controlled by numerous physiological functions less obvious than simply moving the legs. Although many assume that walking is primarily a voluntary activity, in actuality most of the stability regulators are designed into the body in joints and muscles. Even in cases of age, infirmity, or abnormal physiology, the body works to minimize extraneous motion and thereby conserve energy. Throughout the gait cycle, the moments of least stability present the greatest potential for mishap. At heel-strike and push-off, the forces acting against the ground are at their maxima, and almost all slip-and-fall incidents occur at these two points. Insufficient frictional forces between the foot and the ground cause falls at both of these points of the walking cycle. If the foot is unable to plant at heel-strike or push-off, no opposing GRF can be present to brake the motion of the body (in heel-strike) or to push the body forward (in push-off). Information-Processing Model In order to discuss the role of perception and cognition in walking and pedestrian falls fully, the manner in which humans process the information they encounter when walking must also be covered. The steps taken by humans to process information about their relationship to the walking environment have been an active area of human factors research. As humans walk, they are constantly processing information regarding their perception of the walking environment and their interaction with it. Figure 19.2 presents a model of the information-processing system that depicts the major components of human information processing (Sanders and McCormick, 1996). These components are sensory processing, perception, memory, decision-making, attention, and response execution. Sensory Processing The first step for humans to process information is for the body to receive a stimulus. Of primary importance to pedestrians are the visual sense of the eyes, the tactile sense of the feet, and the kinesthetic senses of the body and limb position. Limitations on each sensory system can influence the quality and quantity of information that maybe registered initially and, potentially, all processes that follow. Detection of certain stimuli by a pedestrian may prevent a fall incident from occuring. Perception There are different levels of perception, with simple detection the most basic type. Detection is simply determining the presence of a target. Beyond simply detecting the presence of an object, perception involves identification and recognition of the object, which require the person to identify the class in which an item belongs. Perception of an


545

object depends on the individual’s prior experiences, memory, and associations that have been learned over time.

FIGURE 19.2 Information-processing model. Perception is one very important component of the study of fall incidents. Upon an initial visual scan of the area to be traversed, a pedestrian must be able to perceive a potentially hazardous condition in order to avoid contacting it. If a condition exists that makes it difficult to detect a hazard, the possibility of the pedestrian coming into contact with the hazard is greater. These conditions can include obstructions, confusing patterns, and other lack of sensory cues (predominantly visual) delineating the hazard. The typical pedestrian can see objects ahead and, to a lesser extent, to the sides. The cone of optimum viewing extends about 8 to 10 ft in front of the pedestrian. Objects that are off to the side or closer than 8 ft must be detected by peripheral vision. This usually requires the objects to have high color or pattern contrast, movement, or lighting in order to increase the chance of detection. Memory Memory comprises three components: sensory storage, working memory, and long-term memory. Sensory storage is the temporary storage mechanism for a stimulus. The sensory system associated with the visual system is known as iconic storage. The iconic storage holds an image for a short time, allowing further processing of the image. In order for it to be retained, the information must be encoded and transferred into working memory. In order to transfer the information from sensory storage to working memory, a person must direct his attention to the process. Information is transferred to long-term memory by supplying meaning to the information and relating it to information already stored in


546

long-term memory (semantically coding). To recall this information, it must be analyzed, compared, and related to past knowledge. The more organized the information is, the easier it is to transfer to long-term memory and the easier it is to retrieve. Decision-Making The key component in the information-processing model is the decision-making, which is a complex process in which individuals evaluate alternatives and select a course of action. The process involves seeking information relevant to the pressing decision, estimating probabilities of various outcomes, and attaching values to the expected outcomes. The decision-making process is limited by biases as well as the human capacity to process and evaluate information in order to arrive at an optimum decision. For example, once a pedestrian perceives (or remembers) a condition that he considers hazardous, he then has the option of deciding to select an alternate route or continuing the path while cautiously traversing the hazardous condition. The concept of uncertainty affects the decision-making process. As the number of choices increases, the required decision time and potential for error increase. In general, expectedness is a combination of certainty and experience. Therefore, if a pedestrian is uncertain about a particular walking area, has not perceived a hazard, and has no experience walking in that area, any hazard would be unexpected. Experience with a particular walking surface has three advantages: • It is easier to perceive certain important signs of impending hazard. • A greater number of possible evasive actions are stored in memory. • Understanding of probabilities and outcomes is better. Attention At the very top of the model is the pool of attention resources. Because the other processes (perception, decision-making, working memory, and response execution) draw from this pool, if some of these processes require more from it, less is available for the other processes, whose performance will deteriorate. Practice and learning will decrease the demand for the limited supply of resources from the pool. The four categories of attention are: • Selective: monitoring several sources of information to perform a single task • Focused: maintaining attention on one channel of information and ignoring the other sources • Divided: performing two or more separate tasks simultaneously and attention must be paid to both • Sustained: maintaining attention over prolonged periods of time in order to detect an infrequently occurring signal In order to avoid a fall, adequate and appropriate attention must be paid by the pedestrian to the walking path. If his attention is significantly diverted to other factors in the environment, the pedestrian may not perceive or remember the hazard and a fall may occur.


547

Response Execution After receiving and processing information, people use it to make responses, which can be in the form of motor responses executed by a person to achieve a goal. The motor responses can be influenced by internal and external information and can range from nearly automatic reflexes to which little to no conscious thought is given to voluntary responses in which all steps are consciously executed. In the model, the actual motor response, as influenced by internal and external information, is compared to some desired goal or set point. An error is considered any difference between the goal and the actual response. If sensory information is available during or after the motor response, it is referred to as feedback and can potentially be used to modify subsequent behavior. The Forensic Process in Pedestrian Fall Incidents The HF/E practitioner analyzes situations involved in litigation and offers opinions that may support or refute the legal theories pursued. Legal proceedings are an adversarial process in which lawyers represent opposing parties; in civil cases, they typically represent the plaintiff (injured party) or defendant (the alleged injuring party). The role of the forensic HF/E expert is to offer opinions in his field beyond the knowledge of the ordinary people comprising a jury. The process typically follows the process outlined in Figure 19.3: consultation with client, review of information, inspection of premises, research, formation of opinion, and expression of opinion (Cohen and LaRue, 2000). Consultation with Client The initial step in the forensic consulting process is discussion with the potential client, who usually is an attorney or insurance company representative. In this conversation, the client provides a description of the case and the issues involved. The forensic professional can confirm that the issues in the case are ones that fall within his area of expertise. The client will want to confirm that the expert has appropriate experience and is qualified to offer opinions with respect to the issues in the case. An agreement on fees is generally signed at this stage.


548

FIGURE 19.3 Chart of forensic steps. Review of Information Similar to any other scientific endeavor, the first true step is to review the existing information, which in the legal community is referred to as discovery. It is important for the forensic HF/E professional to review all pertinent discovery in order to understand the relevant background facts of the case. In a typical lawsuit, a variety of information is available regarding the plaintiff in the form of incident reports, statements, and medical records, as well as written and oral answers to legal questions in the form of interrogatories and depositions, respectively. Other important sources of information are witnesses to the incident, if any. Inspection of Site After reviewing the facts of the case, the HF/E professional will conduct an inspection of the location where the incident took place. Unless the site has been dramatically altered or destroyed, an inspection is necessary to gather as much firsthand data as possible. At this inspection, photographs and/or video documenting the scene should be undertaken. It is also important to measure dimensions and conduct testing appropriate to the type of incident. If a fall incident involves a stairway, then measurements regarding the width,


549

stair geometry (tread depths, riser heights, and nosings), and handrails must be gathered. Coefficient of friction (COF) tests are used to measure the slipperiness of a surface for slip cases. Other tests at the site can include illumination testing and slope measurements (Cohen and LaRue, 2002). Research and Formulation of Opinion It is generally beneficial to conduct research prior to developing a forensic HF/E opinion on a case. Research for pedestrian falls can include scientific literature, building codes, industry standards, profes-sional design guidelines, or even comparing how similar facilities have handled the design, construction, or maintenance issues at question. The forensic HF/E professional integrates all the information gathered from the discovery, inspection, and research with his knowledge and experience to form an opinion. The expert’s opinions are strengthened when he can provide a sound authority as the basis for the opinion. The types of bases generally used and the strength of each are discussed below. Expression of Opinion The field of forensic HF/E requires the practitioner to analyze situations involved in litigation and offer opinions regarding the HF/E principles involved. After the expert has formulated his opinions based on relevant case information, inspections, and research, an opinion may be rendered by several methods: reports, deposition, and testimony. An oral or written report is generally prepared for the client. The initial verbal report to the client should outline strengths and weaknesses of the case. It is important that both sides of the case are presented so that the client is aware of weaknesses to the case arguments. At this point, the client can decide if he is interested in pursuing the case further and the HF/E professional can determine if he wishes to be named as an expert in the case. A written report generally contains the results of the site inspection along with the opinions of the HF/E expert. A deposition is a formal question-and-answer period, with the same weight as courtroom testimony, in which the opposing attorney has the opportunity to question the designated expert witness under oath. The topics covered at the deposition usually focus on the expert’s qualifications (education and experience), his opinions regarding the case, and the bases for the opinions. The entire proceedings are typed by a court reporter and put into booklet form where they become part of the case discovery. If a case does not settle earlier in the legal process, it culminates with a trial before a judge and usually a jury (in the U.S.). At the trial, the expert witness will be sworn in and questioned by the retaining attorney (direct examination) and the attorney for the opposing side (cross-examination). The cross-examination is usually an attempt by the opposing attorney to impeach unfavorable testimony. Such questioning may be followed by redirect and recross, during which the attorneys are permitted to ask any desired clarifications to prior questions.


550

Basis for Human Factors Opinions The forensic HF/E professional will integrate the information gathered from the discovery material, inspection of the site, and applicable research with his knowledge and experience to formulate an opinion regarding causation of the fall and subsequent injuries. The expert’s opinion is strengthened when he can provide a sound authority as the basis of the opinion. These opinions may be based upon a variety of authorities, some providing stronger support than others. These bases are illustrated in a hierarchical manner in Figure 19.4, according to their strength and relative ease with which they can be supported (Cohen and LaRue, 1997). Codes and Laws Codes and laws are legal requirements governing a particular jurisdiction. This basis is used when the expert opines that the situation does not meet the provisions of an applicable code or law. When a clear violation exists, an expert may also need to provide an opinion that the violation has causal bearing on the incident. Examples of relevant codes or laws are the International Building Code (IBC) and the federal Americans with Disabilities Act (ADA). Voluntary Consensus Standards Voluntary consensus standards are a group of standards that have been developed by committees associated with standards-setting bodies through an industry consensus process. These standards are not legally binding, but are usually well known throughout the industry; they usually describe design and safety features for the respective industry.

FIGURE 19.4 Basis hierarchy.


551

Examples of these organizations are the American National Standards Institute (ANSI), the American Society for Testing and Materials (ASTM), and the International Standards Organization (ISO). Industry Custom and Practice The way of doing business over the years within a particular industry is considered the industry custom and practice. The custom and practice is typically based on experience and is generally informal; it may not even be documented. In some cases, however, the practices may be formalized as written company policies and procedures. Often, these practices may have evolved from some type of job analysis. These practices can be useful in forming an opinion, especially when evidence is available to support the practice and/or research. Professional Guidelines Most design-related professions have handbooks that offer criteria regarding appropriate and safe design. A citation from an authoritative handbook such as this may be appropriate. Examples commonly used by HF/E professionals include the Illuminating Engineering Society (IES) Lighting Handbook (Kaufman and Haynes, 1987), Woodson’s Human Factors Design Handbook (Woodson et al, 1992), and The Measure of Man and Woman by Henry Dreyfuss Associates (1993). The major disadvantage of citing such an authority as the only opinion basis is that the source may not be widely known, especially by the party deemed responsible, such as the building manager or owner. Scientific Literature Due to backgrounds as research scientists, many HF/E professionals feel most comfortable when citing scientific literature as the basis of their opinions. The primary drawback with using this type of literature as the sole basis of an opinion is that it is typically of very limited circulation. Consequently, citing HF/E scientific literature as an opinion basis may represent information readily available only to HF/ E professionals. Empirical Study Sometimes the specific issue is not significantly addressed by any documented authority and can only be addressed through original empirical research. In addition to the usual areas of possible impeachment of research studies (issues dealing with the appropriateness of experimental design, sample size, bias, confounding, and statistical power), this approach may not always be practical due to time and budget constraints. When feasible, however, research performed to answer specific issues not elsewhere addressed can be helpful in establishing a firm basis for forensic HF/E opinions.


552

Professional Judgment Sometimes, when no other foundational source is available, it becomes necessary for a forensic HF/E professional to base an opinion solely on his professional judgment. This opinion is more powerful when it is based on information extrapolated from previously encountered situations. Generally, the more similar the situations are to the current case, the more credible is the expert’s opinion. An expert’s opinion based solely on his judgment is only as powerful as his experience, credibility, and credentials. Such professional opinion is subject to attack, which may include impeachment of the expert’s credentials, experience, and testimony in similar cases. Consequently, a legitimate and sometimes necessary basis for expert opinion, professional judgment alone is the most difficult to support. 19.4 Pedestrian Fall Incidents and Case Examples When walking, pedestrians generally expect the walking surface to remain uniform. If a hazard exists in the walking surface, it must be perceived or recognized as a hazard by the pedestrian in order to decrease the likelihood of a fall incident. If the hazard is seen and recognized, the next stage is decision-making. The decision whether to avoid the hazardous condition, and, if so, how, is influenced by the pedestrian’s attitudes, behavior, and willingness to take risks. If the pedestrian decides to avoid the hazard, his success rate depends on his ability to do so. This ability is based on factors including anthropometry, biomechanics, and motor skills, along with other human characteristics and skills such as experience, training, and reaction time. Even with the best of intentions, after these steps are followed, random factors or chance can play a role in the process. HF/E principles assert that designs must take into account all sizes, shapes, limitations, and capabilities of the user population. Through learning, experience, and culture, pedestrians have certain expectations regarding walking surfaces, which are generally reinforced daily throughout their lives. When these expectancies are violated, the possibility for error and injury occurs. Many of the conditions that lead to falls are induced by design or situation. The chances of falls occurring due to these situations usually could have been reduced through relatively simple design, construction, or maintenance changes. Pedestrian falls commonly result from uncertainty in the presence of danger. In other words, these incidents occur when a pedestrian unexpectedly encounters a hazard in the walkway. Generally, the fall is initiated when the condition violates a pedestrian’s normal expectations or sufficient sensory cues (usually visual and/or tactile) are not available. The subsequent fall occurs when the pedestrian loses his balance irretrievably and adjustment of posture fails to enable him to remain on his feet. Simply put, designs and conditions that violate a pedestrian’s normal expectation or for which sensory cues are not sufficient to indicate a hazard tend to result in fall incidents. These situations can be categorized into four basic conditions of the walking surface:


553

• Unexpected change in traction of the walking surface • Unexpected change in level of the walking surface • Unexpected impediment in an otherwise level walking surface • Unexpected variation in stair geometry Each of these conditions presents a unique set of hazards and means for prevention. In the following sections, these four conditions are discussed with litigated cases provided as examples. Unexpected Change in Traction of Walking Surface Leading to a Slip When a pedestrian falls due to the loss of traction, he has slipped. These slips are the result of insufficient friction between footwear and the walking surface. A typical slip is shown in Figure 19.5. Slips generally

FIGURE 19.5 Slip diagram. occur at the points of greatest instability in the gait cycle: heel-strike and push-off. At these points, the entire body weight is concentrated on very small support bases: the heels and toes. When the heel touches the ground (heel-strike), the ground pushes back against the heel with GRF. This serves as a brake to the body’s forward momentum. If friction between the ground and the heel is insufficient, the GRF cannot push back against the heel and the foot slides forward. Typical slips include falling down into the split position as the front foot slides forward or falling backward onto the buttocks as the body overcompensates for the forward motion. During push-off, the toes push against the ground and the GRF pushes back against the toes. If traction is insufficient, the foot slides backward. A typical fall at this point in the walking process would result in a fall down into the splits or down onto one knee. It is generally the unexpectedness of the slippery condition that actually precipitates the person’s loss of balance and slip. If a pedestrian realizes that a walking surface is slippery (such as wet or icy), he can adjust his gait in order to traverse the slippery walking surface safely. Once the slippery condition is identified, the pedestrian can direct full attention to the task of walking across the slippery surface, carefully watching and placing each foot. He can take small, deliberate steps, maintain the center of gravity over the feet, and hold out the arms or use a provided handrail or other type of support for balance. By making these adjustments, it is possible to successfully maintain balance, or, if a slip does occur, the pedestrian is prepared for the possibility of the event so that he can relax, roll with the fall, and thus likely avoid serious injury.


554

The remedies for these conditions that lead to slips usually are quite simple. Premises should be designed with uniform, highly tractive flooring surfaces throughout. These premises should be maintained so that pedestrians do not need to be looking constantly for situations such as changes in flooring material, wet floors, or foreign substance such as water or other spills. Wet mopping of the floors should be done when pedestrians are not likely to be in the area; however, in any case, warning signs and cones should be placed in the vicinity to warn individuals of the change in traction. Case Study: Unexpected Change in Traction of Walking Surface Leading to a Slip In one such case, a pedestrian was walking down a carpeted hallway in an office building. As she rounded a corner, the flooring abruptly changed from low-pile commercial carpet into ceramic tile that had just been damp mopped (Figure 19.6). The pedestrian slipped and fell, breaking her right hip. Testimony from the plaintiff indicated that the change in traction between the carpeted and ceramic tile portions of the hallway was completely unexpected to her. The condition was made even more dangerous because the tile was wet from damp mopping. The smooth, nonporous texture of ceramic tile created a film of water on its surface. It was the opinion of the HF/E expert witness that the owners and management of the premises did not maintain a reasonable safety standard of care because they did not provide warning signs or safety cones that should have been used to alert a pedestrian to the presence of a slippery condition. Warning signs or safety cones would have helped the pedestrian perceive that the tile was wet so that she could have adjusted her gait to traverse the damp flooring successfully or avoid it altogether.

FIGURE 19.6 Photograph of change in walking surface traction.


555

HF/E principles dictate the following steps should have been taken in this order: 1. Eliminate the hazard. 2. Guard against access to the hazard. 3. Warn about the hazard (least preferable). Based on these principles, the HF/E expert’s primary opinion was to eliminate the unexpected change in traction by installing uniform flooring, preferably continuing the low-pile carpet throughout the hallway. Second, guarding against the slipping hazard could be accomplished by installing carpet runners with rubber-gripped backing on the ceramic tile portion of the floor or temporarily barricading it. In addition, a cleaning policy instructing the janitorial staff to wet mop only during the very late evening or early morning hours would ensure that the floor was not mopped when personnel were likely to be present. Finally, pedestrians could avoid unexpectedly encountering the wet tile if they were warned of potential slipping dangers with “Wet Floor” signs and safety cones delineating the entire area recently wet mopped. Case Study: Identifiable Change in Surface Traction of the Walking Surface In such a case, the plaintiff walked down a grocery store aisle and stepped on a grate that had been removed from the refrigerator case. The employee was crouching down, working in the dairy case with the shelving grate next to her. This grate was white, while the flooring was composed of darker tan vinyl tiles. In general, prudent pedestrians scan their walking pathway for possible impediments as they traverse the path. Once they start walking, people normally cast their gaze outward and downward approximately 8 to 10 ft in a forward direction. The grate clearly became noticeable at a distance greater than 15 ft (cone of vision). An inspection of the site indicated that the plaintiff, walking at a normal speed, would have had well over 8 sec to view the grate on the floor. Not only did the grate contrast in color with the tile, but it was also quite large (30.5 in. by 23.5 in.). In addition, the employee was crouched over adjacent to the grate, thus providing another visual cue more within the normal line of sight. The HF/E opinion indicated some anomalies in the reported dynamics of the fall. The plaintiff claimed a slip and then a fall forward; most typically, slips result in rearward falls. Second, she stated that both feet slipped on the grating, but this would be physically impossible. Finally, the grate sat 0.75 in. off the ground. Normal foot clearance does not generally exceed 3/8 in. (Cohen, 2000); therefore, the plaintiff would have had to lift her foot higher than what is typical in order to step on the grate. Most people would simply have kicked the grate forward midstride rather than unknowingly step on it.


556

FIGURE 19.7 Misstep diagram. Unexpected Change in Level of Walking Surface Leading to a Misstep Given no other visual cues, a pedestrian expects a level walking surface to remain level. Pedestrian missteps occur when some form of change in level of the walking surface catches the pedestrian unaware. Examples of this type of error-prone condition include the presence of a single riser down or an unexpected change in walking surface to a lower elevation. Typically, the pedestrian fully expects the walkway to remain perfectly level and then, without warning, the elevation of the walkway drops. A typical misstep is shown in Figure 19.7. Such missteps also occur when a pedestrian puts his foot down wrongly (steps down into a hole or on an angle), misses the step, or miscalculates a stride, often due to the misperception of depth when the walking surface drops unexpectedly. As the figure shows, the resulting fall from this type of misstep is usually forward, and can include a stumble of several steps prior to contacting the ground, wall, or other element in the environment. These single steps or other changes in elevation downward usually do not have visual or depth cues to draw the person’s attention to the elevation change. Typically, the flooring material pattern (e.g., carpeting, concrete, tile, or brick) remains the same and the pedestrian cannot distinguish the presence of the downward step. The remedy here is quite simple: eliminate changes in elevation whenever possible. Single risers should be eliminated if they are structurally unnecessary. If deemed necessary or already present and difficult or costly to eliminate, these level changes need to be made conspicuous to pedestrians. The conspicuity can be maximized by enhancing the contrast visibility of the step edge with the surroundings. This can be achieved by providing direct lighting or strip lighting on the step edges (lighting contrast), changing the color of the steps through the use of a different material or painting the step edge (color contrast), or providing abrasive strips or metal nosing strips to provide tactile (touch) cues. The area surrounding the drop in walking surface can also be altered to enhance, rather than detract from, the available depth cues indicative of an approaching change in elevation. For example, placing furniture or a plant near the bottom of the single riser can cue a person that the elevation is changing. Similarly, a handrail can provide support to a pedestrian and can also serve as a visual cue, more within a pedestrian’s normal line of sight, to a change in elevation.


557

Case Study: Unexpected Change in Elevation of Walking Surface Leading to a Misstep Such a case involved an elderly man who had been walking through a parking lot of an auto body shop and died from complications due to misstepping and stumbling forward when the walking surface unexpectedly dropped in level. The gentleman was walking on an asphalt parking lot when he unexpectedly stepped down onto an unpaved portion of the lot (Figure 19.8). He misstepped at the spot where the asphalt dropped in level between 1 and 4 in. He stumbled forward several feet and contacted a chain link fence with his head, causing fatal injuries.

FIGURE 19.8 Photograph of change in walking surface level. At this location, the parking lot was only partially paved, requiring drivers to park their cars straddling an irregular differential of between 1.5 and 4 in. in height (it has since been raised level with the ground). It was the opinion of the HF/E expert that the change in level and material was outside the decedent’s normal line of sight and focus of attention; therefore, the change was difficult to perceive, especially with a car parked directly next to his. Color and other environmental contrasts were not compelling enough to attract the attention of pedestrians. Additionally, people do not normally expect parking lots to have unmarked changes in surface elevations, especially at locations where persons normally would be expected to enter and exit their vehicles. Case Study: Foreseeable Change in Elevation of Walking Surface On some occasions, a HF/E expert will find that the change in a walking surface can be noticed in sufficient time to avoid a loss of balance and fall. One such case involved a plaintiff who was jogging in the early morning hours near a road under construction. Signs and barricades indicated that the road was closed to all traffic. Construction


558

equipment was parked on the site. The plaintiff saw the signs indicating that the road was closed, but instead of continuing on a designated alternative route away from the construction, he decided to jog past the signs and alternative path onto the construction site. As he was running on the site, he stepped down into a 6-foot-deep trench and suffered severe injuries. The view of the hole was unobstructed, well within the plaintiff’s normal line of sight, and could be clearly seen at a sufficient approach distance. The primary area of concern for the HF/E expert was whether or not the construction workers took adequate measures to secure the area. The expert opined that deleting the hole was unnecessary and unfeasible because it was in a clearly marked construction zone. However, the barricades were enough of an impediment that a reasonable person would have altered his behavior and taken the designated alternative route. The jogger took deliberate steps to defeat the barricade and ignore the signs. This behavior implied that the jogger understood the risks and decided to venture onto the job site despite the risks. Once on a construction site, the presence of holes and variations in the walking surface should have been expected. Unexpected Impediment on Walking Surface Leading to a Trip When a pedestrian unexpectedly encounters an impediment in the walking surface, typically a trip occurs. The pedestrian generally falls forward after the leading foot contacts the obstacle or obstruction. A typical tripping incident is shown in Figure 19.9. Trips and the resultant stumbles most often occur during the

FIGURE 19.9 Trip diagram. swing phase of the stride when the forward motion of the foot is halted unexpectedly. This can be a brief impediment (when the foot quickly manages to come free), which usually causes a slight, recoverable stumble, or a longer impediment, when the toe or heel actually becomes caught and a more serious fall results. In both cases, as the figure indicates, the fall and resulting stumble are almost always forward and can result in injuries to the hands, elbows, shoulders, head, or knees. The trips usually occur over impediments that are unmarked and provide no visual cue to alert the pedestrian to their presence. If the impediments were marked or expected, the pedestrian would perceive the existence of the impediment and take appropriate steps to avoid or safely traverse the condition. Examples of situations leading to trip incidents include: unmarked or deceptively painted wheel stops, speed bumps, and ramp edges;


559

potholes in streets; uplifts on sidewalks from tree root intrusion or other causes; or items dropped or placed in areas where they would otherwise not be expected. Research in HF/E and pedestrian safety, such as Cohen (2000), has indicated that uplifts as little as 3/8 in. are sufficient to present undetected and unexpected impediments to persons walking in a normal fashion. From 3/8 to 0.5 in. appears to be the tripping threshold for most pedestrians—subtle enough not to be perceived easily by a pedestrian scanning the walking surface, yet sufficient enough to catch a striding toe or heel. Once again, the remedy is proper design, construction, and maintenance; the design and construction should ensure no unmarked impediments are created and regular inspections and maintenance should ensure tripping hazards do not develop over time. When an impediment is discovered, the best fix would be to eliminate the hazard entirely. When this is not feasible or otherwise desirable, such as in the case of speed bumps and wheel stops, the potential tripping surface should be clearly and unambiguously painted, preferably in bright hazard yellow, to maximize its conspicuity to pedestrians. Indoor tripping hazards, if not possible to eliminate, need to be identified with a change in surface material, lighting, or other indication of the upcoming impediment. Case Study: Unexpected Impediment on Walking Surface Leading to a Trip An unexpected change in elevation involving an unmarked ramp created a hazard in an outdoor produce market. In this case, an elderly man tripped over the ramp edge, stumbled forward several feet, and hit his head on a steel door jam, rendering him a quadriplegic. The HF/E expert witness opined that the ramp edge was unexpectedly close to an infrequently used entrance and was therefore unexpected because it was never within the plaintiff’s field of view. The area was inadequately illuminated and there were no visual cues to indicate the presence of the ramp. There was no color contrast, lighting contrast, or textural change between the ramp and the flooring material. Furthermore, if the ramp edge had been feathered or at least barrier protected by a railing as called for in the building codes, the plaintiff’s injuries would have been prevented. Case Study: Visible Impediment on Walking Surface In some cases, a defect alone does not necessarily indicate the cause of a fall incident. In this case example, a shopper claimed she tripped over a slightly detached section of a retail store end cap display (Figure 19.10). The display was the type commonly used in the retail trades at the end of an aisle. Upon inspection of the store and location of the incident, the HF/E expert determined that the only way for a person’s striding foot to come in contact with the kick plate while making a left turn into a cross-aisle was to walk too close to the end cap and display. By walking this close, the end cap would have presented a fixed obstruction whether or not the kick plate was detached. Furthermore, if the kick plate were detached more than 1 in. (25 mm) as claimed, it should have offered less resistance to the striding foot, thereby preventing altogether or mitigating the plaintiff’s injuries. Also, a kick plate detached by this much would have been readily noticeable, contrasting in color with the tiled floor below.


560

FIGURE 19.10 Falls Due to Irregular Stair Geometry Extensive research over a long period of time has been done in order to determine the best configuration for the design and construction of stairs (see Templer, 1992). Pedestrians develop a specific gait when traversing stairs, based on the expectation that they are uniform in dimension. Typically, pedestrians on stairs do not look at where they are stepping except at the very beginning and end of the flight (Templer et al., 1985). Excessive variations in riser or tread dimensions not only go visually undetected in the initial scan by stairway users, but also disrupt their gait as well. These unexpected variations in geometry disrupt the rhythmic activity of successive foot placements, resulting typically in a loss of balance and, possibly, a fall. For these reasons, the building codes generally allow only up to 3/8-in. (0.95-cm) variation in successive riser heights and tread depths. Case Study: Unexpected Change in Stair Geometry Even seemingly slight variations in stair geometry can lead to serious injuries. In one such case, a spectator was walking down a set of stairs at the auto racetrack when he suddenly misstepped and fell forward. A site inspection revealed that the stairway was composed of 17 prefabricated metal risers with two site-made wooden steps transitioning the metal steps to the ground (Figure 19.11). Each of the risers measured a uniform 7.5 in. (19.05 cm), except for the wooden steps. The riser at the transition from metal to wood measured 6.25 to 6.5 in. (15.8 to 16.5 cm), depending on where the measurement


561

was taken. This was a 1.25- to 1.5-in. (3.2- to 3.8-cm) differential from the previous 17 risers. The HF/E expert determined that this excessive variation of about 1 in. (2.54 cm) caused the pedestrian to misperceive the height between steps, disrupting his gait and thus causing him to fall. In addition to this variation violating the building code requirement for uniformity of riser heights, the wooden steps were poorly maintained and without any nonskid material applied.

FIGURE 19.11 Photograph of nonuniform stair geometry. No case study to date could be identified in which an unexpected change in stair geometry could be defended. However, if a pedestrian decides to descend or ascend a stairway that looks like it is very old and in obvious need of repair, he should expect some irregularity in stair dimensions and be prepared by holding the handrail, watching every step closely, and carefully placing each foot. By perceiving and expecting uneven conditions to exist, a pedestrian could safely traverse even a stairway with gross dimensional variations. In conclusion, the forensic HF/E professional can explain a pedestrian fall in terms of the perceptual, cognitive, and biomechanical abilities and limitations of humans. This scientific information, in conjunction with an analysis of the physical evidence and available testimony, is invaluable in determining for the trier of fact the causes of and means for prevention of pedestrian falls.


562

19.5 Checklist of Main Factors to Consider in Pedestrian Fall Incidents Primary Factor Considerations Pedestrian expectation

Experience with location Memory of hazard Conspicuity of hazard Warning signs and other sensory cues Pedestrian physical condition Age Eyesight Corrective lenses Mobility or other impairment Pregnancy Weight Pedestrian behavior Talking or other task distractions Visual distractions Alcohol or drug impairment Reasonable conduct

Primary Factor Illumination

Indoor level walking surfaces

Outdoor surfaces

Stairs

Single stair

Considerations Adequate lighting/reflectance Shadows Glare Provide contrast for change in level Conformance to applicable building codes Unmarked rises in elevation (single step up) Unmarked drop in elevation (single step down) Physical obstacles Condition of floor surface Coefficient of friction Appropriate and secured carpeting Adequate illumination Appropriate drainage Slip resistance Weather conditions Unmarked rise in elevation (single step, wheel stop, curb) Unmarked drop in elevation (single step down, pothole) Physical obstacle Uniform stair geometry (treads, risers, nosings) Conformance to applicable building codes Proper handrail dimensions Slip-resistant treads Adequate illumination Discernible step edges Possible removal


Ramps

Walking surface traction

Handrails

Maintenance

563

Conspicuity Contrast in color/pattern/material Contrast in lighting (lighting strips) Slip-resistant surface Conformance to applicable building codes Proper ramp geometry Appropriate landings Adequate edge protection Adequate illumination Proper handrail dimensions Coefficient of friction Unexpected change in traction Conspicuous slippery condition Warnings of slippery condition Graspability Firmly attached Height Adequate number Proper extension beyond stairs or ramp Provide appropriate visual cue for change in level Documented maintenance policies and procedures Inspection schedule Documented inspections Matting in appropriate areas Cones and/or barricades used at appropriate times Appropriate floor-finishing products

Defining Terms Coefficient of friction—The degree of traction between the shoe heel/sole and walking surface material; an indicator for slipperiness of the walking surface. It is often abbreviated as COF. Misstep—Loss of balance due to unexpected change in downward level of a walking surface or presence of a hole/dip. Nosing—The front and usually rounded edge of a stair tread; it frequently projects over the riser below it. Riser—The upright (vertical) face of a step. Riser height—The height of one step measured from the top of a step at the nosing to the top of an adjoining step at the nosing. Slip—Loss of balance due to unexpected change in surface traction. Slip resistant—The ability to provide adequate force to resist the tendency of the shoe or foot to slide along a walking surface. Slip resistance is related to a combination of factors including the walkway surface, the footwear heel/sole, and the presence of any foreign material between them.


564

Stairway—A system of two or more risers. Tread—The horizontal surface of a step. Tread depth—The length (from back to front) of a tread as measured from the furthermost extension of adjacent treads. Trip—Loss of balance due to an unexpected change in upward level of a walking surface or presence of an impediment. References Cohen, H.H., 2000, A field study of stair descent, Ergonomics in Design , 8(2), 11–15. Cohen, H.H. and LaRue, C.A., 1997, Advice from an expert witness, Ergonomics in Design , 5(4), 19–24. Cohen, H.H. and LaRue, C.A., 2000, Forensic human factors/ergonomics. Chapter in International Encyclopedia of Ergonomics and Human Factors (London: Taylor & Francis). Cohen, H.H. and LaRue, C.A., 2002, Establishing standards of reasonable care through forensic human factors, Ergonomics in Design , 10(2), 15–18. Henry Dreyfuss Associates, 1993, The Measure of Man and Woman: Human Factors in Design (New York: Whitney Library of Design). Kaufman, J.E. and Haynes, H. (Eds.), 1987, IES Lighting Handbook: Application Volume (New York: Illuminating Engineering Society of North America). Rasch, P.J. and Burke, R.K., 1978, Kinesiology and Applied Anatomy: the Science of Human Movement (6th ed.) (New York: Lea and Febiger). Sanders, M.S. and McCormick, E.J., 1996, Human Factors in Engineering and Design (7th ed.) (New York: McGraw-Hill, Inc.). Templer, J., 1992, The Staircase: Studies of Hazards, Falls, and Safer Design (Cambridge, MA: Massachusetts Institute of Technology Press). Templer, J., Archea, J., and Cohen, H.H., 1985, Study of factors associated with risks of stairway falls, J. Saf. Res. , 16, 183–196. Watkins, J., 1983, An Introduction to Mechanics of Human Movement (New York: MTP Press). Woodson, W.E., Tillman, B., and Tillman, P.L., 1992, Human Factors Design Handbook (2nd ed.) (New York: McGraw-Hill, Inc.).

Further Information Amblard, B., Berthoz, A., and Clarac, F., Eds., 1988, Posture and gait, Proceedings of the 9th International Symposium on Postural and Gait Research (New York: Elsevier Science Publishers, B.V.). Carlson, S., 1972, How Man Moves: Kinesiological Studies and Methods (New York: William Heinemann Ltd.). Cohen, H.H., 1998, Performance based analysis of 60 stairway fall accidents, Proc. Pacific Rim Conf. Building Officials , 9–19. (Maui, HI: International Conference of Building Code Officials) (update in press). Ducroquet, R., Ducroquet, J., and Ducroquet, P., 1968, Walking and Limping: a Study of Normal and Pathological Walking (Philadelphia: J.B. Lippincott Co.). Hyde, A.S., Bakken, G.M., Abele, J.R., Cohen, H.H., and LaRue, C.A., 2002, Falls and Related Injuries: Slips, Trips, Missteps and Their Consequences (Tucson, AZ: Lawyers and Judges Publishing Company, Inc.). Inman, V.T., Ralston, H.J., and Todd, F., 1981, Human Walking (New York: Williams & Wilkins).


565

International Conference of Building Officials, 1997, Uniform Building Code (Whittier, CA: Author). Keeton, W.P. (Ed.), 1984, Prosser and Keeton on the Law of Torts (5th ed.) (St. Paul, MN: West Publishing Co.). McConnill, M.A. and Basmajian, J.V., 1977, Muscles and Movement—a Basis for Human Kinesiology (New York: Robert E. Krieger Publishing Co., Inc.). Schafer, R.C., 1955, Clinical Biomechanics: Musculoskeletal Actions and Reactions (2nd ed.) (New York: Williams & Wilkins). Steindler, A., 1955, Kinesiology of the Human Body under Normal and Pathological Conditions (New York: Charles C. Thomas Publisher). United States Equal Employment Opportunity Commission and the United States Department of Justice, 1992, Americans with Disabilities Act Handbook 59–0 (Washington, D.C.: Authors). Wickens, C.D., 1992, Engineering Psychology and Human Performance (2nd ed.) (New York: HarperCollins Publishers). Zohar, D., 1978, Why do we bump into things while walking? Hum. Factors , 20(6), 671–679.

20 Measurement in Pedestrian Falls Daniel A.Johnson Daniel A.Johnson, Inc.

0–415–28870–3/05/$0.00+$1.50 © 2005 by CRC Press

20.1 Introduction Just as measurement is the basis of all science, accurate and precise measurement is crucial in the evaluation of environmental factors to determine if they have contributed to a pedestrian’s fall. This chapter considers different methods of measuring these factors. Falls cause a tremendous number of injuries and are the leading cause of accidental death for older citizens. The financial effects are phenomenal. In the United States, in 1990, falls cost more than $50 billion. The number of fall injuries was second only to those suffered in automobile collisions (Pauls, 1991). Falls not only have a devastating effect on the injured person and that person’s family, but also can result in expensive lawsuits. Because of this, forensic ergonomists are often asked to investigate the circumstances surrounding a fall. To do this adequately requires taking accurate measurements and comparing them with relevant legal codes and standards that have a basis in ergonomic research. I am unaware of any source that actually describes how to take measurements of these environmental factors accurately. The one exception involves slips, for which considerable effort has been expended in the design and evaluation of equipment to measure the slip resistance between the shoe and a walking surface. A number of books and standards have been promulgated with regard to this equipment (e.g., English, 1996; Bakken, 2002). Various organizations have published standards, texts, and legal codes that recommend or require that the built environment be designed and maintained so that it provides some minimum level of safety for pedestrians. Often these standards and codes call out for stairways and level and sloped walkways to meet certain specific requirements such as the vertical rise of a stair being no greater than 178 mm (7 in.). Other standards call for qualitative values, such as the one requiring walkways to be “slip resistant,” without defining what that means. However, none of the standards or codes describes the method by which the specified dimensions are to be measured. With the exception of slip resistance, how to measure most environmental factors that can cause a fall has not been treated in depth in the ergonomic literature. A goal of this chapter is to examine the issue of measurement as it is related to pedestrian falls and to suggest some optimal methods of performing these measurements.


568

20.2 Importance of Accurate Measurements From a forensic point of view—from a liability point of view—it is important to achieve a high level of accuracy in measuring environmental factors that might relate to a fall. A fall can result when a walkway does not meet some minimum safety standard. Alternatively, a fall may occur because the person was, in some way, predisposed to falling, was inattentive, or was behaving in a manner that increased the likelihood of falling. In the first situation, the Court is more likely to conclude that the fall probably would not have occurred if the walkway had met the standards. In this case, those in charge of that walkway may be held liable for the consequences. In the other situations, in which the walkway did meet standards and the fall occurred, perhaps, because of some behavioral or physiological reason associated with the person who fell, those in charge of that walkway would probably not be held liable. In either case, accurate measurements can assist the Court in determining responsibility for the fall. In yet another set of instances, when falls occur but there are no codes or standards by which a court can judge liability, the forensic ergonomist may be asked to give an opinion as to whether the plaintiff was or was not behaving in an expected or predictable manner. In addition, the forensic ergonomist may be asked if the environmental condition faced by the pedestrian was inappropriate or unexpected. Regardless of the cause of the fall, the ergonomist will need to take accurate measurements of the site where the fall occurred. 20.3 Types of Falls Three basic types of fall are addressed here: trip and falls, slip and falls, and falls on stairs. About 30% of the cases on which I have been consulted have involved slips, 30% have been falls on stairs, 20% have been due to trips, and 7% have been falls on ramps. The remainder includes falls from heights; falls into holes or openings; and falls from equipment, such as ladders and scaffolds. These latter categories, although of concern to the practicing forensic ergonomist, are not addressed here. 20.4 Phases of the Walk Cycle I consider the walk cycle on a flat surface to consist of four phases: the stance phase, toeoff phase, swing phase, and heel-strike phase. On stairs, toe-strike may replace heel-strike in descent as well as ascent. 20.5 Causes of Falls Unintentional falls are nearly always preceded by a misstep, defined here as an unintentional departure from pedestrian gait appropriate for the walkway surface. A misstep can be due to a number of factors such as being jostled by others, a physiological

Measurement in pedestrian falls

569

condition, perceptual inadequacy (e.g., floor covering materials that camouflaged tread nosings), and inattention. Missteps can also result from poor design or inadequate lighting. 20.6 Trip Hazards The most common method by which an object on the ground will trip a person occurs when the foot in the swing phase is stopped short in its forward movement. The swing phase of the walking cycle is initiated on toe-off when the toes are pointing in a downward direction (Jahss, 1982). As the foot travels, forward movement of the knee and dorsiflection cause the toes and foot to start to point forward. At heel-strike, the toes point in a slightly upward direction. However, approximately midway during the swing phase, the toes describe an arc that comes close to touching the surface of the ground. During this phase, an unexpected object rising above the surface can stop the foot’s movement. If this happens, the momentum of the body will carry its center of gravity forward of the ball of the foot in the stance phase, and a stumble or a fall will result (Bakken, 2002). Height of Trip Hazards A vertical rise of as little as 13 mm (0.5 in.) can stop a foot in flight because the toe may come this close to the surface during swing phase (Cohen, 2000). The U.S. Housing and Urban Development (HUD) rules for routes that are expected to be used by those with difficulty walking have specific requirements: 4.5.2 Changes in Level. Changes in level up to 1/4 in. (6 mm) may be vertical and without edge treatment. Changes in level between 1/4 and 1/2 in. (6 and 13 mm) shall be beveled with a slope no greater than 1:2. Changes in level greater than 1/2 in. (13 mm) shall be accomplished by means of a ramp (CFR 24 1995). Similar requirements for safe walkway surfaces have been put forth in two additional documents. American National Standards Institute (ANSI) document ANSI A117.1 (Accessible and Usable Buildings and Facilities) addresses walkways where people with some mobility impairment might be expected. Virtually the same requirements were published in a document published by the American Society for Testing and Materials (ASTM), entitled ASTM F 1637, Standard Practice for Safe Walking Surfaces. This document, however, was not written for a special group of individuals (i.e., handicapped people) but for all people. In fact, it states that its provisions may not be adequate for those with handicaps. Furthermore, as Bakken reports, the International Building Code (IBC), and its predecessor, the Uniform Building Code (UBC), have adopted the entire text of ANSI A117.1 and have stipulated that those requirements are applicable to all individuals. In order to comply with the IBC and UBC, a walkway must also comply with ANSI A117.1. For an excellent review of building codes and standards, and how they relate to pedestrian falls, see Bakken (2002).


570

Thus, it is important not only to measure the height of a suspected trip hazard accurately, but also, if it is over about 6 mm (0.25 in.) and less than about 13 mm (0.5 in.), to measure the beveled edge in order to determine if the slope exceeds 1:2, or 26º. Typical Way of Measuring Trip Hazards In the field, it is not always easy to take accurate and precise measurements of small potential trip hazards using such standard tools as rulers or tape measures. One practical factor that makes it difficult to read a ruler accurately (“ruler” includes tape measures, unless otherwise specified) is that of placing one’s eye close enough to and parallel with the surface to overcome parallax distortion. No accurate measurement of the possible trip hazard can be made if the hazard and the ruler are not in the same plane, parallel to the eye, and at the same distance from the eye. Another practical problem occurs when a fall occurred on a broken and unevenly elevated sidewalk. In this instance, the lifted section, where the plaintiff tripped, might be close enough to 13 mm (0.5 in.) so that it is unclear whether grounds for a lawsuit exist. For example, the height may be about 13 mm (0.5 in.) but the top portion is beveled. The issue, then, is how to take accurate measurements of the rise height and the bevel. Doing so using only a ruler and straightedge is unlikely. Probably even more difficult to measure, using only a ruler, are small but potentially dangerous obstructions on the nosings of some stairways. Cohen looked at whether a raised nail used to fasten a slip-resistant strip to a nosing (the foremost edge of where the stair tread can be) could catch a heel in flight (Cohen, 2000). He found that the heels of six male volunteers cleared the intermediate nosing by an average of 18 mm (0.72 in.) with a range of 5 to 58 mm (0.2 to 2.3 in.). A raised portion on the nosing that protruded more than about 5 mm (0.2 in.) could be considered a trip hazard. Optimal Way of Measuring Trip Hazards A preferable method makes use of a carpenter’s profiling tool—a device with a row of about 180 small stiff wires, each with a diameter of about 0.8 mm (0.3 in.) and a density of about 12 per cm (30 per in.). The typical device is 15 cm (6 in.) long. In-house lab tests show that the tool can provide measurements of vertical risers that are as small as 1 mm (0.04 in.) and as large as 38 mm (1.5 in.). When the wires are pushed down onto a surface such as the sharp rise in a sidewalk, they hold the shape. The profile of the surface can then be traced onto paper and scanned into a computer (see Figure 20.1 through Figure 20.4). A graphics program such as Adobe Illustrator® can then be used to measure the heights and angles of the surfaces (Adobe Systems). In the instance depicted in Figure 20.4, the overall rise was 14 mm (0.54 in.) and the bevel was steeper than 26° for the first 10 mm (0.4 in.), thereby exceeding the limts allowed by HUD. The conclusion from this analysis is that the surface presented a trip hazard.

Measurement in pedestrian falls

571

Profile Tool Accuracy and Reliability A small metal box, which sat flush on a desktop surface, was used to simulate a vertical rise in a walkway. The height of this box was measured at 32.1 mm (1.26 in.) using a caliper. This provided an external check on the accuracy of the profile tool. The fact that it was considerably larger than 13 mm (0.5 in.) was immaterial because the variation in measurements was of interest, not the absolute height. Measurements were performed by arranging the profile tool so that about half of it was over the top of the box and the other half was over the surface of the desk. The sliding pin elements were pushed down until they were in contact with one of the surfaces. The profile was traced onto paper with pencil. The tool was then reset so that all of the pin elements were again parallel, and another measurement was taken. Ten separate profiles were taken, scanned into a computer, and imported into Adobe Illustrator®. Each profile was examined at magnification of up to 800%. Straight lines were scribed through the center of the pencil marks. The “measuring tool” in the Adobe Illustrator® application was used to determine the vertical distance between the upper and lower levels (the top of the box and the desk). At this magnification, the measuring tool reads out to an accuracy of ±41×10− 3 mm (1.6×10− 3 in.). The mean (sd) of the 10 measures indicated that the height of the box above the surface was 32.0 mm (0.14). That is, its height was found to be 1.26 in. (6×10− 3 in.). The CI95 was 32.0 mm ±0.1 (1.26 in. ±0.01), which was consistent with the caliper-measured height of 32.1 mm (1.26 in.). With this potential level of accuracy, the profile tool should allow an investigator to decide if a rise was significantly (p200 Assembly: 28 cm 25.4 cm (10 in) Domestic sq.m: 24 cm (11 in) All others: 25 cm (9.5 in) Prim, school: (9.8 in) 26 cm (10.2 in) High school, etc.: 26 cm (10.2 in) Other facilities: 21 cm (8.3 in) Handrail 86.5–96.5 cm Not ~80 cm Not specified 90–100 cm (35.4– 86.4–96.5 height (34–38 in) specified (~31.5 in) 40 in) cm (34–38 in) Not specified Not specified 3–5 cm (1.2–2 in) 3.2–5.7 cm Handrail 3–5 cm (1.2–2 Not specified (1.25–2.25 diameter in) in) Handrail 4 cm (1.6 in) Not 3 cm (1.2 in) Not specified Not specified 3.3 cm (1.5 from wall specified in) Minimum 10 lux Not Not specified Not specified 100 lux 11 lux

Handbook of Human factors in litigation lighting on specified stairs External, 20 lux e.g., for bridge For subway 100 lux Landing Same as stair Not 120 cm (47 width width up to specified in) (If ht. 110 cm (43 in) stairs>3m) Maximum ramp slope Trip hazard standards? Trip hazard height Slipresistant surface? Slip resistant criteria:

1:6 (9.5°) or less No

10% (4.5°) No

598

(suggested) 11 lux

120 cm (47 in) Not specified (If ht. stairs>3m)

Same as stair width up to 121.9 cm (48 in)

Handbook of human factors in litigation

Handbook of Human Factors in Litigation

Handbook of Human Factors in Web Design

Handbook of human factors and ergonomics methods

Handbook of Human Factors and Ergonomics Methods

Human Factors Interaction Theories in Consumer Product Design (Handbook of Human Factors in Consumer Product Design)

The Handbook of Human Factors in Web Design

Handbook of Human Factors in Medical Device Design

The Handbook of Human Factors in Web Design

Human Factors in Alarm Design

Human Factors In Consumer Products

Human Factors in the Training of Pilots

Human Factors and Ergonomics in Consumer Product Design: Methods and Techniques (Handbook of Human Factors in Consumer Product Design)

Human Factors for Engineers

Macrocognition in Teams (Human Factors in Defence)

Handbook of Cognitive Task Design (Human Factors and Ergonomics)

A Handbook of Transcription Factors

Notfallmanagement: Human Factors in der Akutmedizin

Human Factors in Aviation, 2nd Edition

Human Factors in Lighting, Second Edition

Human Factors in Simulation and Training

Conceptual foundations of human factors measurement

The history of human factors and ergonomics

Aviation Psychology and Human Factors

Ten Questions About Human Error: A New View of Human Factors and System Safety (Human Factors in Transportation)

The Human Factors of Transport Signs

The Universal Access Handbook (Human Factors and Ergonomics)

Ten Questions About Human Error: A New View of Human Factors and System Safety (Human Factors in Transportation)

Handbook of Human Immunology

Corporations and Transnational Human Rights Litigation

Human Factors of Visual and Cognitive Performance in Driving

Handbook of human factors in litigation