Emotion in Memory and Development: Biological, Cognitive, and Social Considerations

Emotion and Memory in Development

Series in Affective Science Series Editors RICHARD J. DAVIDSON and KLAUS R. SCHERER The series was founded in 1982 by Paul Ekman and Klaus R. Scherer. The Nature of Emotion: Fundamental Questions Edited by Paul Ekman and Richard J. Davidson Boo! Culture, Experience, and the Startle Reflex Ronald Simons Emotions in Psychopathology: Theory and Research Edited by William F. Flack Jr. and James D. Laird What the Face Reveals: Basic and Applied Studies of Spontaneous Expression Using the Facial Action Coding System (FACS) Edited by Paul Ekman and Erika Rosenberg

Appraisal Processes in Emotion: Theory, Methods, Research Edited by Klaus R. Scherer, Angela Schorr, and Tom Johnstone Music and Emotion: Theory and Research Edited by Patrik N. Juslin and John A. Sloboda Nonverbal Behavior in Clinical Settings Edited by Pierre Philippot, Robert S. Feldman, and Erik J. Coats Memory and Emotion Edited by Daniel Reisberg and Paula Hertel Psychology of Gratitude Edited by Robert A. Emmons and Michael E. McCullough

Shame: Interpersonal Behavior, Psychopathology, and Culture Edited by Paul Gilbert and Bernice Andrews

Thinking about Feeling: Contemporary Philosophers on Emotions Edited by Robert C. Solomon

Affective Neuroscience: The Foundations of Human and Animal Emotions Jaak Panksepp

Bodily Sensibility: Intelligent Action Edited by Jay Schulkin

Extreme Fear, Shyness, and Social Phobia: Origins, Biological Mechanisms, and Clinical Outcomes Edited by Louis A. Schmidt and Jay Schulkin Cognitive Neuroscience of Emotion Edited by Richard D. Lane and Lynn Nadel The Neuropsychology of Emotion Edited by Joan C. Borod Anxiety, Depression, and Emotion Edited by Richard J. Davidson Persons, Situations, and Emotions: An Ecological Approach Edited by Hermann Brandstätter and Andrzej Eliasz Emotion, Social Relationships, and Health Edited by Carol D. Ryff and Burton Singer

Who Needs Emotions? The Brain Meets the Robot Edited by Jean-Marc Fellous and Michael A. Arbib What the Face Reveals: Basic and Applied Studies of Spontaneous Expression Using the Facial Action Coding System (FACS), Second Edition Edited by Paul Ekman and Erika L. Rosenberg The Development of Social Engagement: Neurobiological Perspectives Edited by Peter J. Marshall and Nathan A. Fox Handbook of Emotion Elicitation and Assessment Edited by James A. Coan and John J. B. Allen Emotion and Memory in Development: Biological, Cognitive, and Social Considerations Edited by Jodi A. Quas and Robyn Fivush

Emotion and Memory in Development Biological, Cognitive, and Social Considerations

Edited by Jodi A. Quas Robyn Fivush

1 2009

1 Oxford University Press, Inc., publishes works that further Oxford University’s objective of excellence in research, scholarship, and education. Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam

Copyright © 2009 by Oxford University Press, Inc. Published by Oxford University Press, Inc. 198 Madison Avenue, New York, New York 10016 www.oup.com Oxford is a registered trademark of Oxford University Press. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of Oxford University Press. Library of Congress Cataloging-in-Publication Data Emotion and memory in development : biological, cognitive, and social considerations / edited by Jodi A. Quas and Robyn Fivush. p. cm.—(Series in affective science ; 24) Includes bibliographical references and index. ISBN 978-0-19-532693-2 1 Memory in children 2 Emotions in children 3 Child psychology I Quas, Jodi A II Fivush, Robyn BF723.M4E46 2009 155.4′1312—dc22 2008043029

9 8 7 6 5 4 3 2 1 Printed in the United States of America on acid-free paper

Preface Jodi A. Quas, University of California, Irvine Robyn Fivush, Emory University

T

he question of how well children can recall and discuss emotional experiences is one that has received considerable attention, not only in scientific research, but also in clinical contexts, in the legal system, and by the public. Knowledge concerning the role of emotions generally and emotional distress specifically in children’s emerging cognitive abilities has implications for understanding how children attend to and process information, how children react to emotional information, and how that information affects their development and functioning over time. Of practical importance, increasing numbers of children have been involved in legal settings as victims or witnesses to violence, highlighting the need to determine the extent to which children’s eyewitness reports of traumatic experiences are accurate and complete. Also, in educational contexts, children may be asked to perform under conditions of heightened emotional arousal (e.g., when they are spontaneously called upon to recount their knowledge in front of classmates). The emotion that results from teacher and peer evaluations has implications for children’s performance and perceived competence. Finally, in clinical contexts, the ability to recount emotional events is emerging as a significant predictor of psychological outcomes. How children learn to describe emotional experiences and the extent to which they can do so coherently may play a critical role in their well-being. Clinicians must be aware of children’s emerging ability to describe stressors in order to develop and implement appropriate interventions. Given the broad range of implications that relate to understanding how emotions affect children’s memory, it is not surprising that a considerable amount of theorizing, research, and writing has been devoted to the topic of when and how emotions, particularly emotional distress and trauma, influence children’s memory. In 2006, we (Fivush and Quas) decided that it was time to bring together leading scientists conducting this research to discuss the current state of knowledge in this field. We thus held a small conference, supported by Emory University through the Emory Cognition Project, focused on emotion and memory v

in development. At the conference, more than a dozen top scientists in the United States and abroad whose work was relevant to the study of emotion and memory came together. We discussed our research, integrated findings, shared ideas, and elaborated on the theoretical and practical significance of the evidence to date. What emerged was not only a broader understanding of the state of the field, but also the clear need to incorporate findings from numerous related fields as we launch the next wave of research in this important field of inquiry. The current volume represents the outcome of the conference. Scientists expanded upon their ideas to address broad questions concerning how emotions and stress affect memory across development. One key question addressed by the volume’s contributors concerns why children react so differently to emotional experiences and subsequently come to understand those experiences differently. Contributors addressed this question by studying sources of individual differences that influence children’s emotional reactions, coping, and emotional understanding (Compas, Campbell, Robinson, & Rodriguez; Laible & Panfile), narrative processes (Oppenheim & Koren-Karie; Sales), and event memory (Baker-Ward, Ornstein, & Starnes; Chae, Ogle, & Goodman; Peterson & Warren). Of importance, sources range from those at the individual level (e.g., biological) level to those at the social-contextual level (e.g., parentchild discourse, attachment relationships, living with a chronic illness, living in an abusive environment) in which children’s lives are embedded. Thus, consistent with models of studying developmental phenomena at all levels of analysis, or “from neurons to neighborhoods” (Alexander & O’Hara), scientists interested in children’s memory for emotional events must consider multiple sources of influence across different levels. A second key question that appears across numerous chapters in this volume focuses on the precise emotional nature of the stressful event that children are remembering, or at least being asked to remember. That is, researchers have conceptualized and studied “emotion” and “stress” in different ways. For instance, some researchers who study emotion and memory investigate how well children remember discrete arousing or fear-eliciting events (e.g., tornados, invasive medical procedures) (Chae, Ogle, & Goodman; Peterson & Warren). Other researchers examine how exposure to chronic stress affects children’s memory functioning, including their general autobiographical memory (Greenhoot, Johnson, Legerski & McCloskey; Wiik & Gunnar), and the neurological structures underlying memory functioning (Carver & Cluver). A third approach, often taken by researchers interested in children’s physiological responses, attachment, or parent-child communication, involves investigating children’s and families’ reactions to and narratives about mildly arousing negative events (e.g., interpersonal conflicts, laboratory and naturalistic challenges, and/or brief encounters that evoke anger, sadness, or disgust) (Laible & Panfile; Sales; Oppenheim & Koren-Karie). Finally, in the neurobiological field, researchers interested in stress have studied the effects of stress on neurological and physiological processes linked to memory encoding, consolidation, and retrieval (Carver & Culver; Wallin, Quas, & Yim; and Wiik & Gunnar). The different foci have implications for evaluating children’s interpretations of and responses during vi

PREFACE

particular events and their later ability to recount those events accurately and completely (Bauer; Salmon & Conroy). As such, it is imperative to merge findings across the different conceptualizations and more clearly articulate the specific topics being studied to better understand the role of emotions broadly and distress and trauma specifically in children’s memory. A final key question raised by many of the contributors concerns how memories of emotional, especially traumatic, events are similar to and different from memories of everyday or mundane events (Chae, Ogle, & Goodman). Answers to this question remain elusive, especially in the developmental literature. One reason for this uncertainty stems from whether research focuses on group or individual differences. That is, when children recalling an emotional event (e.g., an inoculation) are compared to children recalling a nonemotional event (e.g., a well-child visit), emotional events are almost always recalled better than mundane events. However, when recall among children who have all experienced the same emotional event is examined, huge individual differences emerge in children’s understanding of the event, the magnitude of their emotional reactions, and the extent and accuracy of children’s memory (Baker-Ward, Ornstein, & Starnes). Thus, it is imperative to look beyond the general question “What are the overall effects of emotion on memory?” and instead ask “Why do individual children react differently to particular events and how (and when) do these reactions affect their later narrative processes and memory?” (Bauer; Fivush; Thompson).

Individual Chapter Overview Our volume is heuristically organized into four parts, although as will be evident quite quickly, the ideas presented overlap across the parts. However, the topics are often approached from different perspectives across the chapters, and, by taking these perspectives into account, novel insights into individual children’s memory for emotional events can be gleaned. In Part I, leading researchers who have studied emotion, primarily distress, and memory describe the current state of the field. Three research teams (Baker-Ward, Ornstein, & Starnes; Chae, Ogle, & Goodman; Peterson & Warren) who have examined how children remember a range of discrete distressing personal experiences. These teams describe their ongoing programs of research. Chae, Ogle, and Goodman open by outlining how attachment may influence children’s emotional reactions to and later memory for personally distressing experiences. Next, Baker Ward, Ornstein, and Starnes describe how children’s understanding of emotional events, both before events take place and as they unfold (what the authors termed extended encoding), affects children’s later memory. Peterson and Warren then describe findings from an ongoing line of work focused on children’s memory for injuries and hospital visits. The work has been instrumental in testing how stress, in conjunction with multiple individual-difference characteristics, relates to children’s memory. In the final chapter of the part, Greenhoot, Johnson, Legerski and Preface

vii

McCloskey review the state of research concerning stress and autobiographical memory development. They describe mechanisms that may underlie poor general autobiographical abilities among adolescents formerly exposed to trauma and present data from their ongoing longitudinal research that has tested several of these mechanisms. Overall, the chapters in Part I not only review current findings, but also draw explicit links to the topics discussed in subsequent Parts (e.g., attachment; chronic stress) and highlight the applications of the findings to theories concerning autobiographical memory development more broadly and to applied questions concerning children’s eyewitness abilities and developmental psychopathology. Part II comprises four chapters that highlight first the interrelations among coping, emotion regulation, attachment, and parent-child reminiscing; and second, the potential relations of these factors to children’s memory for stressful experiences. Although such topics have traditionally not been a primary focus in studies of children’s memory for stressful events, the literature reviewed in the chapters certainly demonstrates the relevance of coping, emotion regulation, and attachment for understanding how well parents and children reminisce about past negative events and children’s memory. Compas, Campbell, Robinson, and Rodriguez begin the part with a description of their dual process model of coping. They then discuss how executive functioning, especially working memory, may affect children’s use of controlled coping strategies, and how different coping processes may influence the occurrence of automatic intrusive memories. Oppenheim and Korens-Karie provide an overview of the secure-base function of attachment figures. The authors then describe several studies concerning parent-child discourse about emotional events. Findings suggest that emotionally matched narratives are reflective of a psychological secure base between a caregiver and child, and that these narratives have implications for a range of child outcomes. Third, Laible and Panfile discuss how parent-child conversations about emotional experiences, particularly after they occur, shape children’s emotion understanding, regulation capabilities, and well-being generally. Laible and Panfile further argue that it is critical to consider not only parent-child conversations, but also parent-child attachment, when evaluating how children come to understand negative emotional experiences. Finally, Sales describes how, via parent-child conversations, children find meaning in stressful events. She also discusses findings from several studies that indicate children’s coping and parent-child attachment may moderate the process by which parents and children talk about both discrete stressful experiences and chronic life stressors. Part III contain four chapters, each focused on different aspects of neurobiological and physiological stress-response systems and the implications of activation of these systems for children’s memory. First, Alexander and O’Hara, in a compelling chapter, integrate psychobiological processes with other elements of the developmental context to illustrate the complexity of the relations between emotion and memory. The researchers specifically argue that it is not enough to take into account the nature of the emotional stimulus or characteristics of the individual. Instead, a transactional model approach is needed to understand more viii

PREFACE

fully how emotional experiences are retained in memory across development. In the next two chapters, Wiik and Gunnar, followed by Carver and Cluver, review in considerable detail evidence from several fields (e.g., nonhuman animal studies, investigations of adults’ physiological responses) concerning how children’s stress reactivity and exposure to chronic stress likely affect neurological and cognitive systems. Wiik and Gunnar focus on how stress likely affects children’s hypothalamic pituitary adrenal axis functioning, a key neuroendocrine regulatory system in humans. Carver and Cluver focus primarily on stress effects on brain structures involved in memory encoding and consolidation in children. In the fourth chapter, Wallin, Quas, and Yim describe how discrete stress responses according to multiple physiological systems, directly and in conjunction with sociocontextual factors, may influence children’s memory for stressful experiences. Overall, these chapters highlight the utility of considering biological factors, in conjunction with sociocontextual factors, to understand the many ways in which stress affects children’s memory, both broadly and for specific incidents. Finally, a critical feature of the volume is the final parts, which contains four integrative chapters each written from a somewhat different theoretical and applied perspective. These authors comment on the fi ndings discussed in the prior chapters, and, in doing so, synthesize the different literatures reviewed. Fivush takes a broad approach to integrating the chapters, further demonstrating the need to consider multiple levels of analysis when studying children’s memory. She persuasively argues that, not only are developmental level and individual differences complexly intertwined across time, but both are modulated by socioemotional contexts that provide the framework for the development and expression of memory and emotion. She then systematically combines evidence presented across the chapter to demonstrate this point and to suggest several important next steps for research. The next three authors in this part review specific issues raised across the volume. Thompson focuses on the important role that parent-child relationships play in children’s emotional reactions to, understanding of, and, later memory for, negative experiences. Thompson also describes key questions that arise when attempting to integrate literature across the chapters, questions whose answers will provide much-needed insight into the mechanisms that likely underlie the influence of close relationships on emotion and memory across development. Bauer provides perhaps the most critical evaluation of the research presented in the volume. She emphasizes that researchers, in their attempt to focus on specific influences on children’s memory for stressful events, have often oversimplified what is a complex, multifaceted phenomenon: how children attend to, interpret, and later reconstruct prior stressful experiences. As she clearly articulates, if the field is to continue to advance, future research must take a more comprehensive approach in studying stress and memory across development. Finally, Salmon and Conroy change directions somewhat and highlight the practical implications of research to date in two critical areas. These researchers describe the forensic relevance of the research for evaluations of child witnesses’ accounts of traumatic events to which they were exposed. The researchers then turn to the clinical relevance Preface

ix

of the research for, most notably, clinical evaluations of the effects of emotion on children’s event memory, but also for how emotional memories affect children’s functioning over time and for treatment possibilities of these effects.

Conclusions The overarching purpose of both the Emory Cognition Project Conference and this volume was to bring together researchers from diverse fields whose work had implications for understanding the complex links between emotion and memory in development. We hoped that, through the different authors’ presentations, interactions, discussions, and writing, common trends would be evident. And, from these common trends, we anticipated that a new, innovative research agenda would emerge. As is evident from the contributions of this volume, we have achieved success. The different contributors, from varying perspectives, provide valuable insight not only into the complex ways in which emotions likely affect memory, but also into the powerful mediating and moderating roles of children’s age and biological response proclivities, parental and familial influences, and the broader context within which children recount prior emotional autobiographical experiences. This insight lays the foundation for the next wave of research in this arena, and we look forward to taking part in this endeavor for years to come.

x

PREFACE

Contents

Contributors xiii I

Stress and Memory, Empirical Evidence

1. Remembering Negative Childhood Experiences: An Attachment Theory Perspective 3 Yoojin Chae, Christin M. Ogle, and Gail S. Goodman 2. Children’s Understanding and Remembering of Stressful Experiences 28 Lynne Baker-Ward, Peter A. Ornstein, and Lauren P. Starnes 3. Injuries, Emergency Rooms, and Children’s Memory: Factors Contributing to Individual Differences 60 Carole Peterson and Kelly L. Warren 4. Stress and Autobiographical Memory Functioning 86 Andrea Follmer Greenhoot, Rebecca J. Johnson, John-Paul Legerski, and Laura A. McCloskey II Stress, Coping, and Parent-Child Narratives 5. Coping and Memory: Automatic and Controlled Processes in Adaptation to Stress 121 Bruce E. Compas, Laura K. Campbell, Kristen E. Robinson, and Erin M. Rodriguez 6. Mother-Child Emotion Dialogues: A Window into the Psychological Secure Base 142 David Oppenheim and Nina Koren-Karie xi

7. Mother-Child Reminiscing in the Context of Secure Attachment Relationships: Lessons in Understanding and Coping with Negative Emotions 166 Deborah Laible and Tia Panfile 8. Creating a Context for Children’s Memory: The Importance of Parental Attachment Status, Coping, and Narrative Skill for Co-Constructing Meaning Following Stressful Experiences 196 Jessica McDermott Sales III

Stress, Physiology, and Neurobiology

9. An Integrated Model of Emotional Memory: Dynamic Transactions in Development 221 Kristen Weede Alexander and Karen Davis O’Hara 10. Development and Social Regulation of Stress Neurobiology in Human Development: Implications for the Study of Traumatic Memories 256 Kristen L. Wiik and Megan R. Gunnar 11. Stress Effects on the Brain System Underlying Explicit Memory 278 Leslie J. Carver and Annette Cluver 12. Physiological Stress Responses and Children’s Event Memory 313 Allison R. Wallin, Jodi A. Quas, and Ilona S. Yim IV Integration and New Directions 13. Co-constructing Memories and Meaning over Time Robyn Fivush 14. Relationships, Stress, and Memory Ross A. Thompson

343

355

15. Complications Abound, and Why That’s a Good Thing Patricia J. Bauer

374

16. Emotion and Memory in Development: Clinical and Forensic Implications 394 Karen Salmon and Rowena Conroy Author Index 415 Subject Index 427

xii

CONTENTS

Contributors

Kristen Weede Alexander Department of Child Development California State University Sacramento, CA

Yoojin Chae Department of Psychology University of California Davis, CA

Lynne Baker-Ward Department of Psychology North Carolina State University Raleigh, NC

Annette Cluver Department of Psychology University of California San Diego, CA

Patricia J. Bauer Department of Psychology Emory University Atlanta, GA

Bruce E. Compas Department of Psychology and Human Development Vanderbilt University Nashville, TN

Laura K. Campbell Department of Psychology and Human Development Vanderbilt University Nashville, TN

Rowena Conroy Royal Children’s Hospital Melbourne, Australia Robyn Fivush Department of Psychology Emory University Atlanta, GA

Leslie J. Carver Department of Psychology University of California San Diego, CA

xiii

Gail S. Goodman Department of Psychology University of California Davis, CA Andrea Follmer Greenhoot Department of Psychology University of Kansas Lawrence, KS Megan R. Gunnar Institute of Child Development University of Minnesota Minneapolis, MN Rebecca J. Johnson Children’s Mercy Medical Center Kansas City, MO Nina Koren-Karie Department of Psychology University of Haifa Mr. Carmel, Israel Deborah Laible Department of Psychology Lehigh University Bethlehem, PA John-Paul Legerski Department of Psychology University of Kansas Lawrence, KS Laura A. McCloskey Institute for Research on Women and Gender University of Michigan Ann Arbor, MI Christin M. Ogle Department of Psychology University of California Davis, CA xiv

CONTRIBUTORS

Karen Davis O’Hara Department of Child Development California State University Sacramento, CA David Oppenheim Department of Psychology University of Haifa Mt. Carmel, Israel Peter A. Ornstein Department of Psychology University of North Carolina Chapter Hill, NC Tia Panfile Department of Psychology Lehigh University Bethlehem, PA Carole Peterson Department of Psychology Memorial University of Newfoundland St. John’s, Canada Jodi A. Quas Department of Psychology and Social Behavior University of California Irvine, CA Kristen E. Robinson Department of Psychology and Human Development Vanderbilt University Nashville, TN Erin M. Rodriguez Department of Psychology and Human Development Vanderbilt University Nashville, TN

Jessica McDermott Sales Department of Behavioral Sciences and Health Education Emory University Atlanta, GA

Allison R. Wallin Department of Psychology and Social Behavior University of California Irvine, CA

Karen Salmon School of Psychology Victoria University of Wellington Wellington, New Zealand

Kelly L. Warren Department of Psychology Memorial University of Newfoundland St. John’s Canada

Lauren P. Starnes Department of Psychology North Carolina State University Raleigh, NC

Kristen L. Wiik Institute of Child Development University of Minnesota Minneapolis, MN

Ross A. Thompson Department of Psychology University of California Davis, CA

Ilona S. Yim Department of Psychology and Social Behavior University of California Irvine, CA

Contributors

xv

This page intentionally left blank

I

Stress and Memory, Empirical Evidence

This page intentionally left blank

1 Remembering Negative Childhood Experiences An Attachment Theory Perspective Yoojin Chae Christin M. Ogle Gail S. Goodman

A

ttachment theory provides a powerful framework for understanding children’s memory and suggestibility for emotional life experiences. In particular, attachment theory can serve as a basis for predictions about individual differences in the ways children respond to and remember stressful incidents. Beginning with research on children’s memory for distressing medical procedures (Goodman, Hepps, & Reed, 1986; Goodman, Bottoms, Schwartz-Kenney, & Rudy, 1991; Goodman, Hirschman, Hepps, & Rudy, 1991) and extending more recently to research on children’s memory for criminal victimization (e.g., Bidrose & Goodman, 2000; Hershkowitz, Lanes, & Lamb, 2007; Leander, Christianson, & Granhag, 2007), several decades of scientific study on children’s memory for stressful events have quickly elapsed. A substantial body of empirical knowledge has thus accumulated. It is now time for the field to turn to theory testing. In our own research, for theoretical guidance, we find that attachment theory offers substantial insights This chapter is based on work supported in part by the National Science Foundation (grant 0545413). Any opinions, findings, conclusions, or recommendations expressed in this article are those of the authors and do not necessarily refl ect the views of the National Science Foundation. We thank Stephanie Head for editorial assistance. Address correspondence to Dr. Gail S. Goodman, Department of Psychology, University of California, 1 Shields Avenue, Davis, CA 95616 ([email protected]).

3

about memory for distressing childhood experiences, in particular, those that activate the attachment system. In this chapter, we first provide a brief overview of attachment theory, and then review empirical findings from various laboratories on significant relations of attachment with children’s memory and suggestibility for emotional, attachment-evoking information, focusing on (1) associations between children’s attachment and their memory/suggestibility for attachment-related information, and (2) associations between parents’ attachment and children’s memory/suggestibility for such information. Following this, we consider potential mechanisms underlying the relations. We also discuss the information-processing stages (e.g., encoding, retrieval) during which the attachment effects may be operative. We then turn to a more in-depth description of our studies.

Attachment Theory Attachment theory was formulated by John Bowlby (1958) to explain the importance of early caregiver-child relationships for personality development and emotion regulation. Based on the degree to which their caregivers are available and provide support when needed, infants form expectations or internal working models (IWMs), which are viewed, theoretically, as fairly stable mental representations of the self and the caregiver (Bowlby, 1980; Bretherton & Munholland, 1999). These IWMs are then used to interpret others’ intentions and actions, anticipate future behaviors, and regulate responses, particularly in times of stress (Bowlby, 1969). The Strange Situation, which consists of a series of separations from and reunions with a caregiver, is a laboratory situation developed to assess the quality of infant-caregiver attachment (Ainsworth, Blehar, Waters, & Wall, 1978). Specifically, the Strange Situation consists of eight phases, involving, for example, the parent and child being in an unfamiliar room, a stranger entering, the parent leaving, and the parent and child reuniting. Based on infant behaviors in the Strange Situation, infant-caregiver attachment styles were originally categorized into three patterns: secure, insecure-avoidant, and insecure-anxious. Securely attached infants have IWMs reflecting available and responsive caregivers in stressful or threatening situations. These infants seek proximity, contact, and communication with their caregivers during reunions, are easily soothed by the caregiver, and generally cope with stress effectively. Insecure-avoidant infants represent their caregivers as unwilling or unavailable to soothe negative affect and thus tend to avoid or ignore the caregivers upon reunions despite experiencing a great amount of distress during separations. Insecure-anxious infants represent their caregivers as 4

STRESS AND MEMORY, EMPIRICAL EVIDENCE

inconsistently available or inconsistently supportive and thus sometimes cling excessively to a caregiver to avoid separation but also display angry, resistant behavior upon reunion. A fourth category, disorganized/ disoriented, was later conceptualized for infants who fail to maintain a coherent attachment strategy for dealing with reunion and exhibit sequences of behavior that seem to lack a clear goal or that evince a collapse of strategy (Main & Solomon, 1990). These four attachment patterns were first defined for infants, but similar attachment patterns have been identified in children at preschool age (Main, Kaplan, & Cassidy, 1985), although more stressful events (e.g., starting kindergarten) may be needed to evoke distress in children after infancy (Quas, Murowchick, Bensadoun, & Boyce, 2002). Attachment theory can also be applied to adults’ intimate relationships. Based on what appears to be continual effects of attachment styles on affectional bonds and behaviors in close relationships (e.g., with romantic partners) in adulthood, Hazan and Shaver (1987) expanded the theory to encompass adults’ attachment, which is hypothesized to be based on experiences with primary attachment figures (e.g., parents). Adult attachment styles (and parent-child attachment styles as well; Fraley & Spieker, 2003) are characterized by two relatively independent dimensions, avoidance and anxiety (Fraley & Shaver, 2000). Attachment avoidance, reflected by discomfort with closeness and intimacy (Brennan, Clark, & Shaver, 1998; Fraley, Davis, & Shaver, 1998; Fraley & Shaver, 1997), refers to attempts to deactivate, down-regulate, or inhibit attachment-system activation (Edelstein & Shaver, 2004). Avoidant attachment is further divided into a dismissing avoidant style (i.e., purporting not to need or like closeness and interdependence) and a fearful avoidant style (i.e., desiring closeness while also fearing some of its consequences). Attachment-anxiety, involving preoccupation with the availability and responsiveness of attachment figures (preoccupied attachment; Mikulincer, Birnbaum, Woddis, & Nachmias, 2000; Mikulincer, Gillath, & Shaver, 2002), is characterized by hyperactivation of the attachment system (Cassidy, 2000). Adults who score low on both avoidance and anxiety dimensions are considered securely attached. Individual differences in adult attachment have been associated with parenting behaviors, for example, with high scores on avoidance and/or anxiety being related to negative models of parenthood and parent-child relationships, such as stricter, harsher disciplinary practices, lower levels of warmth, tendencies to be easily aggravated by children, less perceived ability to relate to children, and colder relations with children (Rholes, Simpson, & Blakely, 1995; Rholes, Simpson, Blakely, Lanigan, & Allen, 1997). These models in turn influence children’s strategies of affect regulation, as evident, at least under stressful situations, in more severe Remembering Negative Childhood Experiences

5

emotional reactions demonstrated by children of insecure parents than by children of secure parents, who might expect their parents to provide safety and protection (Goodman, Quas, Batterman-Faunce, Riddlesberger, & Kuhn, 1994, 1997). Indeed, when children experience stressful or attachment-activating incidents, such as a painful medical procedure, adults’ attachment styles have important implications for responsiveness to and support of children (Edelstein et al., 2004).

Attachment and Memory Attachment theorists have proposed that attachment styles, and associated IWMs, affect not only relationship behavior but also children’s memory for attachment-related experiences (for a review, see Alexander, Quas, & Goodman, 2002). Individual differences in children’s own attachment and parents’ attachment may influence the kinds of information and extent to which children encode, elaborate on, and later retrieve and report the information. As one example, avoidant children whose bids for care have been rejected or belittled are theorized to develop a nonconscious strategy, called “defensive exclusion,” which limits processing of stressful events, with the goal of preventing activation of the attachment system, such as the negative affect associated with reminders of attachmentrelated loss or extended separations (Bowlby, 1980; Main, 1990). Such limits on processing have implications for memory. In contrast, anxious children are theorized to be hypervigilant to attachment-related stressors because of their overconcern for attachment-related issues (Mikulincer et al., 2000, 2002). However, studies have demonstrated inconsistent findings regarding the anxiety dimension and memory. Attachment orientation has also been related to children’s suggestibility. Bruck and Melnyk (2004) conclude, in a review of individual differences in suggestibility, that attachment shows promise of being consistently related to suggestibility: Children who are insecurely attached or whose parents are insecurely attached are more suggestible than their secure counterparts. This topic is particularly important in view of evidence that abused children, who are often questioned in forensic settings, are likely to have high rates of insecure attachment (Barnett, Ganiban, & Cicchetti, 1999). We make several basic assumptions in our research program on attachment and memory. First, we assume that many traumatic and stressful events trigger the attachment system. We further propose that internal working models (IWMs) act as sets of expectations that influence levels of distress and memory performance in attachment-related stressful situations. We also assume, based in part on our empirical 6

STRESS AND MEMORY, EMPIRICAL EVIDENCE

findings, that parental attachment is relevant to how children cope with stressful situations during and after they occur, which affects children’s memory and resistance to false suggestion. Finally, we assume that individual differences in attachment are important moderators of the extent to which threatening information is attended and processed, and thus remembered.

Children’s Attachment and Memory/Suggestibility Studies have shown that children’s attachment to a caregiver may be a valuable predictor of their memory for attachment-relevant information. Kirsh and Cassidy (1997) investigated associations between threeyear-olds’ attachment orientations, assessed in infancy using the Strange Situation procedure, and memory for attachment-related stimuli. Children were read stories in which mothers’ reactions to their children’s requests for help varied to correspond to three different attachment-related IWMs (i.e., secure mothers responded sensitively; avoidant mothers rejected their children; anxious mothers provided exaggerated responses). On a subsequent cued-recall test about the stories’ content, secure children recalled both responsive and rejecting stories better than did avoidant and anxious children, even with general cognitive functioning controlled statistically. Alexander and Edelstein (2001) also found that attachment security assessed by the Attachment Q-sort (Waters & Deane, 1985), completed by an adult familiar with the child’s behavior, was positively related to threeto five-year-olds’ memory for attachment-related stories, even with statistical control of age and temperamental characteristics. Overall, across these studies, the findings underscore the importance of attachment orientations in predicting children’s memory for attachment issues, over and above contributions of age, general cognitive functioning, and temperament. However, when to-be-remembered information did not evoke attachment concerns, a different pattern regarding attachment and memory was revealed. Alexander and Edelstein (2001) found no significant relations between children’s memory for a nonattachment-related play event (e.g., making hot chocolate) and their attachment security. Also, in a study by Belsky, Spritz, and Crnic (1996), which investigated relations between boys’ attachment styles assessed in the Strange Situation at 12 months and recognition memory at three years for details from nonattachmentrelevant puppet shows (e.g., receiving a birthday present, spilling juice), boys with secure attachment histories remembered more positive than negative information, whereas boys with avoidant- or anxious-attachment histories recalled more negative than positive information. Fewer studies have been published on children’s attachment and suggestibility. Quas, Qin, Schaaf, and Goodman (1997) proposed that Remembering Negative Childhood Experiences

7

insecure children might be particularly susceptible to the demand characteristics of an interview (i.e., social pressure to agree with an interviewer) because they might be emotionally needy or nervous and therefore more likely to comply with an interviewer in an effort to receive approval and thus less likely to report their own memory. Indeed, Clarke-Stewart, Malloy, and Allhusen (2004) observed significant associations between children’s attachment at 15 months, assessed in the Strange Situation, and suggestibility for nonattachment-related activities (e.g., administration of standard developmental assessments) at 5 years: Children lacking a secure attachment relationship with their mothers displayed increased suggestibility (e.g., agreements that the examiner had undressed them). This finding implies that the link between children’s attachment security and suggestibility exists even for information not directly related to attachment issues, possibly because insecurely attached children are more nervous and distressed in social interactions with an unfamiliar adult.

Parental Attachment and Memory/Suggestibility Based on reasoning and research supporting the intergenerational transmission of attachment (Benoit & Parker, 1994; Steele, Steele, & Fonagy, 1996), parental attachment has at times been used as a proxy measure for children’s attachment. However, there are reasons to be interested in parental attachment itself as an influence on children’s memory and suggestibility. Parents’ mental representations of attachment may be related to their sensitivity and responsiveness to their children’s attachment signals (e.g., their children’s distress, needs for comfort; Edelstein et al., 2004; Goodman et al., 1997; Van IJzendoorn, 1995). Research has shown the functional role of parental attachment in children’s memory and suggestibility for stressful incidents. As discussed in greater detail later in this chapter, in a series of studies on children’s recollections of medical procedures, such as well-child inoculations and an invasive medical test (i.e., voiding cystourethrogram fluouroscopy, also called VCUG), we found that children of avoidantly attached parents provided more inaccurate memory reports and displayed heightened suggestibility, particularly under high levels of distress, compared to children of securely attached parents. In contrast, Alexander and Schaaf (2001) reported that parental attachment was not significantly related to children’s memory for a less stressful, nonattachment-related play event. Thus, as is the case for children’s attachment style and children’s memory/suggestibility, parental attachment style and children’s memory/suggestibility seem to be particularly interrelated in regard to events that evoke attachment issues.

8

STRESS AND MEMORY, EMPIRICAL EVIDENCE

As implied by the intergenerational transmission of attachment, parental attachment style predicts children’s attachment style. Secure parents may be better able to perceive their child’s attachment signals accurately and react adequately. Avoidant parents may rebuff their child’s attachment behavior, because they prefer not to encounter strong emotions and close contact. Anxious parents may be overly focused on their own attachment experiences and thus respond inappropriately and at times excessively. According to theory and evidence, these differences in parents’ responsiveness to their children may in turn determine children’s attachment to their parents (Ainsworth et al., 1978; Bowlby, 1969; Campbell et al., 2004; DeWolff & Van IJzendoorn, 1997; Van IJzendoorn, 1995). Hence, sensitive and responsive parents are more likely to have securely attached children, whereas insensitive and unresponsive parents run a higher risk of having insecurely attached offspring. Consequently, the documented associations between parents’ attachment and children’s memory/ suggestibility for attachment-evoking events may be due to the quality of children’s attachment to their parents. However, significant relations between parental and child attachment have not always been demonstrated (e.g., Fivush & Sales, 2006). Thus, the link between the two attachment orientations cannot yet be assumed to be underlying the relations between parental attachment and children’s memory/suggestibility. Other mediators relevant to parental attachment might be responsible for the relations.

Mechanisms Underlying Attachment-Memory/ Suggestibility Relations Although mechanisms underlying attachment-memory/suggestibility relations await elucidation, researchers have proposed possible mediators. These include attentional focus, parent-child rehearsal, and sensitivity to interview context. Based on a review of previous theory and evidence, we discuss the role of these potential mediators in the mnemonic processes of encoding, storage, and retrieval.

Encoding Attachment-related IWMs and associated mental strategies likely act as affective and cognitive filters that influence preattentive processing and selective attention. This is a way in which attachment theorists retain the psychoanalytic notion of “defense” (Thompson, Laible, & Ontai, 2003). More specifically, based on Bowlby’s notion of defensive exclusion, avoidant children, who have learned to associate attachmentsystem activation with rejection and distress, are thought to regulate their

Remembering Negative Childhood Experiences

9

attention to (i.e., defend against) materials with emotional, attachmentrelated themes. If potentially upsetting information is not fully processed, the attachment system is less likely to be activated, thus reducing psychological pain or discomfort but at the same time impairing memory. Findings from the few studies of attachment-related differences in attention are consistent with this idea. In the study by Kirsh and Cassidy (1997) referred to earlier, when children were shown several sets of drawings depicting different mother-child dyads engaged in positive, negative, and neutral interactions, avoidant children looked away from the drawings more than did secure and anxious children. In addition, when shown different sets of drawings depicting mother-child dyads engaged in positive interactions and adult dyads expressing neutral affect, avoidant and anxious children looked away from the mother-child drawings more than did secure children. Additional evidence comes from a study by Main et al. (1985), which showed that infant attachment classification predicted six-year-olds’ attention to family photographs: Avoidant children were more likely than secure or anxious children to avoid looking at the photographs and actively turned away from them. In short, extant studies provide support for the theoretical link between children’s insecure, particularly avoidant, attachment and inattention or avoidance of attention to attachment-related information. Findings from research with adults provide enticing hints regarding the extent to which insecure children’s memory deficits and suggestibility result from restricted attention at encoding. Fraley, Garner, and Shaver (2000) examined adult avoidant attachment in relation to encoding of attachment-related information and decay of the information over time. In Fraley and colleagues’ study, some participants were interviewed about attachment-related issues immediately after encoding (about 3 minutes); other participants were interviewed at various delay intervals (3–21 days). Results suggested that, compared to individuals low in avoidance, highly avoidant individuals were less attentive to attachmentrelated information initially. However, the two groups forgot the encoded information at a similar rate. These findings imply that encoding and attentional mechanisms underlie avoidant adults’ memory deficits, as opposed to subsequent repression or lack of elaboration. If these findings generalize to children, memory deficits of avoidant children may be due to limited attentional processes at encoding similar to those used by avoidant adults. Parental attachment style also has implications for the amount and type of information children encode in attachment-eliciting situations, which may be mediated by parents’ responsiveness and support. As described in greater detail later in this chapter, Edelstein et al. (2004) found that parents high in avoidant attachment were less supportive 10

STRESS AND MEMORY, EMPIRICAL EVIDENCE

toward their children during an inoculation if their children were highly upset, whereas this pattern was reversed for less avoidant parents. These variations in parental behaviors in challenging situations may influence children’s distress level and ability to cope with stress, which in turn may affect children’s focus of attention. However, further research is necessary to make clear the mediating role of attentional processes at the encoding phase in the relations between attachment and memory/ suggestibility.

Storage Attachment also has implications for children’s retention of information in memory. Secure children who have developed IWMs of a responsive and supportive caregiver may feel comfortable thinking about stressful experiences and be able to handle aversive emotions. In contrast, avoidant children are likely to eschew attachment-related elaboration, desiring to keep the attachment system from being strongly activated. In addition, they may suppress or distort the details of attachment-related information to regulate their emotions (Bowlby, 1958; Bretherton & Munholland, 1999). Anxious children may have difficulties in dealing with negative emotions evoked by thoughts about stressful incidents, excessively focusing on the stressors and themselves. Thus, overall, insecure compared to secure children might be less able to think back to distressing incidents in a coherent and organized way. A study by Mikulincer and Orbach (1995) with adults provides support for these theoretical ideas. When asked about early childhood experiences associated with certain emotions, such as anxiety, sadness, or happiness, and asked to rate how they had felt at the time of the event, avoidant individuals reported less intense emotions than did anxious individuals in the sadness and anxiety episodes. Secure people fell in between the two insecure groups but did not significantly differ from them. In addition, anxious individuals were unable to repress negative affect or to inhibit emotional spreading, whereas secure people accessed negative memories easily without being overwhelmed. Insecure individuals’ discomfort with emotional issues might further contribute to lack of coherency in parent-child discourse concerning stressful information (Bretherton, 1990, 1996; Thompson, 2000). Indeed, research has established that the style and content of parent-child rehearsal are a function, at least in part, of the attachment status of the parent and child (Newcombe & Reese, 2004; Reese & Farrant, 2003). Specific to emotional issues, due to difficulty in confronting and coping with negative emotions, insecure mother-daughter dyads avoid elaborating on such emotions, whereas secure dyads openly expand on both positive Remembering Negative Childhood Experiences

11

and negative emotions (Farrar, Fasig, & Welch-Ross, 1997). Similarly, secure mother-child dyads discuss a potentially threatening topic (e.g., misbehavior) with more emotional openness than insecure dyads (Laible & Thompson, 2000). Of interest, in a study conducted by Laible (2004), although secure, compared to insecure, dyads discussed negative emotions more frequently, attachment security was not significantly related to the discussion of positively valenced emotions during memory conversations. Also, a study by Fivush and Vasudeva (2002), in which the majority of the mother-child dyads talked about highly positive experiences, revealed that the attachment status of mother-child dyads was not significantly related to their tendency to openly discuss emotions. Hence, attachment orientations seem to matter most when parents and children discuss negative emotions and negative experiences (Bowlby, 1988). These differences in reminiscing conversations may in turn shape children’s storage of information and subsequent recollection of that information (Fivush, 1994; Hudson, 1990; Nelson, 1993). Talking openly and fluently about emotions may make emotions and emotion-related memories more accessible and less threatening for children when they reflect on previous emotional experiences. Children may also learn communication patterns and emotional content, as reflected in their own use of emotion references, affective perspective taking, and early conscience development, through shared-memory talk with their parents (Laible, 2004; Laible & Panfile, this volume; Laible & Thompson, 2000; Leichtman, Skowronek, & Pillemer, 2005; Oppenheim & Waters, 1995). A study by Favez (1997) showed that mothers who included both memory prompts (e.g., “And then what happened?”) and affective prompts (e.g., “Were you sad?”) about a stressful separation video had children who remembered the video content better than did mothers who included only memory prompts.

Retrieval Sensitivity to interview context as a function of attachment style is another possible mediator of the attachment-memory/suggestibility relations. Although children are generally hesitant to confront and/or talk about highly negative incidents (Engelberg & Christianson, 2002), insecurely attached children and/or children of insecure parents, who are more nervous and distressed with an adult interviewer, may be particularly reluctant to share emotionally distressing experiences. Furthermore, such children may be more likely to simply accept interviewers’ suggestions, compared with their secure counterparts.

12

STRESS AND MEMORY, EMPIRICAL EVIDENCE

Some empirical evidence implies that an unfamiliar or nonsupportive interviewer may raise attachment concerns in children, contributing to poor performance among insecurely attached children, even when tobe-remembered events are unlikely to have evoked attachment issues. Davis et al. (1998) found that children of insecurely attached parents were more suggestible when interviewed about a play event by a cold, professional interviewer than by a warm, supportive interviewer, whereas the performance of children of secure parents did not vary as a function of the interviewer’s level of support. The authors reasoned that children of insecure mothers might be similarly insecure, thus being more apprehensive and distressed during the interview and having more difficulty resisting a neutral interviewer’s suggestions. That is, attachment style may not affect children’s memory per se, but rather, through children’s generalized IWM of relationships, attachment style may influence sensitivity to social interactions with an unfamiliar and/or nonsupportive adult.

Our Research on Attachment and Memory/ Suggestibility In this section, we delve into greater detail about our research on attachment, memory, and suggestibility. In doing so, we hope to illustrate the specific influence on memory/suggestibility of parental as well as child attachment. Before considering the research, a few general points about stress and memory are essential to address. Substantial debate exists as to whether, in the face of stressful, or even traumatic, experiences, memory accuracy suffers or benefits. Based on a comprehensive review, Christianson (1992) proposed that, during highly distressing events, memory for main stressors is strengthened, whereas memory for peripheral details is impaired. Others have argued that memory accuracy peaks for moderately arousing events and declines when events are so distressing as to activate defensive processes (Deffenbacher, Bornstein, Penrod, & McGorty, 2004). Both of these views may be true, at least to a certain extent, regarding memory for stressful childhood experiences. Our research reveals that overall, Christianson’s (1992) formulation applies well to a majority of people (e.g., more secure individuals) when interviewed about stressful events that likely activate the attachment system, but that Deffenbacher et al.’s (2004) ideas may be applicable to a subset of individuals (e.g., avoidant individuals), specifically those for whom the processing of stressful, attachment-related information is characterized by defensive exclusion.

Remembering Negative Childhood Experiences

13

Parental Attachment, Parental Responsiveness, and Child Memory/Suggestibility for VCUG The first study to uncover significant relations between parental attachment and children’s memory concerned 3- to 10-year-olds’ recollections of VCUG (Goodman et al., 1994, 1997). This medical test is at times ordered by doctors when a child has urinary track problems (e.g., bladder infections, excessive bed wetting). The procedure involves the following: The child and parent go to the Radiology Department of a hospital, at which point the child undresses, puts on a hospital gown, and lies on an x-ray table. Next, an x-ray technician enters and takes initial x-rays. A nurse then arrives to wash the child’s genitals. With the child forcibly held down, the nurse catheterizes the child through the urethra, which can be painful and scary. Then, in some hospitals, the standard procedure is to require the child’s parent to leave the room before a doctor enters to take additional x-rays. Finally, the parent is reunited with the child. This medical procedure, in some of its formal characteristics, has certain similarities to the Strange Situation. Although the VCUG entails direct assault (penetration) of the child’s body, whereas the Strange Situation does not, both procedures involve a parent and child in an unfamiliar setting, strangers and the parent entering and leaving the room, and a parent-child reunion. Thus, it was of interest to include an attachment measure in the study. However, given the age range of the children tested (3 to 10 years), a single child attachment measure was not feasible. Fortunately, an adult attachment measure that could be easily administered to the parents was available (Bartholomew & Horowitz, 1991; Hazan & Shaver, 1987), and the parent completed it a few weeks after the VCUG, when the child’s memory and suggestibility regarding the invasive medical procedure was also tested. As would be expected based on an attachment-theory orientation, parental attachment style was correlated with children’s level of distress and later suggestibility. Parents with a more avoidant attachment style had more suggestible children, whereas parents with a relatively secure attachment style had children who made fewer errors to misleading questions. In multivariate path models, parental attachment was a stronger predictor of child memory errors than was child distress. Remarkably, parental attachment was an even stronger predictor than child age for some memory measures. We further found that the avoidant parents were more likely than secure parents to report that they had not prepared the child for the VCUG, had not talked much to the child about it after, avoided physically comforting the child, and did not have time to attend to the 14

STRESS AND MEMORY, EMPIRICAL EVIDENCE

child’s feelings. Despite these relations, a mediational analysis (Baron & Kenny, 1986) indicated that these factors did not mediate the correlations between parental attachment and child memory performance. In a second VCUG study, Quas et al. (1999) also showed that parental insecurity was predictive of inaccuracies in three- to 13-year-olds’ reports about VCUG even when the medical procedure had been experienced months to years earlier. Children of fearful-avoidant parents made more errors to specific questions and fewer do-not-know responses to misleading questions. In addition, children of dismissingly avoidant parents displayed heightened suggestibility. These first two studies showed important relations between parental attachment, children’s distress, and children’s memory. It was crucial to determine, however, if the results would generalize to other stressful situations. We also wondered if we could capture differences in parents’ reactions to children’s distress while the stressful event was unfolding. We thus turned to videotaping children while they received inoculations as a regular part of their health care. In this way, we could also include a wider range of families.

Parental Attachment, Parental Responsiveness, and Child Memory/Suggestibility for Inoculations Edelstein et al. (2004) examined relations between adult attachment and parental responsiveness toward 3- to 7-year-olds during well-child inoculations. Compared to VCUGs, inoculations are typically less stressful. However, for some children (e.g., those who receive six shots in one sitting), inoculations can be quite upsetting. We hypothesized that relations between attachment and parental responsiveness would be most evident for the children who became particularly distressed during inoculations. Previous research indicated that the amount of emotional support provided to a romantic partner is predicted by an interaction between attachment-related avoidance and partner distress (Fraley & Shaver, 1998; Simpson, Rholes, & Nelligan, 1992). Specifically, avoidant individuals became less emotionally supportive of their partners when those partners became more distressed. To the extent that adults’ self-reported attachment styles reflect their relationship orientations more generally, an interaction between parental avoidance and children’s levels of distress was expected to predict parental responsiveness. That is, as children become more upset, more avoidant parents were anticipated to become less supportive, whereas the opposite was expected for more secure parents. It was important to ensure, however, that the findings were not due to child temperament or parental personality. Given the possible relations Remembering Negative Childhood Experiences

15

among child temperament, parental personality, and children’s reactions to stressful events, these factors were also considered in predicting parental responsiveness. In the study, parent-child dyads were observed while children received doctor-ordered inoculations at a county immunization clinic. The event was videotaped and coded independently for children’s distress and parental responsiveness. Distress was evaluated before, during, and immediately following the inoculation. In addition, a global evaluation of children’s overall reaction was made. Parental responsiveness was assessed using the Emotional Availability Scales (Biringen, Robinson, & Emde, 1998), which measure the affective quality of parent-child interactions along four emotional availability dimensions: sensitivity, structuring, nonintrusiveness, and nonhostility. Parental attachment style was measured using the Relationship Scales Questionnaire (RSQ, Griffin & Bartholomew, 1994), children’s temperament was assessed by the Children’s Behavior Questionnaire (Ahadi, Rothbart, & Ye, 1993), and parental personality was measured using the NEO Five-Factor Inventory (Costa & McCrae, 1992). These are all well-regarded, standardized measures used frequently in psychological research. A positive relation was found between adult avoidant attachment and children’s distress during the inoculation such that having a parent who scored higher on attachment avoidance was predictive of more child distress during the inoculation. Hierarchical regression revealed that the influence of parental avoidance on children’s distress remained significant after controlling for parental personality and children’s temperament. Attachment anxiety was unrelated to children’s distress. Significant relations also emerged for adult avoidant attachment and parental responsiveness. Parents scoring higher on attachment avoidance were rated as less responsive to their children during the inoculation. An interaction between adult avoidant attachment and children’s emotional reaction predicted parental responsiveness, such that parents scoring high on avoidance were less responsive to their children, especially when their children were highly distressed. It thus appeared that we could capture attachment-related differences in parental behavior toward children during a stressful event itself. Put simply, as the children became more distressed during the inoculations, secure parents became nicer to their children, and avoidant parents became less positively (and often more negatively) engaged. Given that stress level and parental responsiveness during a distressing incident might influence children’s attention to the stressor at the encoding phase and thus affect memory accuracy, we wondered if parental attachment would predict children’s memory and suggestibility about the inoculations. 16

STRESS AND MEMORY, EMPIRICAL EVIDENCE

Thus, Alexander et al. (2002a) investigated relations between adults’ self-reported attachment and their children’s memory and suggestibility for the inoculation event studied by Edelstein et al. (2004). It was predicted that children of parents with secure adult attachment orientation would have better memory for the inoculation and greater resistance to suggestion than children of insecurely attached parents. Moreover, it was expected that for most children, greater distress would be associated with better memory for the inoculation (Christianson, 1992), but that for children of avoidant parents, greater distress would be associated with worse memory (Deffenbacher et al., 2004). Regarding the latter association, not only might defensive exclusion be involved but several additional factors might be at work as well. For example, our VCUG studies (Goodman et al., 1997) indicated that children of avoidant parents were less likely to have been prepared for the procedure, which might contribute to poorer encoding, and less likely to have received parental help to cope emotionally afterwards, that is, at the storage phase. Moreover, during the inoculation event, Edelstein et al.’s (2004) results showed that the children of avoidant adults had to deal with a hostile parent in addition to the shots, thus perhaps encoding less information about the inoculation. Children between the ages of three and seven were interviewed about getting their shots approximately two weeks after the shots had taken place. Memory interviews included free-recall questions, and direct and yes/no questions that were either specific or misleading in nature. Selfratings of parental attachment styles were obtained using the Relationship Questionnaire (RQ; Bartholomew & Horowitz, 1991) and the RSQ (Griffin & Bartholomew, 1994). Results revealed a positive relation between children’s distress and the accuracy of their memory reports. However, this relation was found only for children of parents who scored low on attachment avoidance, that is, parents who were more secure. Children of parents with high avoidant-attachment scores tended to show poor memory for free recall and yes/no questions about the inoculation, especially when the children had been highly distressed. The memory findings of Alexander et al. (2002a) thus mirrored the parental responsiveness fi ndings of Edelstein et al. (2004). The results suggest that the beneficial effect of stress on memory (Christianson, 1992; Goodman et al., 1991) may not extend to children of parents with high scores of attachment avoidance. The findings described thus far demonstrate the usefulness of attachment theory in predicting and understanding children’s memory for medical procedures. It was important to determine if the findings would extend beyond this domain. We were fortunate to be able to test the generalizability of our findings to the realm of adults’ and adolescents’ memory for childhood trauma, specifically the trauma of child sexual abuse. Remembering Negative Childhood Experiences

17

Attachment and Memory for Child Sexual Abuse Edelstein et al. (2005) examined the relation between adult attachment and long-term memory for child sexual abuse. The adults and adolescents had participated in a longitudinal study of the emotional outcomes for child sexual-abuse victims involved in the legal system (Quas et al., 2005). As already mentioned, previous research indicated that, although emotion generally enhances memory, individuals with avoidant-attachment styles may have memory deficits for highly emotional information. Thus, an interaction between child sexual abuse severity and attachment avoidance was hypothesized: Individuals low on avoidant attachment were expected to demonstrate the most accurate memory for severe child sexual abuse, whereas individuals high on attachment avoidance were expected to show worse memory for more severe abuse. The nature of the victims’ abuse experiences varied widely, including on dimensions of severity (e.g., from less severe child sexual abuse, such as fondling over the clothes, to highly traumatic child sexual assault, such as kidnap with rape at gunpoint), which permitted a test of the hypothesis that individuals’ with avoidant attachment styles would show greater memory deficits for particularly negative and severe abuse incidents. For the study, adults and adolescents with documented histories of child sexual abuse were interviewed 12 to 21 years after the alleged abuse ended. Detailed information (derived from prosecutors’ files, police reports, etc.) regarding participants’ prior abuse experiences was examined to assess memory accuracy for the abuse. That is, memory accuracy was evaluated based on discrepancies between the memory reports in comparison to the original documentation of sexual abuse during the victims’ childhoods. Abuse severity was calculated based on factors such as abuse duration, extent of sexual contact, and degree of force and injury sustained, and was used as an index of the emotional intensity of the abuse incident. The degree to which participants discussed their abuse experiences with friends and family members was also assessed. Adult attachment was measured using the RQ (Bartholomew & Horowitz, 1991). Regression analyses revealed the expected interaction between abuse severity and avoidant attachment. Child sexual-abuse victims with low scores on avoidant attachment showed greater memory accuracy for cases of severe abuse, whereas victims scoring high on avoidant attachment demonstrated worse memory for severe abuse. Furthermore, avoidant individuals were less likely to have discussed the abuse with others. Frequency of discussion was related to memory accuracy for the abuse, such that the less one talked about the abuse, the less accurate was memory for the abuse. Attachment anxiety was unrelated to memory 18

STRESS AND MEMORY, EMPIRICAL EVIDENCE

accuracy, irrespective of abuse severity. These fi ndings demonstrate the importance of understanding the influence of attachment in investigations of memory for real-life traumas. Similar to Alexander et al.’s (2002a) results on parental attachment and children’s memory for stressful events, the findings suggest that the beneficial influence of emotion on memory may not extend to individuals with avoidant attachments. Also, despite differences in whose attachment orientations were being measured across studies (i.e., parental attachment or victims’ own attachment), consistent findings emerged.

Attachment and False Memory Thus far, we have discussed memory for childhood events that were truly experienced. Does attachment also relate to memory for what has not occurred? That is, can attachment theory help us understand false memory? Although we are in the early phases of research on this topic, some promising leads have emerged. In our first foray into this area, we examined possible relations between false autobiographical memory and adult attachment (Qin, Ogle, & Goodman, in press). Based on research, as reviewed in this chapter, showing associations between insecure parental attachment and children’s memory inaccuracies (e.g., Quas et al., 1999), it was predicted that adult insecure attachment styles would be associated with higher levels of false autobiographical memory in adults. Similar to the “lost in the mall” false memory paradigm first employed by Loftus and Pickrell (1995), our paradigm involved asking parents of adult participants to supply information regarding various childhood events (e.g., going to the hospital for an injury) either experienced or not experienced by the adult participants before age five. Participants were then asked to recall details of three true events supplied by their parents and one false event. Adult attachment styles were measured in both parents and adult participants using the RQ (Bartholomew & Horowitz, 1991). Of particular relevance to the present chapter, the results showed that parents’ fearful-avoidant attachment predicted their adult children’s higher false memory scores. Parental attachment style was not significantly associated with participants’ true memory for autobiographical events. Of interest, adult participants’ own attachment styles were not predictive of true or false autobiographical memory reports. In our second attachment and false memory study, Schaaf, Alexander, and Goodman (2008) investigated relations between false memory, suggestibility, and attachment style in three- to five-year-olds. Of particular interest were potential mediators between parental attachment and children’s false memory resistance. Again using a procedure similar to Loftus Remembering Negative Childhood Experiences

19

and Pickrell’s (1995) “lost in the mall” paradigm, parents were fi rst interviewed to verify four true and four false events for their children. Children were then questioned repeatedly (in a single session) about true and false events by one interviewer in a leading fashion and after that by another interviewer who asked direct (nonleading) questions. Parental attachment was measured using the RQ (Bartholomew & Horowitz, 1991) and an adapted version of the RSQ (Griffin & Bartholomew, 1994), on which parents rated their level of agreement with statements concerning close relationships. To assess children’s parent-child attachment security, parents also completed the Attachment Q-Sort, which consists of a large number of cards, each describing a child’s specific secure-base behavior (Waters & Deane, 1985). Finally, children’s behavioral problems were measured using the Child Behavior Checklist (Achenbach & Edelbrock, 1983). It was hypothesized that higher attachment security for both children and parents would predict greater resistance to suggestion. Path analyses revealed parental insecure attachment predicted children’s insecure attachment. The latter predicted behavioral problems in the children, which predicted suggestibility. At last, a possible mediator of parental attachment and child memory errors was identified. However, findings concerning attachment and behavior problems in this study must be interpreted with caution, given that both parent and child attachment scores, as well as the behavioral adjustment scores, were based on parental ratings. We are currently conducting research to remediate this potential problem (e.g., Chae et al., 2007).

Remaining Questions The study of attachment and memory is still in its infancy, but it shows great promise for furthering scientific understanding of children’s memory and suggestibility for emotional events. Numerous issues remain; we mention just a subset of them here. Clearly, further work is needed to identify mediators of the attachment and memory relations we and others have uncovered. Moreover, although in this chapter we have focused on avoidant attachment and defensive exclusion, interesting relations may exist between anxious attachment and memory (Alexander et al., 2002a; Melinder et al., 2008). It is also important to examine the specificity of attachment-memory relations; for example, research is needed to determine if attachment orientation predicts memory mainly for attachmentrelated information or other types of information as well. Finally, the field would profit from additional research concerning whether attachment influences encoding, storage, and/or retrieval processes.

20

STRESS AND MEMORY, EMPIRICAL EVIDENCE

Practical Implications We emphasized theory in this chapter, but practical implications should be addressed as well, albeit with caution given that developmental science on attachment and memory is still relatively new. In regard to the legal context, maltreated children and their parents are likely to have insecure attachments (Carlson, Cicchetti, Barnett, & Braunwald, 1989; Cicchetti & Manly, 2001). Our findings may thus have implications for forensic interviews and courtroom testimony. For example, insecurely attached children may need additional rapport building so that sufficient trust can be established to aid resistance to misleading questions. Furthermore, experts who testify in court should be informed that important individual differences exist in trauma memory. There are also clinical implications of attachment and memory research. Given the potential importance of parental support for children’s mental health, clinicians may need to help avoidant parents better deal with children’s distress. Children who themselves are more avoidant might also profit from intervention regarding their ability to process rather than defend against memories of stressful experiences.

Conclusion Scientists have, for many decades, debated the mechanisms underlying memory for stressful childhood events. We believe attachment theory offers a viable explanatory framework for predicting important individual differences in memory for distressing experiences, at least those that activate the attachment system. When we are frightened or hurt, we are biologically programmed to want our loved ones to protect us. Expectations about whether or not we will receive protection may pervade encoding, storage, and/or retrieval of distressing childhood memories.

References Achenbach, T. M., & Edelbrock, C. (1983). Manual for the Child Behavior Checklist and Revised Child Behavior Profile. Burlington, VT: University of Vermont, Department of Psychiatry. Ahadi, S. A., Rothbart, M. K., & Ye, R. (1993). Children’s temperament in the US and China: Similarities and differences. European Journal of Personality, 7, 359–377. Ainsworth, M. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the strange situation. Potomac, MD: Erlbaum.

Remembering Negative Childhood Experiences

21

Alexander, K. W., & Edelstein, R. S. (Eds.). (2001). Children’s attachment and memory for an experienced event. Poster presented at the biennial meeting of the Society for Research in Child Development, Minneapolis, MN. Alexander, K. W., Goodman, G. S., Schaaf, J. M., Edelstein, R. S., Quas, J. A., & Shaver, P. R. (2002a). The role of attachment and cognitive inhibition in children’s memory and suggestibility for a stressful event. Journal of Experimental Child Psychology, 83, 262–290. Alexander, K. W., Quas, J. A., & Goodman, G. S. (2002b). Theoretical advances in understanding children’s memory for distressing events: The role of attachment. Developmental Review, 22, 490–519. Alexander, K. W., & Schaaf, J. M. (2001). Individual differences in children’s recounting of stressful events. Paper presented at the biennial meeting of the Society for Research in Child Development, Minneapolis, MN. Barnett, D., Ganiban, J., & Cicchetti, D. (1999). Maltreatment, negative expressivity, and the development of Type D attachments from 12 to 24 months of age. Monographs of the Society for Research in Child Development, 64, 97–118. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology, 61, 226–244. Belsky, J., Spritz, B., & Crnic, K. (1996). Infant attachment security and affective-cognitive information processing at age 3. Psychological Science, 7, 111–114. Benoit, D., & Parker, K. C. H. (1994). Stability and transmission of attachment across three generations. Child Development, 65, 1444–1456. Bidrose, S., & Goodman, G. S. (2000). Testimony and evidence: A scientific case study of memory for child sexual abuse. Applied Cognitive Psychology, 14, 197–213. Biringen, Z., Robinson, J. L., & Emde, R. N. (1998). The Emotional Availability Scales, 3rd Edition. Unpublished coding manual, Department of Human Development and Family Studies, Colorado State University. Bowlby, J. (1958). The nature of the child’s tie to his mother. International Journal of Psycho-Analysis, 39, 350–373. Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. New York, NY: Basic Books. Bowlby, J. (1980). Attachment and loss: Vol. 3. Loss: Sadness and depression. New York, NY: Basic Books. Bowlby, J. (1988). A secure base: Clinical applications of attachment theory. London, England: Routledge. Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment: An integrative overview. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 46–76). New York, NY: Guilford Press. Bretherton, I. (1990). Open communication and internal working models: Their role in the development of attachment relationships. In 22

STRESS AND MEMORY, EMPIRICAL EVIDENCE

R. A. Thompson (Ed.), Nebraska Symposium on Motivation: Vol. 36. Socioemotional development (pp. 59–113). Lincoln, NB: University of Nebraska Press. Bretherton, I. (1996). Internal working models of attachment relationships as related to resilient coping. In G. G. Noam & K. W. Fischer (Eds.), Development and vulnerability in close relationships (pp. 3–27). Mahwah, NJ: Erlbaum. Bretherton, I., & Munholland, K. A. (1999). Internal working models in attachment relationships: A construct revisited. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 89–111). New York, NY: Guilford. Bruck, M., & Melnyk, L. (2004). Individual differences in children’s suggestibility: A review and synthesis. Applied Cognitive Psychology, 18, 947–996. Campbell, S. B., Brownell, C. A., Hungerford, A., Spieker, S. J., Mohan, R., & Blessing, J. S. (2004). The course of maternal depressive symptoms and maternal sensitivity as predictors of attachment security at 36 months. Development and Psychopathology, 16, 231–252. Carlson, V., Cicchetti, D., Barnett, D., & Braunwald, K. (1989). Disorganized/ disoriented attachment relationships in maltreated infants. Developmental Psychology, 25, 525–531. Cassidy, J. (2000). Adult romantic attachments: A developmental perspective on individual differences. Review of General Psychology, 4, 111–131. Chae, Y., Goodman, G. S., Augusti, E-M., Larson, R. P., Alley, D., Culver, M., & Hansen, R. (2007). Children’s memory for a distressing event: The roles of parental attachment and authenticity. Manuscript in preparation. Christianson, S. (1992). Emotional stress and eyewitness memory: A critical review. Psychological Bulletin, 112, 284–309. Cicchetti, D., & Manly, J. T. (2001). Operationalizing child maltreatment: Developmental processes and outcomes. Development and Psychopathology, 13, 755–757. Clarke-Stewart, K. A., Malloy, L. C., & Allhusen, V. D. (2004). Verbal ability, self-control, and close relationships with parents protect children against misleading suggestions. Applied Cognitive Psychology, 18, 1037–1058. Costa, P. T., & McCrae, R. R. (1992). The Revised NEO Personality Inventory professional manual. Odessa, FL: Psychological Assessment Resources. Davis, S. L., Bottoms, B. L., Guererro, R., Shreder, E., Krebel, A., Reyes, R., Rohacs, J., & Stein, A. (1998). Attachment style, social support, and children’s eyewitness reports. Poster presented at the annual meeting of the Midwestern Psychological Association Convention, Chicago, IL. Deffenbacher, K. A., Bornstein, B. H., Penrod, S. D., & McGorty, E. K. (2004). A meta-analytic review of effects of high stress on eyewitness memory. Law and Human Behavior, 28, 687–706. DeWolff, M. S., & Van IJzendoorn, M. H. (1997). Sensitivity and attachment: A meta-analysis on parental antecedents of infant attachment. Child Development, 68, 571–591. Edelstein, R. S., Alexander, K. W., Shaver, P. R., Schaaf, J. M., Quas, J. A., Lovas, G. S., & Goodman, G. S. (2004). Adult attachment style and Remembering Negative Childhood Experiences

23

parental responsiveness during a stressful event. Attachment and Human Development, 6, 31–52. Edelstein, R. S., Ghetti, S., Quas, J. A., Goodman, G.S., Alexander, K., Redlich, A., & Cordon, I. (2005). Avoidant attachment and memory for child sexual abuse. Social and Personality Psychology Bulletin, 31, 1549–1560. Edelstein, R. S., & Shaver, P. R. (2004). Avoidant attachment: Exploration of an oxymoron. In D. Mashek & A. Aron (Eds.), Handbook of closeness and intimacy (pp. 397–412). Mahwah, NJ: Lawrence Erlbaum Associates. Engelberg, E., & Christianson, S. A. (2002). Stress, trauma, and memory. In M. L. Eisen, J. A. Quas, & G. S. Goodman (Eds.), Memory and suggestibility in the forensic interview (pp. 143–163). Mahwah, NJ: Lawrence Erlbaum Associates. Farrar, M. J., Fasig, L. G., & Welch-Ross, M. K. (1997). Attachment and emotion in autobiographical memory development. Journal of Experimental Child Psychology, 67, 389–408. Favez, N. (1997). Patterns of maternal emotional regulation and the narratives of an affective event by preschoolers. Poster presented at the biennial meeting of the Society for Research in Child Development, Washington, DC. Fivush, R. (1994). Constructing narrative, emotion and self in parent-child conversations about the past. In U. Neisser & R. Fivush (Eds.), The remembering self: Accuracy and construction in the life narrative (pp. 136–157). New York, NY: Cambridge University Press. Fivush, R., & Sales, J. M. (2006). Coping, attachment, and mother-child narratives of stressful events. Merrill-Palmer Quarterly, 52, 125–150. Fivush, R., & Vasudeva, A. (2002). Remembering to relate: Socioemotional correlates of mother-child reminiscing. Journal of Cognition and Development, 3, 73–90. Fraley, R. C., Davis, K. E., & Shaver, P. R. (1998). Dismissing avoidance and the defensive organization of emotion, cognition, and behavior. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 249–279). New York, NY: Guilford Press. Fraley, R. C., Garner, J. P., & Shaver, P. R. (2000). Adult attachment and the defensive regulation of attention and memory: Examining the role of preemptive and postemptive defensive processes. Journal of Personality and Social Psychology, 79, 816–826. Fraley, R. C., & Shaver, P. R. (1997). Adult attachment and the suppression of unwanted thoughts. Journal of Personality and Social Psychology, 73, 1080–1091. Fraley, R. C., & Shaver, P. R. (1998). Airport separations: A naturalistic study of adult attachment dynamics of separating couples. Journal of Personality and Social Psychology, 75, 1198–1212. Fraley, R. C., & Shaver, P. R. (2000). Adult romantic attachment: Theoretical developments, emerging controversies, and unanswered questions. Review of General Psychology, 4, 132–154. Fraley, R. C., & Spieker, S. (2003). Are infant attachment patterns continuously or categorically distributed? A taxometric analysis of strange situation behavior. Developmental Psychology, 39, 387–404.

24

STRESS AND MEMORY, EMPIRICAL EVIDENCE

Goodman, G. S., Bottoms, B. L., Schwartz-Kenney, B. M., & Rudy, L. (1991). Children’s testimony about a stressful event: Improving children’s reports. Journal of Narrative & Life History, 1, 69–99. Goodman, G. S., Hepps, D., & Reed, R. S. (1986). The child victim’s testimony. In A. Haralambie (Ed.). New issues for child advocates. Phoenix, AZ: Arizona Council of Attorneys for Children. Goodman, G. S., Hirschman, J., Hepps, D., & Rudy, L. (1991). Children’s memory for stressful events. Merrill-Palmer Quarterly, 37, 109–158. Goodman, G. S., Quas, J. A., Batterman-Faunce, J. M., Riddlesberger, M., & Kuhn, J. (1994). Predictors of accurate and inaccurate memories of traumatic events experienced in childhood. Consciousness and Cognition, 3, 269–294. Goodman, G. S., Quas, J. A., Batterman-Faunce, J. M., Riddlesberger, M. M., & Kuhn, J. (1997). Children’s reactions to and memory for a stressful event: Influences of age, anatomical dolls, knowledge, and parental attachment. Applied Developmental Science, 1, 54–75. Griffin, D., & Bartholomew, K. (1994). Models of the self and other: Fundamental dimensions underlying measures of adult attachment. Journal of Personality and Social Psychology, 67, 430–445. Hazan, C., & Shaver, P. R. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 59, 270–280. Hershkowitz, I., Lanes, O., & Lamb, M. E. (2007). Exploring the disclosure of child sexual abuse with alleged victims and their parents. Child Abuse & Neglect, 31, 111–123. Hudson, J. A. (1990). The emergence of autobiographical memory in motherchild conversation. In R. Fivush & J. A. Hudson (Eds.), Knowing and remembering in young children (pp. 166–196). New York, NY: Cambridge University Press. Kirsh, S., & Cassidy, J. (1997). Preschoolers’ attention to and memory for attachment relevant information. Child Development, 68, 1143–1153. Laible, D. J. (2004). Mother-child discourse in two contexts: Links with child temperament, attachment security, and socioemotional competence. Developmental Psychology, 40, 979–992. Laible, D. J., & Thompson, R. A. (2000). Mother-child discourse, attachment security, shared positive affect, and early conscience development. Child Development, 71, 1424–1440. Leander, L., Christianson, S-V., & Granhag, P. (2007). A sexual abuse case study: What children remember and report. Psychiatry, Psychology, & Law, 14, 120–129. Leichtman, M. D., Skowronek, J. S., & Pillemer, D. B. (2005). Talking or not about events of the day: Effects on preschoolers’ long-term memories. Poster presented at the biennial meeting of the Society for Research in Child Development, Atlanta, GA. Loftus, E. F., & Pickrell, J. E. (1995). The formation of false memories. Psychiatric Annals, 25, 720–725. Main, M. (1990). Cross-cultural studies of attachment organization: Recent studies, changing methodologies, and the concept of conditional strategies. Human Development, 33, 48–61. Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and adulthood: A move to the level of representation. In I. Bretherton & E. Waters (Eds.), Growing points in attachment theory and research. Remembering Negative Childhood Experiences

25

Monographs of the Society for Research in Child Development, 50 (1–2, Serial No. 209), 66–104. Main, M., & Solomon, J. (1990). Procedures for identifying infants as disorganized/disoriented during the Ainsworth Strange Situation. In M. T. Greenberg & D. Cicchetti (Eds.), Attachment in the preschool years: Theory, research, and intervention (pp. 121–160). Chicago, IL: University of Chicago Press. Melinder, A., Alexander, K. W., Goodman, G.S., Cho, Y. I., Thorensen, C., Lonnum, K., & Magnussen, S., (2008). Children’s eyewitness memory: A comparison of two interviewing strategies. Manuscript submitted for publication. Mikulincer, M., Birnbaum, G., Woddis, D., & Nachmias, O. (2000). Stress and accessibility of proximity-related thoughts: Exploring the normative and intraindividual components of attachment theory. Journal of Personality and Social Psychology, 78, 509–523. Mikulincer, M., Gillath, O., & Shaver, P. R. (2002). Activation of the attachment system in adulthood: Threat-related primes increase the accessibility of mental representations of attachment figures. Journal of Personality and Social Psychology, 83, 881–895. Nelson, K. (1993). The psychological and social origins of autobiographical memory. Psychological Science, 1, 1–8. Newcombe, R., & Reese, E. (2004). Evaluations and orientations in motherchild narratives as a function of attachment security: A longitudinal investigation. International Journal of Behavioral Development, 28, 230–245. Oppenheim, D., & Waters, H. A. (1995). Narrative processes and attachment representation: Issues of development and assessment. Monographs of the Society for Research in Child Development, 60, 1076–1081. Qin, J. J., Ogle, C. M., & Goodman, G. S. (in press). Adults’ memories of childhood: True and false reports. Manuscript submitted for publication. Journal of Experimental Psychology: Applied. Quas, J. A., Goodman, G. S., Bidrose, S., Pipe, M.-E., Craw, S., & Ablin, D. S. (1999). Emotion and memory: Children’s long-term remembering, forgetting, and suggestibility. Journal of Experimental Child Psychology, 72, 235–270. Quas, J. A., Goodman, G. S., Ghetti, S., Redlich, A., Edelstein, R., Alexander, K., Cordon, I., & Jones, D. P. H. (2005). Childhood sexual assault victims: Long-term outcomes of testifying in criminal court. Monographs of the Society for Research in Child Development, Vol 70, Serial No. 280. Quas, J. A., Murowchick, E., Bensadoun, J., & Boyce, W. T. (2002). Predictors of children’s cortisol activation during the transition to kindergarten. Journal of Developmental and Behavioral Pediatrics, 23, 304–313. Quas, J. A., Qin, J., Schaaf, J. M., & Goodman, G. S. (1997). Individual differences in children’s and adults’ suggestibility and false event memory. Learning and Individual Differences, 9, 359–390. Reese, E., & Farrant, K. (2003). Social origins of reminiscing. In R. Fivush & C. A. Haden (Eds.), Connecting culture and memory: The social construction of an autobiographical self (pp. 29–48). Mahwah, NJ: Lawrence Erlbaum Associates. 26

STRESS AND MEMORY, EMPIRICAL EVIDENCE

Rholes, W. S., Simpson, J. A., & Blakely, B. S. (1995). Adult attachment styles and mothers’ relationships with their young children. Personal Relationships, 2, 35–54. Rholes, W. S., Simpson, J. A., Blakely, B. S., Lanigan, L., & Allen, E. A. (1997). Adult attachment styles, the desire to have children, and working models of parenthood. Journal of Personality, 65, 357–385. Schaaf, J., Alexander, K. W., & Goodman, G. S. (2008). Children’s false memory and true disclosure in the face of repeated questions. Journal of Experimental Child Psychology, 100, 157–185. Simpson, J. A., Rholes, W. S., & Nelligan, J. S. (1992). Support seeking and support giving within couples in an anxiety provoking situation: The role of attachment styles. Journal of Personality and Social Psychology, 62, 434–446. Steele, H., Steele, M., & Fonagy, P. (1996). Associations among attachment classifications of mothers, fathers, and their infants. Child Development, 67, 541–555. Thompson, R. A. (2000). The legacy of early attachments. Child Development, 71, 145–152. Thompson, R. A., Laible, D. J., & Ontai, L. L. (2003). Early understandings of emotion, morality, and self: Developing a working model. In R. V. Kail (Ed.), Advances in child development and behavior (Vol. 31, pp. 139–171). San Diego, CA: Academic Press. Van IJzendoorn, M. H. (1995). Adult attachment representations, parental responsiveness and infant attachment: A meta-analysis on the predictive validity of the Adult Attachment Interview. Psychological Bulletin, 117, 387–403. Waters, E., & Deane, K. E. (1985). Defining and assessing individual differences in attachment relationships: Q-methodology and the organization of behavior in infancy and early childhood. Monographs of the Society for Research in Child Development, 50, 41–65.

Remembering Negative Childhood Experiences

27

2 Children’s Understanding and Remembering of Stressful Experiences Lynne Baker-Ward Peter A. Ornstein Lauren P. Starnes

I

n a discussion of realism and abstraction in art, Georgia O’Keeffe (1922) focused on the transformation of experience and observed: “Nothing is less real than realism. Details are confusing. It is only by selection, by elimination, by emphasis that we get at the real meaning of things.” These constructive activities—so critical for the artist—are at the core of our efforts to understand and remember the events that we experience. Indeed, we are not passive observers of the events that make up our lives. Our impressions of our experiences are influenced by our expectations, our goals, and our knowledge of the world (Baker-Ward, Ornstein, & Principe, 1997; Bartlett, 1932; Paris & Lindauer, 1976). These impressions, moreover, are reflected in the ways in which our experiences come to be represented in memory, and the encoding processes that lead to the construction of these representations can be quite extended in time (see Baker-Ward et al., 1997). Indeed, after the initial encoding of an event, a variety of factors that arise from both endogenous (e.g., our own “internal” cognitive processes) and exogenous (e.g., our interactions with others) sources may transform our understanding and consequently alter the memory representation. From this perspective, autobiographical memory involves a set of constructive processes that can continue long after an experience has ended but which nonetheless influence our understanding and memory of the events 28

of our lives (see Fivush & Baker-Ward, 2005). In this chapter, we consider these constructive processes as they impact children’s understanding of emotionally significant experiences and their subsequent memory for these events.

Extended Encoding of Personal Experiences A Constructivist Account of Memory The constructivist perspective that is central to much research on cognition and its development emphasizes the active involvement of an individual in seeking to understand his or her experiences. From this point of view, understanding an event as it unfolds involves the application of underlying knowledge—broadly defined—to modify and embellish the “objective stimulus” record, in the process “going beyond the information given,” to use Bruner’s (1957) celebrated phrase. As we see it, both semantic (i.e., generic) and episodic (i.e., specific) memory representations are used in the interpretive process—sometimes quite unconsciously— as are self-schemata (e.g., internal working models of relationships with parents—see Oppenheim & Koren-Karie, this volume) and an individual’s vast storehouse of domain-specific knowledge (e.g., of the rules of soccer or chess). To a considerable extent, understanding involves the transformation of an “objective” event that may be experienced by a group of participants, i.e., Underwood’s (1963) nominal stimulus, into a personally relevant experience, i.e., Underwood’s functional stimulus. The roots of the constructivist position as it applies to explorations of memory can be seen in Binet and Henri’s (1894a, 1894b) studies of children’s superior recall of sentences, in comparison to their memory for words, and Bartlett’s (1932) explorations of adults’ recall of unfamiliar folk tales. Binet and Henri found that the basic idea underlying a long passage was more resistant to forgetting than were the actual words, leading them to suggest that the children in their studies were abstracting the underlying meaning of the material on the basis of their existing knowledge. Moreover, in an influential exploration of reconstructive processes in remembering, Bartlett (1932) reported that English adults distorted aspects of the North American folk tales that they were asked to remember, interpreting this difficult-to-understand material in terms of their own cultural experiences and then remembering their interpretations. Consistent with Bartlett’s findings of knowledge-based distortions, other researchers (e.g., Paris & Lindauer, 1977) have shown that the recall of meaningful prose passages can often be characterized by transformations and omissions of the original material. These effects reflect Understanding and Remembering of Stressful Experiences

29

the operation of constructive processes that are set in motion by underlying knowledge and can be seen in the integration of ideas both within and across sentence boundaries (Barclay & Reid, 1974; Bransford & Franks, 1972) and in the inferences that are imposed on the material (e.g., Paris & Lindauer, 1976).

The Impact of Constructive Processes There is ample evidence that these constructive processes operate at every phase of the information processing sequence, i.e., from initial encoding through memory storage, retrieval, and reporting (Baker-Ward et al., 1997). An individual’s prior knowledge is used routinely—most often unconsciously, but sometimes deliberately—to interpret ongoing events, with the resulting interpretations impacting the deployment of attention and the encoding of information in memory. Thus, the initial representation of an experience that is established in memory is driven clearly by knowledge-based constructive processes. However, it is also important to indicate that these processes can be quite extended in time and can lead to changes in understanding and interpretation. Indeed, both endogenous (e.g., reflection and rumination) and exogenous (e.g., conversation and exposure to misleading information) factors can affect the extended encoding of an experience, with corresponding modifications of the representation. Moreover, later in the information-processing sequence, the impact of knowledge can affect remembering as the details of a specific experience fade over time and generic information is used to supplement what can be remembered (Myles-Worsley, Cromer, & Dodd, 1986; Ornstein, Merritt, Baker-Ward, Furtado, Gordon, & Principe, 1998). Further, changes over time in general knowledge can lead to a reworking of what is remembered, as memory for a specific experience may be transformed on the basis of current understanding and beliefs (Greenhoot, 2000; Ross, 1989).

Constructive Processing over Time Although constructive processes influence all aspects of remembering— from “stimulus” input to retrieval and reporting—the impact of forces from within and “outside” the child during the retention interval is of special importance for the underlying memory representation. In terms of endogenous forces, personal reflection on recently experienced events can certainly alter one’s perspective, as can rumination and efforts to make sense of one’s personal experiences (Bruner, 1990), and thus modify the state of representations in memory (Baker-Ward et al., 1997). As discussed below, such internally driven, postevent processing may be most likely to occur when the experiences in question are infused with 30

STRESS AND MEMORY, EMPIRICAL EVIDENCE

emotion. Moreover, individuals may be aware of the operation of this processing, at least at some points in the retention interval, and may even initiate them deliberately (e.g., as in an attempt to understand how an injury occurred or why a relationship ended). On the other hand, the modification of a representation in memory may result from the spread of activation in semantic memory or other knowledge-based processes of which the individual is unaware, as is the case with instances of autosuggestion (Binet, 1900; see also Baker-Ward et al., 1997). The DeeseRoediger-McDermott (DRM) Illusion (Roediger & McDermott, 1995), in which participants erroneously report having seen a target word (e.g., “sleep”) as part of a list of stimuli with which the word was highly associated (e.g., “bed” and “dream”), provides one example. Complementing these endogenous influences on the underlying memory representations are exogenous forces that also serve to extend the encoding process. For example, conversations with others (peers or adults) about an experience, particularly one that is not entirely understood, can lead to changes in interpretation that impact the representation (Principe et al., 2006), as can access to accounts in the media or family photo albums and videotapes of similar experiences (Ornstein, Larus, & Clubb, 1991; Principe, Ornstein, Baker-Ward, & Gordon, 2000). In addition, memory distortions that arise from the presentation of contradictory information, as in the “misinformation effect” in the area of eyewitness testimony (Loftus, 1993), can be seen as reflecting the processes that are ordinarily involved in extended encoding. Moreover, related to studies of the misinformation effect are investigations of suggestibility in which the aim is that of simulating the use of misleading or coercive questioning of child witnesses (see Bruck & Ceci, 1999). The extensive literature on memory distortion notwithstanding, both endogenous and exogenous forces may also operate to maintain and enhance as well as to diminish the accuracy of the recollection of personal experiences. Based on research in cognitive psychology, it seems reasonable to expect that at least under some conditions the addition of information can make event representations more accessible and more likely to be maintained over long delays (see Baker-Ward et al., 1997; Fivush & Baker-Ward, 2005). For example, if initially unrelated aspects of an event are subsequently connected by the discovery of a causal relation between them, a more interconnected representational structure would be created in memory. As a consequence, thinking about one of the components should activate the other in memory, thereby increasing the accessibility and long-term maintenance of the information in memory. In addition, forgetting over time can be reduced considerably when information in memory is reactivated during the retention interval through repeated exposure to aspects of the original, a phenomenon Understanding and Remembering of Stressful Experiences

31

described as reinstatement (Brainerd & Ornstein, 1991; Campbell & Jaynes, 1966; Hudson & Sheffield, 1999). Within this perspective, then, the representation of an experience in memory begins, but certainly does not end, with the initial encoding of an event. Information acquired after an experience may change the meaning of that experience, rendering salient details that subsequently remain in memory but that might otherwise have been forgotten. Hence, a technologist’s decision to repeat an x-ray or a physician’s worried expression may be retained as components of the individual’s personal narrative if a medical condition is subsequently diagnosed, but may be readily forgotten if the patient is later given a clean bill of health. Similarly, a child who discovers that Santa Claus is mythical may reinterpret and remember the shopping bags discovered in a closet or the parental voices overheard on Christmas Eve, which now become central rather than peripheral components of the holiday experience. Moreover, two recent studies of adults’ story recall indicate that simply retelling a story prior to a later recall assessment can have a substantial effect on the underlying memory representation, consistent with the notion of extended encoding. Marsh (2007) reported that memory for the stories changed after initial encoding and storage, and Wang and Ross (2005) concluded that individuals’ memory representations eventually become amalgams of the original information and additional content derived from their retellings, and even their thoughts.

Stress, Emotion, and Extended Encoding Among children as well as adults, the events that comprise our everyday lives are not infrequently associated with some degree of stress, defined in terms of physiological and emotional reactions to threats from the outside world. Whereas traumatic events involve by definition (DSM-IV) a realistic threat to life itself, the threats experienced in more typical situations constitute real or perceived risks to one’s sense of safety (e.g., in experiencing a painful but necessary medical procedure) or status and self-worth (e.g., in experiencing a blocked goal) (see McEwan, 2000). Within the context of the “same” event (e.g., getting cut from the team), different individuals may experience alternative emotional reactions as a consequence of their varying interpretations of the experience (Lazarus, 1977), which are derived in part from their individual temperaments and personal histories. In response to these divergent emotions, different courses of action may be undertaken (Campos, Mumme, Kermonian, & Campos, 1994). A child who reacts to a stressful experience with anger, for example, may respond with aggression, whereas another who experiences sadness may withdraw from the scene. As a consequence of these diverging 32

STRESS AND MEMORY, EMPIRICAL EVIDENCE

approaches to a situation, alternative aspects of the event may become more or less relevant (Dolan, 2002; Easterbrooks, 1959). The individual who becomes hypervigilant in preparing to defend the self, for example, may focus on different aspects of the situation than will a child who withdraws from the scene and “shuts down.” In this regard, emotion can be expected to affect encoding and consequently determine to some extent the representation of an event in memory. It is also the case that alternative emotional experiences and expressions are likely to be associated with different opportunities and motivation for constructive processing over time as an individual reappraises a changing situation. For example, a child’s pride in her accomplishment could provide the impetus for extensive discussions with others about the details of the positive experience, resulting in an augmented representation in memory. In contrast, a child who feels shame regarding her performance could avoid revisiting the episode in question with others. Similarly, the sequelae of emotionally salient events may include multiple opportunities for the reinstatement or augmentation of memory (see Sales, this volume). For example, a child’s emotional displays may evoke explanations and reassurances from adults, or, alternatively, opportunities for interpretation and understanding may be limited if an episode such as sexual assault is silenced by the community (Fivush, 2000). It is also possible that events accompanied by self-conscious emotions, such as pride or guilt, may be distorted or reinterpreted over time (see Ross, 1989). Hence, emotion has implications for remembering at each point in the information-processing sequence (see Gordon, Schroeder, Ornstein, & Baker-Ward, 1995; Ornstein, Larus & Clubb, 1991).

An Approach to the Investigation of Extended Encoding Knowledge-based constructive processes thus play an important role in children’s memory for personally experienced events and may be especially salient when the events being remembered are emotional in tone. As such, our focus here is on understanding and remembering emotionally charged experiences among boys and girls between 3 and 10 years of age. To illustrate these processes, we present several investigations from our own research programs, focusing on (a) children’s understanding of personally significant experiences as they unfold and their subsequent memory for these events, and (b) the influences on remembering that may transpire after these events have occurred. Our discussion will highlight some of the ways in which emotions may affect understanding, extended encoding, and consequently remembering, in development. Understanding and Remembering of Stressful Experiences

33

Finally, we present some directions for further investigation and examine some implications of our work for forensic and clinical practice. For over 15 years, in collaboration with our late colleague Betty N. Gordon and a number of talented students and postdoctoral fellows, we have studied children’s memory for naturally occurring, personally significant experiences. In most of these investigations, our work has been motivated by questions regarding young children’s capacity to serve as witnesses in legal proceedings. Assuming that what cannot be remembered cannot be reported, we have charted young children’s retention of the details of their medical experiences over extended periods of time as a function of varied conditions of reporting. For overviews of this research, see Ornstein, Baker-Ward, Gordon, and Merritt (1997) and Gordon, Baker-Ward, and Ornstein (2001); see also Peterson (this volume) for a discussion of another research program involving children’s memory for medical treatment. For several reasons, we decided to use medical procedures as the tobe-remembered events in our studies. First, these procedures can serve as analogue events for the experiences about which children are often called upon to testify in legal proceedings. Consider, for example, the fact that during a physical examination of a child (a) is in partial and/ or complete states of undress, (b) is handled by an adult who is often unknown and of the opposite sex, and (c) can experience some degree of stress and pain. Of course, we readily acknowledge the important distinctions between such parent-ally sanctioned events and episodes of abuse; see Sales, this volume, for a discussion of the role of the parent in children’s responses to emotional experiences. Importantly, because the procedures that we observed were medically indicated, they provided an ethically appropriate venue for the investigation of children’s memory for stressful experiences. Second, through the cooperation of parents and medical professionals, it was possible to specify the components and details of each individual child’s medical experience. Hence, we were able to determine the accuracy of the child’s report of the event. Third, although the children’s involvement in the research did not affect their medical care, we could assess and in some cases manipulate a variety of factors that could impact recall, including the child’s prior knowledge of the specific medical procedure(s) and the timing of the memory interviews (e.g., Ornstein, Baker-Ward, Gordon, Pelphrey, Tyler, & Gramzow, 2006). Although much of our work has involved routine pediatric examinations (e.g., Baker-Ward, Gordon, Ornstein, Larus & Clubb, 1993; Greenhoot, Ornstein, Gordon, & Baker-Ward, 1999; Ornstein et al., 2006), we have also carried out studies of children’s memory for other medically related events, when they seemed appropriate for addressing 34

STRESS AND MEMORY, EMPIRICAL EVIDENCE

specific particular research questions. These events have included dental examinations and restorations (Vandermaas, Hess, & Baker-Ward, 1993), a urinary bladder catheterization procedure (Merritt, Ornstein, & Spicker, 1994), and emergency room treatment following a facial laceration (Burgwyn-Bailes, Baker-Ward, Gordon, & Ornstein, 2001). Moreover, recently we have augmented our research on children’s memory for medical experiences with investigations involving young participants’ reports of another event that has significance in their lives, namely, a final game in a season-ending soccer tournament (e.g., Baker-Ward, Eaton, & Banks, 2005). This work was initiated specifically to explore the effects of emotion on remembering across a delay interval. Although members of both the winning and losing teams participated in the same activity, at least at the level of the nominal stimulus, their experiences were typically quite divergent with regard to the emotions that accompanied the outcome of the game. Hence, within this setting, we could explore the effects of alternative emotions on aspects of the encoding and subsequent retrieval and reporting of the details of the game. The experiences examined in this chapter thus range from painful diagnostic procedures to recreational athletic competitions. The same general approach, however, characterized each investigation. Working with expert informants (physicians, dentists, nurses, and radiological technologists; coaches and referees), we defined on an a priori basis the standard components or “features” that comprise each type of experience, i.e., those aspects of the event that would be typically be encountered by each participant. Whereas the majority of these features were integral components of the event, across individual children there were some variations in the experiences. (For example, although most children receiving pediatric check-ups would have their vision screened, this typical aspect of the examination might be omitted for children who receive services from ophamologists.) Hence, to enable an examination of the accuracy of the children’s memory reports, observers recorded as each medical procedure unfolded the presence or absence of each of the event features that were listed on checklists. When possible, these records were augmented with video recordings. Children’s memory reports were obtained in individual interviews, in which an experienced examiner began with very general, open-ended questions (“Tell me what happened . . . ”) to elicit the child’s narrative account of the experience. In the investigations of medical procedures, the retention of each feature was assessed with a hierarchically organized protocol, such that increasingly specific prompts (closed questions to yes/no probes) were presented to the child when information was not forthcoming in response to more general queries. Consistent with the increased focus on children’s emotional responses and interpretation of their experiences, the interview Understanding and Remembering of Stressful Experiences

35

protocols for the soccer studies were modified somewhat, with increasing support provided for the generation of a narrative account of the experience, and with wh- questions addressing all features (including those provided in the narrative) presented in a separate portion of the interview. All interviews also included some plausible questions about actions that were not components of the event. Included in each study was a measure of the child’s level of arousal and/or emotional reactions to the event, although the nature of these measures varied across investigations. In some cases, it was possible to augment ratings of the children’s levels of stress with behavioral analyses of their video-recorded responses during the procedures. The studies presented below (cf. Merritt et al., 1994) are limited in that they did not include assessments of the multiple response systems through which stress may effect remembering (see Wallin, Quas, & Yim; and Wiik & Gunnar; this volume, for a discussion of the importance of obtaining such data in understanding the stress-memory linkage). In future research, we plan on devoting attention to the multiple pathways through which arousal, as moderated by knowledge, coping, and other variables, affects children’s memories for the important things that happen to them.

Studies of Extended Encoding and Remembering Prior Knowledge and Memory for a Pediatric Examination A recent investigation from our laboratory illustrates the facilitative effects of domain-relevant knowledge on children’s recall of a medical check-up (Ornstein et al., 2006). In most studies of the impact of prior knowledge on memory, knowledge has been explored by contrasting the performance of groups of individuals who differed in expertise in domains such as chess (e.g., Chi, 1978) or soccer (Schneider, Körkel, & Weinert, 1989). Moreover, in a few investigations, the feature of a salient event such as a visit to the doctor (as opposed to the participant) has been taken as the unit of analysis, with the critical question being that of the differential recall of event components that varied along a scale of children’s understanding, as defined by normative data (Clubb, Nida, Merritt, & Ornstein, 1993; Ornstein et al., 1997). In contrast to these approaches, Ornstein et al. (2006) included a knowledge assessment at the individual level and thus could compare knowledge and memory scores within the same participants. To assess the children’s knowledge, half of the sample of four-toseven-year-olds was interviewed regarding the components of a pediatric examination prior to their own medical check-ups; the remaining half 36

STRESS AND MEMORY, EMPIRICAL EVIDENCE

was assigned to a control condition to verify that the knowledge interviews did not affect subsequent recall performance. All of the children were asked to report the component features of their physical examinations immediately after their check-ups, and then again after a delay of six months. As shown in Figure 2–1, knowledge of the details of routine physical examinations increased with age, and the children’s knowledge scores predicted both initial, r = .53, p < .001, and delayed recall, r = .30, p = .05. The results of regression analyses indicated that knowledge accounted for a significant proportion of the variance in open-ended recall at the initial interview, even after age and vocabulary scores had been entered into the model.

A First Dental Visit: Preparation and Recall

80%

80%

60%

60% Memory Scores

Memory Scores

Given this demonstration of the impact of knowledge on children’s memory for a familiar physical examination, is it possible to observe knowledge effects on three-year-olds’ reports of a novel and possibly stressful procedure? To examine this issue, Baker-Ward and Ornstein (in preparation) recruited a sample of 46 three-year-old children who were scheduled for their very first dental examination. Through the cooperation of their dentist, we set out to manipulate the children’s preparation (and, presumably, their knowledge) for this initial experience. With the assistance of the parents, we were also able to explore the effects of naturally occurring preparation, including their deliberate attempts to increase their children’s understanding of the nature and importance of dental treatment, as well as the children’s incidental exposure to relevant information.

40%

R Sq Linear = 0.092

40%

20%

20% R Sq Linear = 0.277

0%

0% 0%

20%

40%

Knowledge Scores

60%

0%

20%

40%

60%

Knowledge Scores

Figure 2–1. Scatterplots illustrating the associations between knowledge and open-ended recall at the immediate (left panel) and six-month (right panel) interviews. Note that when participants have identical scores, their data points overlap. (Adapted from Ornstein et al., 2006, Figure 3. Used by permission from the American Psychological Association.) Understanding and Remembering of Stressful Experiences

37

Three experimental conditions were established, and the children were randomly assigned to one of three groups that differed with regard to the materials they received from us about two weeks before their checkups. The families in the first group were mailed a specially prepared video that showed two child models receiving complete dental treatments by the children’s dentist and a dental hygienist. Those in the second group, a manipulation control condition, were mailed our own production of “Healthy Snacks for Healthy Teeth.” This video was comparable to the experimental tape in length and with regard to parents’ ratings of its appeal, but it conveyed no knowledge relevant to the dental visit. Moreover, the families assigned to third group, a baseline control condition, received only a letter confirming their participation in the research. To determine the impact, if any, of our videos, the children’s knowledge of the features to be included in the dental examination was assessed briefly when they arrived at the dentist’s office. This interview consisted only of a very general prompt (“What will happen when you see the dentist?”), and the children provided little information about their expectations. Even though the parents in both video conditions indicated that they complied with our instructions to play the videos on at least two occasions (M = 2.33), the performance of the three groups on this knowledge assessment was quite similar. Indeed, the children in the knowledge, control, and baseline conditions reported only 1.87, 1.25, and 0.86 features of the examination, respectively. The children’s dental treatment, which included an examination, teeth cleaning, and the application of fluoride, was video-recorded, but was otherwise unaffected by their participation in the research. Six weeks after the visit, all of the children were interviewed at a university laboratory. We used a protocol consistent with our general approach, as described above. The questions were again hierarchically structured, with very general, open-ended questions followed by increasingly specific probes, and the children’s retention of standard features of the visit was assessed. In contrast to expectations—but consistent with the failure to find significant differences in children’s knowledge, as a function of the videotape manipulation—no overall group effects were present; the experimenter-provided preparation did not enhance recall. Although our experimental manipulation of information in an attempt to differentially prepare children for their fi rst dental examination was not successful, we were quite aware that the children were prepared for the check-up in other ways. Did such preparation—which we assume results in more knowledge—affect retention? To examine this question, we tabulated all the preparation activities reported by the parents on the questionnaires they completed at the dentist’s office. Parents reported a total of 12 different types of preparation activities (e.g., reading books 38

STRESS AND MEMORY, EMPIRICAL EVIDENCE

about dentists). We then asked a group of nine experts, all of whom were practicing dentists, to rank-order the preparation activities in terms of their presumed effectiveness. Attesting to the potential importance of naturally occurring experiences, the preparation we provided, viewing a specially prepared video, ranked behind visiting the dental office in advance and observing a sibling’s dental treatment. On the basis of these rankings, we reconstituted the groups. The High Preparation condition was composed of the 16 children who had experienced one or more of the three most effective techniques, whereas the Medium Preparation group included 10 children who had had exposure to one or more of the next three techniques. The remaining 20 children in the Low Preparation Group received no specific preparation at all. Interestingly, as can be seen in Figure 2–2, these post hoc groups did differ with regard to memory performance, with the participants in the High Preparation group recalling significantly more of the features than did the children in either the Medium or Low Preparation groups, whose performance did not differ. Hence preparation that presumably resulted in greater understanding of the event was associated with higher levels of memory performance, at least when naturally occurring experience is considered. Of course, quasi-experimental findings must be interpreted cautiously. We can, however, rule out a number of potential alternative explanations for the effect, including preparation group differences in months of age, trait anxiety, and child temperament, as assessed by parental questionnaires. In addition, there were no between-group similarities in responses to questions about actions that were not part of the dental visit, making it seem unlikely that the High Preparation group’s better recall

0.6

Proportion Recall

0.5 0.4 0.3 0.2 0.1 0 High

Medium Preparation Group

Low

Figure 2–2. Proportion of standard features of their first dental examinations recalled by three-year-old children in three post hoc preparation conditions.

Understanding and Remembering of Stressful Experiences

39

performance can be attributed to a response bias that could have influenced the measure of total recall. It is also the case that the preparation effect cannot be attributed to reductions in anxiety. An original purpose of the investigation was to lower anxiety during an event in order to assess the effects on memory. However, the children were quite relaxed during their dental examinations, as indicated by parents’ and hygienists’ ratings and by an observational measure of anxiety, the Behavioral Profile Rating Scale (BPRS; Melamed & Lumley, 1988), and no overall group differences in anxiety were found. Finally, one additional limitation of the present research design must be noted: because we did not assess memory immediately after the dental visit, it is possible that the effects of preparation observed here operated during the retention interval instead of at encoding, although there is no evidence for this effect.

The Provision of Information During an Invasive Procedure An additional investigation in our laboratory was carried out by Ornstein, Principe, Hudson, Gordon, and Merritt (1997) to examine the effects of children’s understanding of what is happening to them on their subsequent memory of a highly stressful experience. In this follow-up to a previous study (Merritt et al., 1994), 24 children between three and eight years of age were observed as they underwent an invasive radiological procedure involving urinary bladder catheterization and were interviewed on three occasions about their experience. The procedure was a voiding cystourethrogram (VCUG), which is performed when a physician needs an evaluation of the urinary tract, as is often the case in children with repeated bladder infections. The 20 to 30 minute VCUG procedure begins with the child being placed on a table beneath the fluoroscopy camera. The genital area is then uncovered and cleansed, and a catheter is inserted into the urethra and passed into the urinary bladder. Contrast fluid is infused until the bladder is filled to capacity, and fluoroscopic filming is then performed. The procedure ends with the child being instructed to urinate on the table, a process that leads to the forced expulsion of the catheter. Following our standard procedures, we identified 10 specific features of the VCUG and used a hierarchically structured interview protocol to assess the retention of these features at each of three memory interviews. Half of the children were interviewed immediately after the VCUG; all of them were questioned in their homes at both six and 16 weeks after the procedure. Although the children and their families did not receive a uniform orientation to the procedure, the technologists provided some patients with more information than others. The technologists’ talk to the young 40

STRESS AND MEMORY, EMPIRICAL EVIDENCE

patients was classified as including a procedural narrative when they were provided with an account containing a description of (a) the catheter and its insertion, (b) the infusion of contrast fluid, and (c) the filling of the bladder. Following up Principe et al.’s (1996) reanalysis of Merritt et al.’s (1994) data, we thought it possible that presentation of information in this narrative form may have permitted the child to better understand the event as it unfolded. With an enhanced understanding of the experience, the child could construct a more elaborated representation of the event, which could serve as a foundation for subsequent remembering. Moreover, because the features addressed in the procedural narrative must have taken place in order for some other components to occur (e.g., the child’s bladder must be filled before it can be emptied), information regarding these “enabling” features could be expected to affect the child’s understanding of other components of the VCUG as well. Two post hoc groups were formed, consisting of 12 children who received a procedural narrative from their technologists and 12 who were not given this initial overview of the procedure. As shown in Figure 2–3, children who received the procedural narrative reported a greater proportion of the features of the event at each occasion. As it turned out, however, the technologists were more likely to provide the procedural narrative to the older children, whom, we suspect, they thought would benefit from the information. The post hoc Narrative Group had a mean age of 76 months, compared with only 53 months for the children in the No Narrative Group. When the recall data were statistically adjusted for age, the facilitative effect of the procedural narrative on recall remained at the initial interview, but was no longer significant at the delayed assessments. It is possible that participation in the memory interview itself served to enhance remembering across occasions among the children who had not received a procedural narrative and may have had little understanding of what had happened to them. Indeed, the interview could have provided additional information or assisted the child in creating an organization structure that could support subsequent reports. Providing some support for this possibility, Starnes and Baker-Ward (under review) found that four-year-olds who received an immediate interview following an experimenter-provided, novel play event evidenced enhanced recall and greater resistance to retroactive interference from an intervening experience than did children who were not interviewed after the experience. Alternatively, it is possible that the enhancement provided by the procedural narrative in the VCUG study resulted in more extensive but weaker encoding, and hence the additional information was not maintained in memory. Nonetheless, these data provide further evidence that children’s understanding of an experience facilitates their encoding of the event. Understanding and Remembering of Stressful Experiences

41

Narrative

No Narrative

1 0.9

Proportion Recall

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Initial

6-Week

16-Week

Timing of Interview

Figure 2–3. Proportion of standard features of the VCUG procedure recalled at three different delay intervals by children in two post hoc narrative conditions.

Changing Expectations and Memory of a Soccer Match In another investigation from our laboratory, Eaton (2003) explored the ways in which children’s changing understanding of an experience may alter their goals and hence their deployment of attention during an event. In her dissertation research, Eaton examined the memory of 69 ten-year-old athletes of their final game in an end-of-season soccer tournament that culminated in a league championship. The details of play were noted during the match by observers who had varsity-level soccer experience, and video recordings were made to provide a means of verifying additional information. Memory for details of the match was assessed immediately after the game in interviews conducted in tents set up on the periphery of the field, and six weeks later in the children’s homes. The interviews began with a request for a narrative account of the match, followed by the set of wh- questions regarding predefined game components. The player’s free recall reports were divided into propositions, which were classified as reflecting central or peripheral game components. Applying Peterson and Bell’s (1996) definitions, central propositions referred to aspects of the game that directly affected play

42

STRESS AND MEMORY, EMPIRICAL EVIDENCE

or the outcome, whereas peripheral components were outside the match or did not affect play. The standard set of wh- questions referenced predetermined match components, and these elicited recall probes were similarly comprised of items addressing central features (e.g., “What was the final score?”) and peripheral features (e.g., “What color uniform did the other team wear?”). In addition to the memory assessments, each player also provided ratings of the importance of the event, perceived performance (both individual and for the team as a whole), and an estimate of the point at which the participant was confident of the outcome of the game. In addition, the children’s emotional reactions prior to and subsequent to play were obtained, using a context reinstatement technique based on the Cognitive Interview (see Fisher, Brennan, & McCauley, 2002). Participants on the winning and losing teams did not differ in their reported desire to win the game, and the ratings confirmed our expectation that the game was of great important to the young athletes. Indeed, more than a quarter of the children in the sample endorsed the statement: “I wanted to win as much as I’ve ever wanted anything in my life.” Not surprisingly, the participants on the winning teams reported that they were not confident of the outcome of the game until the second half, or even the final minutes of play. In contrast, the players from the losing teams could be classified into informal groups of those who identified the first half, second half, and end of play as the point at which they were certain that their team would lose. We predicted that children who believed a victory was still possible would be more vigilant and hence more focused on the central components of play than would those had believed that a loss was inevitable. At both the initial and delayed interviews, the results were consistent with this prediction. As illustrated in Figure 2–4, the players on the losing teams who reported believing into the second half or at the end of the game that a win was possible responded correctly to more wh- questions regarding central components of play (e.g., “How many times was a player called for off-sides?”; “How many times did the other team score on a penalty kick?”) than did those who thought the outcome had been decided in the first half. Some additional evidence supports the interpretation that the group differences arose from the deployment of attention. The groups did not differ significantly in their responses to questions addressing peripheral components of the game (e.g., “How many times did a ball from another game come onto your field?”), suggesting that the between-group differences in central recall cannot be attributed to a general disengagement among the children who saw the outcome as a foregone conclusion.

Understanding and Remembering of Stressful Experiences

43

0.8

Proportion Central Recall

0.7

Initial 6-Week Delay

0.6 0.5 0.4 0.3 0.2 0.1 0 1st Half

2nd Half

End

Knowledge of Outcome

Figure 2–4. Proportion of central game components reported by the children on losing teams who reported knowing the game outcome at alternative points in play.

Further, we compared the recall performance of the players on both losing and winning teams who reported that they did not know what the outcome of the game would be until the very end of play. These groups did not differ with regard to their recall of either the central or peripheral components of their games. This pattern of results suggests that the deployment of attention may mediate the relation between memory and arousal in children’s memory for salient experiences, at least under some conditions (see Easterbrook, 1959). The investigations described in this section are clearly limited by their quasi-experimental designs and by their small sample sizes. Nonetheless, taken as a whole, the work provides evidence for the importance of children’s understanding of their experiences on their encoding and subsequent retention of the event. More specifically, our studies of children’s memory for a physical examination and dental check-up indicate that prior knowledge influences performance, and our exploration of the VCUG provides evidence for the importance of recently acquired procedural information. In contrast, the soccer match study indicates clearly that children’s evolving understanding of the situation can affect the deployment of attention and hence subsequent recall; although all players were exposed to the same nominal stimulus—i.e., the tournament game—their differing perceptions of the

44

STRESS AND MEMORY, EMPIRICAL EVIDENCE

likely outcome led to contrasting functional stimuli, with corresponding differences in remembering.

A Novel Pediatric Examination: Expectations and Long-Term Recall Whereas the research reported above addresses the role of knowledge and understanding on the encoding of events as they transpire, these investigations did not include an examination of changes in the memory representation as a result of extended encoding. In this section, we explore the implications of some additional research for the understanding of possible knowledge-driven mechanisms through which these representations may be altered—in ways that may either disrupt or enhance veridical recall—after events have taken place. It has long been known that constructive processes driven by generic knowledge may alter children’s reports of their experiences over time. A study from our own research program illustrates the ways in which children’s expectations, based on their own past experiences, can affect their reports over an extended delay interval. Ornstein et al. (1998) presented four- and six-year-old children with a specially designed medical examination, administered by a licensed pediatrician in her office, in which some expected, typical features of the check-up were omitted and other novel features were included. Hence, for example, the physician omitted checking the child’s heart, but added the unusual component of measuring the child’s head circumference. Both at the initial and the 12-week delayed interview, the children were more successful in reporting the typical as compared with the atypical features in response to open-ended prompts. However, at the delayed interview, the children experienced difficulty in differentiating between typical features that had and had not been in the mock examination than they did with atypical present versus omitted components. Further, in contrast to previous investigations in which spontaneous intrusions are quite rare, the children reported a number of features that had not been included in the specially constructed examination, all of which were typically included in routine pediatric check-ups. Moreover, the older children, who presumably had more generic knowledge regarding the event, generated the majority of these intrusions. These findings are consistent with investigations of children’s reliance on generic representations such as scripts as episodic memory fades (Myles-Worsley, Cromer & Dodd, 1986) and with research examining the effects of event repetition on children’s suggestibility (e.g., Roberts & Powell, 2006). In addition to changes in memory over time that arise from the activation of existing knowledge, it is possible that knowledge-based Understanding and Remembering of Stressful Experiences

45

reasoning and interpretive processes may contribute to changes over time in the underlying memory representation. At present, few studies have documented the mnemonic consequences of such cognitive activity for children’s event recall. Tsethlikai and Greenhoot (2006) recently explored these issues in the context of nine- to 11-year-olds’ recall of stories. These children initially heard a story that they were told to imagine had happened to them. The participants subsequently heard either a control story or one in which another person’s perspective on the events in the story justified a change in its original interpretation. In response to the provision of this alternative perspective, the children in the experimental condition substantially changed their initial recall by presenting a character in a more positive light and attempting to reconcile details from both perspectives by adding new information to their reports. Some recent work in our laboratory, although by no means definitive, has implications for examining the ways in which changes in understanding an emotionally salient experience may alter its memory. We begin with the presentation of some findings from our investigations of children’s memory for soccer games. The data, we believe, provide evidence for a dynamic relationship between understanding and remembering over time.

Constructive Activity among Soccer Players on Winning vs. Losing Teams The findings from another study of soccer players’ recall of the details of an important match suggest a linkage between constructive activity and the emotion associated with the game. Baker-Ward, Eaton, and Banks (2005) asked 48 ten-year-old girls who participated in a soccer tournament to rate their individual performance on a 1 (“My worst game ever”) to 5 (“My best game ever”) Likert scale, both at an initial interview conducted within one week of the final game and at a delayed interview five weeks later. Not surprisingly, the players on winning teams rated the quality of their individual performance significantly higher than did their less-successful opponents. Interestingly, there was an interaction between team outcome and time. The children who enjoyed a victory rated their performance slightly higher at the second interview than at the first, whereas the young athletes who coped with a loss saw themselves as having played more poorly as time progressed. Additional results from this investigation indicate that the emotional valence of an experience is associated with the opportunity for extended encoding. At the delayed interview, a subset of the young soccer players was asked, “How often did you talk with someone about the game?” These respondents had all participated in the same tournament game 46

STRESS AND MEMORY, EMPIRICAL EVIDENCE

as members of opposing teams; hence, differences in their engagement in conversations about the event should be due to the game outcome, rather than to extraneous variations in their experiences (such as extreme weather or the attendance of a local celebrity at the game). The children who played on the winning team reported that they had discussed the game with others an average of 4.67 times over the five weeks between the match and the delayed interview. In contrast, the players from the losing team reported only about half as many conversations (M = 2.25). Given the importance of parent-child conversations about salient experiences on children’s memory (see Sales, this volume) and research on the effects of story retellings on adults’ recall performance (Marsh, 2006), it seems reasonable to predict that a different propensity to engage in conversations about an experience would affect long-term memory for the event. Moreover, although we do not have access to the specifics of these conversations, it seems likely that members of winning as opposed to losing teams talked about different aspects of the match.

Evaluations of Play and Extended Coding among Young Athletes As discussed above, Eaton (2003) demonstrated that children’s changing perceptions of the likelihood of winning or losing a soccer match influenced attentional deployment and the encoding of game-related information. Further results from this study suggest a linkage between the players’ personal evaluations of their team’s performance and their postgame focus over an extended period of time on alternative components of the match. As a general indicator of the young athletes’ reaction to the game the players rated their team’s performance (as well as the quality of their individual play) on a scale of 1 (“Worst game ever”) to 7 (“Best game ever”). Four groups were formed on the basis of these ratings, with ns ranging from 10 to 29 children. Because of the infrequent use of very negative ratings, children who reported one of the lower three ratings were included in one group who perceived team play during the match as “OK” to “Worst Ever.” Additional groups perceived team play as “Pretty Good,” “Really Good,” or “Best Ever.” Rating category was independent of the game outcome. The children’s memory for details of the match was assessed immediately after play and again after a delay of six weeks. As discussed above, the young players first provided a narrative account of the match, and then responded to a series of standard wh- questions regarding predefined game components. The free-recall reports were divided into propositions, which were classified as reflecting central or peripheral game components. In an analysis of the proportion of central propositions in the children’s Understanding and Remembering of Stressful Experiences

47

game narratives, the 19 children who rated the tournament game as their team’s “best game ever” differed across time from the groups who saw the game as less salient. These participants used a greater proportion of central propositions at the second interview (M = .61, SD = .27) than they did at the first (M = .42, SD = .29), whereas the children who evaluated play less positively did not differ over time in the proportion of central propositions they included in their reports. The groups did not differ with regard to the length of the narratives, eliminating a possible confounding variable. Moreover, they did not differ with regard to their rates of inclusion of propositions coded as peripheral information. With regard to elicited recall, the groups who rated team performance higher also tended (ps < .06) to respond correctly to a greater proportion of wh- questions addressing both central game components than did the other groups, although this effect did not change across time. Although the above findings reflect the operation of extended encoding processes, there are a number of reasons to expect that extended encoding may be limited in children’s reflection on the outcomes of recreational sports competitions. In remembering soccer games, the children in our investigations reported an expected, familiar experience. Moreover, although the soccer tournaments in which they took part were clearly important occurrences for the participants at the time at which they took place, the outcome is unlikely to have had a continuing effect on the young players’ lives. The tournaments in which they competed were organized by their recreational league, in which participation and good sportsmanship rather than winning were emphasized. In addition, the young athletes had extensive exposure to wins and losses. Parents of the participants in one of our soccer studies reported that their 10-year-old children had, on average, more than five years of experience in playing organized soccer. And, for these young players as well as sports fan everywhere, “there is always next year.” Consequently, understanding the experience in most cases may not require extended reflection or the provision of additional information from others. In contrast, other experiences, such as events that result in blocking the individual’s goals (see Stein & Liwag, 1997) or experiences that impose threats to the individual’s sense of selfworth or that are inconsistent with previous perceptions of others (see Ross, 1989) may be more likely to activate extended encoding.

Changes in Children’s Injury Reports over a One-Year Delay To explore the possibility of an increased likelihood of extended encoding in more emotionally challenging situations, we returned to the data obtained in a previous investigation of children’s memory for the details 48

STRESS AND MEMORY, EMPIRICAL EVIDENCE

of emergency treatment following a facial laceration (Burgwyn-Bailies et al., 2001). In this study, 24 three- to seven-year-olds experienced an injury that required treatment by a plastic surgeon, and we assessed their memory within two days of the accident and after delays of six weeks and one year. Although our interview protocol included open-ended and wh- questions regarding the children’s accidents as well as their surgical treatment, the report by Burgwyn-Bailes et al. (2001) focused only on the minor surgeries; indeed, the children’s accounts of their accidents have not been previously systematically examined. Of relevance to the present issue, the accidents had consequences for the children’s subsequent experience in at least some cases, as caregivers implemented new rules or strategies to insure their children’s future safety. For example, a six-year-old who cut his cheek with his new Cub Scout knife had to surrender his “whittling chip” and could no longer participate in some den activities. In other instances, the aftermath of the injury included questions of blame or fault. The children’s reports of 15 components of the injuries that they experienced were obtained, including nine central (e.g., the ongoing activity at the time of the injury) and six peripheral (e.g., weather, clothing the child was wearing) features. The initial reports were accurate and fairly extensive, with the young participants reporting an average of 10.8 of these 15 components. Few changes in these reports were observed at the six-week interview, but after the one-year delay the children reported about 70% of the features of the accidents that had been reported initially, omitting an average of 1.79 of the features. Little new information was provided after a year; the children reported a mean of only 0.49 previously undisclosed accident components. Interestingly, there were some contradictions in long-term recall such that each participant on the average altered the report by 1.38 components in a manner that was inconsistent with the original report. Although about two-thirds of these contradictions involved peripheral information, central features of the experience were also altered, including something as basic as the cause of the injury. Indeed, 10 children’s accounts included at least one such potentially consequential contradiction across the one-year interval. To illustrate, a second grade student provided this account of his injury one day after it occurred: I was playing basketball with my teacher and I was trying to steal. She sort of accidentally . . . you know, she was running and I just came across like this and we sort of collided. And she outweighed me. I got the worst of it and hit my head on the pavement.

One year later, the child gave us this report of his injury:

Understanding and Remembering of Stressful Experiences

49

I was playing pass the ball with my teacher. The ball hit the rim. I ran to get it and the teacher . . . I tripped over her foot. I fell and hit my head against the ground.

Note that the teacher’s role in the incident has been altered in the delayed report. She is described at the first interview as the unintentional agent of the injury, having inadvertently knocked down her student when they collided. In contrast, the child describes her role as passive in the delayed report; it is his own action in tripping over her foot that caused his fall. Could it be that serving as an agent of an injury that required emergency treatment, even inadvertently, was inconsistent with this child’s perception of a nurturing teacher, prompting a reconstruction of this episode? The complexity involved in the development of the understanding of intentional versus unintentional actions in assigning blame (Kohlberg, 1976) could increase the likelihood of such reworking of the experience. Of course, examples such as this one can only yield hypotheses, not conclusions. To explore more fully the impact of changing understanding over time on children’s memory, future research efforts should explore changes in memory for complex experiences that have consequences that extend beyond the temporal duration of the events themselves.

Future Directions In this chapter, we have suggested that the processes involved in the encoding of information in memory are knowledge-driven and extended in time. Children’s understanding of the events that they experience is critical for subsequent remembering, and comprehension is driven by both endogenous and exogenous forces that may operate long after an event has ended. In turn, our understanding of children’s abilities to remember events, especially those that are emotionally laden, requires an analysis of the factors that influence the establishment and modification—through extended encoding—of representations in memory. We readily acknowledge the limitations in the research examples that we have presented, including the post hoc nature of a number of the analyses, the small sample sizes in many studies, and, significantly, our limited measurement and conceptualization of emotion and stress (see Alexander and O’Hara, this volume, for discussion of the operationalization of stress and emotion). These important qualifications notwithstanding, we have provided evidence for the importance of exposure to various sources of understanding on children’s encoding of an experience. These influences include adults’ efforts to prepare children for specific events and children’s perceptions of critical aspects of an 50

STRESS AND MEMORY, EMPIRICAL EVIDENCE

experience as it unfolds. Regarding the effects of preparation on understanding, further studies should examine the changes in children’s knowledge representations that result from these efforts (see Baker-Ward et al., 1997). An understanding of the dynamics of changing memory representations should facilitate more precise explorations of linkages between understanding and remembering. In addition, our preliminary examination of the relation between presumed changes in children’s goals during a soccer match and the encoding of the details of this experience suggests the importance of developing methods for “on-line” assessments as events unfold. The initial explorations of children’s extended encoding provided here suggest that participants’ varying emotional reactions to an event can influence the likelihood of subsequent experiences that can, in turn, alter understanding and memory. Young soccer players who enjoyed a victory reported that they were more likely to engage in retellings of the game than were those who endured a defeat. Moreover, individuals’ perceptions of their own performance during the game were altered as a consequence of the game outcome. Further, some preliminary data suggest that young athletes who evaluated their play more positively, regardless of the outcome of the competition, focused more over time on central components of the event. Clearly, the hypotheses resulting from these findings must be tested in investigations designed specifically for this purpose. Such additional research must document children’s precise emotional reactions and assess contemporaneously their diverse experiences in revisiting the target event. To elucidate the processes involved in extended encoding, it will be important to explore changes in children’s understanding in the context of compelling experiences that have continuing significance for the individual. Even a championship recreational league match may have few continuing consequences, but a competition that results in a decision about the direction of a young athlete’s further athletic career (e.g., progressing to a competitive travel league or remaining on a recreational league) may require further interpretation and reassessment. Much remains to be learned about linkages between understanding and memory across development. Given our commitment to a constructivist approach to memory, we have focused on age-related changes in domain-relevant knowledge as contributors to corresponding changes in remembering (see Ornstein et al., 1998, for a review). We still continue to examine knowledge as one of a number of mediators of developmental change (Ornstein et al., 2006), but our perspective on knowledge has been broadened, and we find it useful to emphasize that understanding involves more than knowing. In addition to the contributions of expertise, our emerging perspective mandates that we direct our attention to a number of additional domains of development. For example, changes Understanding and Remembering of Stressful Experiences

51

across age in children’s general reasoning abilities contribute in important ways to developmental changes in understanding, as the capacity to link components of an experience in a logical method may enhance subsequent remembering (Baker-Ward et al., 1997). In addition, developmental transitions in the likelihood of the occurrence of “efforts after meaning,” individuals’ propensity to revisit and reinterpret emotionally significant events, must be examined to determine changes over time in understanding (see Fivush & Baker-Ward, 2005). Although few developmental analyses are currently available, Pennebaker and Stone’s (2003) identification of changes in language use over the lifespan supports the importance of such explorations. Children between eight and 14 years of age, in contrast to a group of 15- to 24-year-old participants, mentioned fewer cognitive processes—including causal and insight terms—in their personal narratives. It also seems likely that younger children’s exposure to external sources of information, including conversations with caregivers, has particular significance for their understanding of emotional experiences, both before and after they transpire. Clearly, cross-sectional and ultimately longitudinal analyses of children’s understanding and remembering are needed.

Implications for Practice Researchers in children’s memory, particularly those whose work is motivated in part by the needs of clinical and legal professionals in interpreting eyewitness testimony, have long been interested in the relation between levels of stress at encoding and the extent of recall and suggestibility (see, e.g., Gordon et al., 1995). The chapters in this volume attest to the complexity of the issues involved in addressing the stress-memory linkage and the importance of examining moderators of performance. Clearly, one of these important moderators is the child’s understanding of the event as it unfolds. Research on the linkage between children’s understanding and their subsequent memory of stressful experiences has relevance for both forensic development and pediatric psychology. Given the need for future research, as discussed above, we outline here two basic implications of our work. First, assessments of children’s testimony must incorporate an evaluation of the child’s understanding of the to-be-reported event. At the time that an experience transpires, variations in children’s understanding may partially mediate age differences in performance. Even younger preschoolers, in the presence of domain-relevant information, may provide relatively extensive accounts of what has happened to them. Without such understanding, their reports may be limited, particularly with regard 52

STRESS AND MEMORY, EMPIRICAL EVIDENCE

to free recall, and more extensive cueing may be needed to elicit the information that was encoded. Given the risks of response bias, especially among very young children (see Baker-Ward, Ornstein, Gordon, Follmer, & Clubb, 1995), such cueing must be provided cautiously, with a reliance on closed rather than specific questions whenever possible. The conceptualization of encoding as a process that extends beyond the event also conveys the importance of examining experiences that may maintain or augment memory, as well as those that may disrupt recall. Some apparent inconsistencies in children’s reports over time, for example, may represent changes in the understanding of the relevance of aspects of an experience, and hence may not constitute evidence of memory contamination (see Gordon & Follmer, 1994). Opportunities for retelling a story about an experience in informal situations as well as formal interviews must also be examined. Further, the extent to which therapeutic interventions may have transformed a child’s perspective on an experience, and hence the underlying memory representation, must be carefully evaluated. Second, enhancing children’s understanding of a forthcoming experience can be expected to affect the likelihood that they will experience anxiety during the event. In general, greater levels of understanding will certainly decrease overall anxiety, and a recent investigation by Salmon, McGuigan, and Pereira (2006) provides direct support for this generalization. These researchers extended our analysis of the effects of information provided by the technologists in conjunction with an invasive VCUG procedure (Principe et al., 1997), as discussed above. Young children were randomly assigned to conditions differing with regard to the provision of information about the procedure both before it was implemented and as it unfolded. The participants who received complete information as well as distraction, in comparison to groups receiving only partial information and distraction or no psychosocial intervention, demonstrated lower levels of observed behavioral distress during the VCUG procedure and better memory performance after a delay of a week. The findings suggest an effective and efficient means of decreasing children’s anxiety and increasing their compliance during medical procedures. It should be emphasized, however, that building understanding rather than simply providing knowledge is necessary to obtain such benefits. Salmon (2006), in an application of laboratory-based research to the preparation of children for medical procedures, concludes that information must be specifically rather than generally related to the event. Hence, parental discussions that convey reassurance and general instructions for behavior may not prepare children in an effective manner. It may also be the case that preparation has a cost, at least during the initiation of a medical procedure. In our investigation of the effects of preparation on children’s memory for a first dental visit as Understanding and Remembering of Stressful Experiences

53

discussed above (Baker-Ward & Ornstein, in preparation), children in the High Preparation group actually had significantly greater levels of anxiety during the first portion of the examination, as measured through a behavioral-observation coding scheme, than did children receiving less preparation. Of course, the post hoc nature of the design mandates caution in interpreting this finding, and it is possible that parents provided more preparation to children whom they perceived as more anxious. However, parental ratings of child trait and state anxiety and temperament did not differentiate the preparation groups. It must be emphasized that overall levels of anxiety were low and that the children were quite compliant during the dental examination. Potential collaborations with pediatric psychologists, child-life specialists, and medical professionals offer opportunities to establish causal linkages between stress and remembering. Whereas ethical practice prohibits creating high levels of anxiety in most situations involving children, it is certainly appropriate to work toward reducing children’s stress when they face painful medical procedures. Preparation has long been recognized as a means of alleviating anxiety, and it is apparent that effective preparation enhances understanding. In addition to identifying the variables associated with the most effective interventions, research addressing children’s preparation for difficult experiences can enable the determination of the pathways through which stress at encoding and during retention impacts on remembering. But one thing is clear: the complex relations between stress and memory can be elucidated only when children’s changing understanding of challenging experiences is addressed.

Acknowledgment: The research on children’s memory for medical procedures discussed in this chapter was supported in part by grant HD 32214 to Peter A. Ornstein from the United States Public Health Service. We also extend our appreciation to the medical professionals who cooperated with these investigations, the recreation program staff members and community volunteers who facilitated our efforts, and the parents and children who willingly participated. The research we review represents the continuing involvement of members of our laboratory groups at UNC-CH and NC State, and we gratefully acknowledge their contributions to our research programs. Kimberly L. Eaton’s valued collaboration made possible the research on children’s memory for soccer games. Thanks are also due Elaine Burgwyn-Bailes and Mary Koenig Styers for their assistance in the preparation of this chapter.

54

STRESS AND MEMORY, EMPIRICAL EVIDENCE

Address correspondence to Lynne Baker-Ward, Department of Psychology, NC State University, Raleigh, NC 27695-7650. E-mail: [email protected]. References Baker-Ward, L., Eaton, K. L., & Banks, J. B. (2005). Young soccer players’ reports of a tournament win or loss: Different emotions, different narratives. Journal of Cognition and Development, 6, 507–527. Baker-Ward, L., Gordon, B. N., Ornstein, P. A., Larus, D. M., & Clubb, P. A. (1993). Young children’s long-term retention of a pediatric examination. Child Development, 64, 1519–1533. Baker-Ward, L., & Ornstein, P. A. (in preparation). The effects of naturallyoccurring and experimenter-provided preparation on three-year-old children’s memory for a first dental visit. Baker-Ward, L., Ornstein, P. A., & Principe, G. F. (1997). Revealing the representation: Evidence from children’s reports of events. In P. van den Broek, P. Bauer, & T. Borg (Eds.), Developmental spans in event comprehension and representation: Bridging fictional and actual events (pp. 79–107). Mahwah, NJ: Lawrence Erlbaum Associates. Baker-Ward, L., Ornstein, P. A., Gordon, B. N., Follmer, A., & Clubb, P. A. (1995). How shall a thing be coded? Implications of the use of alternative procedures for scoring children’s verbal reports. In M. S. Zaragoza, J. R. Graham, G. C. N. Hall, R. Hirschman, & Y. S. Ben-Porath (Eds.), Memory and testimony in the child witness (pp. 61–85). Thousand Oaks, CA: Sage. Barclay, J. R., & Reid, M. (1974). Semantic integration in children’s recall of discourse. Developmental Psychology, 10, 277–281. Bartlett, F. C. (1932). Remembering: A study in experimental and social psychology. New York, NY: Cambridge University Press. Binet, A. (1900) La suggestibilité [On suggestibility]. Paris, France: Schleicher. Binet, A., & Henri, V. (1894a). La mémoire des mots. L’Année Psychologique, 1, 1–23. Binet, A., & Henri, V. (1894b). La mémoire des phrases (mémoire des idées). L’Année Psychologique, 1, 24–59. Brainerd, C. J., & Ornstein, P. A. (1991). Children’s memory for witnessed events: The developmental backdrop. In J. Doris (Ed.), The suggestibility of children’s recollections: Implications for eyewitness testimony (pp. 10–20). Washington, DC: American Psychological Association. Bransford, J. D., & Franks, J. J. (1972). The abstraction of linguistic ideas: A review. Cognition, 1, 211–249. Bruck, M., & Ceci, S. J. (1999). The suggestibility of children’s memory. Annual Review of Psychology, 50, 419–439. Bruner, J. S. (1957). On going beyond the information given. In J. S. Bruner, E. Brunswik, L. Festinger, F. Heider, K. F. Muenzinger, C. E. Osgood, & D. Rapaport (Eds.), Contemporary approaches to cognition (pp. 41–69). Cambridge, MA: Harvard University Press. Bruner, J. S. (1990). Acts of meaning. Cambridge, MA: Harvard University Press.

Understanding and Remembering of Stressful Experiences

55

Burgwyn–Bailes, E., Baker-Ward, L., Gordon, B. N., & Ornstein, P. A. (2001). Children’s memory for a minor medical emergency procedure after one year: Individual differences in recall and suggestibility. Applied Cognitive Psychology, 15, 1–24. Campbell, B. A., & Jaynes, J. (1966). Reinstatement. Psychological Review, 73, 478–480. Campos, J. J., Mumme, D. L., Kermoian, R., & Campos, R. G. (1994). A functionalist perspective on the nature of emotion. Monographs of the Society for Research in Child Development, 59 (2–3. Serial No. 240), 284–303. Chi, M. T. H. (1978). Knowledge structures and memory development. In R. S. Siegler (Ed.), Children’s thinking: What develops? (pp. 73–96). Hillsdale, NJ: Lawrence Erlbaum Associations. Clubb, P. A., Nida, R. E., Merritt, K. & Ornstein, P. A. (1993). Visiting the doctor: Children’s knowledge and memory. Cognitive Development, 8, 361–372. Dolan, R. J. (2002). Emotion, cognition, and behavior. Science, 298 (5596), 1191–1194. Easterbrook, J. A. (1959). The effect of emotion on cue utilization and the organization of behavior. Psychological Review, 66, 183–201. Eaton, K. (2003). Memory and emotion: The influence of valence on children’s memory for a salient event. Unpublished doctoral dissertation, NC State University, Raleigh. Fisher, R. P., Brennan, K. H., & McCauley, M. R. (2002). The cognitive interview method to enhance eyewitness recall. In M. L. Eisen, J. A. Quas, & G. S. Goodman (Eds.), Memory and suggestibility in the forensic interview (pp. 265–286). Mahwah, N. J.: Lawrence Erlbaum Associates. Fivush, R. (2000). Accuracy, authorship, and voice: Feminist approaches to autobiographical memory. In P. Miller & E. Scholnick (Eds.), Towards a feminist developmental psychology (pp. 85–106). New York, NY: Cambridge University Press. Fivush, R. & Baker-Ward, L. (2005). The search for meaning: Developmental perspectives on internal state language in autobiographical memory. Journal of Cognition and Development, 6, 455–462. Goodman, G. S., Quas, J. S., Batterman-Faunce, J. M., Riddlesberger, M. M., & Kuhn, J. (1997). Children’s reactions to and memory for a stressful event: Influences of age, anatomical dolls, knowledge and parental attachment. Applied Developmental Science, 2, 54–74. Gordon, B. N., & Follmer, A. (1994). Developmental issues in judging the credibility of children’s testimony. Journal of Clinical Child Psychology, 23, 283–294. Gordon, B. N., Schroeder, C. S., Ornstein, P. A., & Baker-Ward, L. (1995). Clinical implications of research on memory development. In T. Ney (Ed.), Child sexual abuse cases: Allegations, assessment and management (pp. 99–124). New York, NY: Brunner/Mazel. Gordon, B. N., Baker-Ward, L., & Ornstein, P. A. (2001). Children’s testimony: A review of research on memory for past experiences. Clinical Child and Family Psychology Review, 4, 157–181. Greenhoot, A. F. (2000). Remembering and understanding: The effects of changes in underlying knowledge on children’s recollections. Child Development, 71, 1309–1328. 56

STRESS AND MEMORY, EMPIRICAL EVIDENCE

Greenhoot, A., Ornstein, P. A., Gordon, B. N., & Baker-Ward, L. (1999). Acting out the details of a pediatric check-up: The effects of interview condition on individual differences in children’s memory reports. Child Development, 70, 363–380. Hudson, J. A., & Sheffield, E. G. (1999). The role of reminders in young children’s memory development. In L. Balter & C. S. Tamis-LeMonda (Eds.), Child psychology: A handbook of contemporary issues (pp. 193– 214). New York, NY: Psychology Press. Kohlberg, L. (1976). Moral stage and moralization: The cognitivedevelopmental approach. In T. Lickona (Ed.), Moral development and behavior: Theory, research, and social issues (pp. 84–107). New York, NY: Holt, Rinehart & Winston. “I Can’t Sing So I Paint” [Interview with Georgia O’Keeffe]. (1922: December 5). New York Sun. Lazarus, R. S. (1977). Cognitive and coping processes in emotion. In A. Monat & R. S. Lazarus (Eds.), Stress and coping (pp. 145–158). New York, NY: Columbia University Press. Loftus, E. F. (1993). Made in memory: Distortions in memory after misleading communications. In D. L. Medin (Ed.), The psychology of learning and motivation: Advances in research and theory. Vol. 30 (pp. 187– 215). San Diego, CA: Academic Press. Marsh, E. L. (2007). Retelling is not the same as recalling. Psychological Science, 16, 16–25. McEwan, B. S. (2000). The neurobiology of stress: From serendipity to clinical relevance. Brain Research, 886, 172–189. Melamed, B. G., & Lumley, M. A. (1988). Behaviour profile rating scale. Dictionary of behavioral assessment. Oxford, England: Pergamon. Merritt, K. A., Ornstein, P. A., & Spicker, B. (1994). Children’s memory for a salient medical procedure: Implications for testimony. Pediatrics, 94, 17–23. Myles-Worsley, M., Cromer, C. C., & Dodd, D. H. (1986). Children’s preschool script reconstruction: Reliance on general knowledge as memory fades. Developmental Psychology, 22, 22–30. Ornstein, P. A., Baker-Ward, L., Gordon, B. N., & Merritt, K. A. (1997). Children’s memory for medical experiences: Implications for testimony. Applied Cognitive Psychology, 11, S87–S104. Ornstein, P. A., Baker-Ward, L., Gordon, B. N., Pelphrey, K. A., Tyler, C. S., & Gramzow, E. (2006). The influence of prior knowledge and repeated questioning on children’s long-term retention of the details of a pediatric examination. Developmental Psychology, 42, 332–344. Ornstein, P. A., Larus, D., & Clubb, P. A. (1991). Understanding children’s testimony: Implications of research on the development of memory. In R. Vasta (Ed.), Annals of Child Development (pp. 145–176). London, England: Jessica Kingsley. Ornstein, P. A., Merritt, K. A., Baker-Ward, L., Furtado, E., Gordon, B. N., & Principe, G. F. (1998). Children’s knowledge, expectation, and long-term retention. Applied Cognitive Psychology, 12, 387–405. Ornstein, P. A., Principe, G. F., Hudson, A. E., & Gordon, B. N., & Merritt, K. A. (1997, April). Procedural information and stress as mediators of children’s long-term recall. In J. A. Quas & G. S. Goodman (Chairs), Individual differences in children’s memory and Understanding and Remembering of Stressful Experiences

57

suggestibility: New research findings and directions. Symposium paper presented at the meetings of the Society for Research in Child Development, Washington, D. C. Ornstein, P. A., Shapiro, L. R., Clubb, P. A., Follmer, A., & Baker-Ward, L. (1997). The influence of prior knowledge on children’s memory for salient medical experiences. In N. Stein, P. A. Ornstein, C. J. Brainerd, & B. Tversky (Eds.), Memory for everyday and emotional events. (pp. 83–111). Mahwah, NJ: Lawrence Erlbaum Associates. Paris, S. G., & Lindauer, B. K. (1976). The role of inference in children’s comprehension and memory for sentences. Cognitive Psychology, 8, 217–222. Pennebaker, J. W., & Stone, L. D. (2003). Words of wisdom: Language use over the life span. Journal of Personality and Social Psychology, 85, 291–301. Peterson, C. & Bell, M. (1996). Children’s memory for traumatic injury. Child Development, 67, 3045–3070. Principe, G. F., Kanaya, T., Ceci, S. J., & Singh, M. (2006). Believing is seeing: How rumors can engender false memories in preschoolers. Psychological Science, 17, 243–248. Principe, G. F., Myers, J. T., Furtado, E., Merritt, K., & Ornstein, P. A. (1996, March). The relationship between procedural information and young children’s recall of an invasive medical procedure. In L. Baker-Ward (Chair), The role of individual differences in young children’s reports of salient personal experiences. Conference on Human Development, Birmingham, AL. Principe, G. F., Ornstein, P. A., Baker-Ward, L., & Gordon, B. N. (2000). The effects of intervening experiences on children’s memory for a physical examination. Applied Cognitive Psychology, 14, 59–80. Roberts, K. P., & Powell, M. B. (2006). The consistency of false suggestions moderates children’s reports of a single instance of a repeated event: Predicting increases and decreases in suggestibililty. Journal of Experimental Child Psychology, 94, 68–89. Roediger, H. L. III, & McDermott, K. B. (1995). Creating false memories: Remembering words not presented on lists. Journal of Experimental Psychology: Learning, Memory and Cognition, 21, 803–814. Ross, M. (1989). Relation of implicit theories to the construction of personal histories. Psychological Review, 96, 341–357. Salmon, K. (2006). Commentary: Preparing young children for medical procedures: Taking account of memory. Journal of Pediatric Psychology, 31, 859–861. Salmon, K., McGuigan, F., & Pereira, J. K. (2006). Optimizing children’s memory and management of an invasive medical procedure: The influence of procedural narration and distraction. Journal of Pediatric Psychology, 31, 522–527. Schneider, W., Körkel, J., & Weinert, F. E. (1989). Domain-specific knowledge and memory performance: A comparison of high- and lowaptitude children. Journal of Educational Psychology, 81, 306–312. Starnes, L. and Baker-Ward, L. (under review). An initial interview enhances pre-kindergarten children’s subsequent event recall and resistance to retroactive interference. Stein, N. L., & Liwag, M. D. (1997). Children’s understanding, evaluation, and memory for emotional events. In P. W. van den Broek, P. J. Bauer, & T. Bourg, (Eds.), Developmental spans in event comprehension and 58

STRESS AND MEMORY, EMPIRICAL EVIDENCE

representation: Bridging fictional and actual events (pp. 129–235). Mahwah, NJ: Lawrence Erlbaum Associates. Tsethlikai, M., & Greenhoot, A. F. (2006). The influence of another’s perspective on children’s recall of previously misconstrued events. Developmental Psychology, 42, 732–745. Underwood, B. J. (1963). Stimulus selection in verbal learning. In C. N. Cofer & B. S. Musgrave (Eds.), Verbal behavior and learning: Problems and processes (pp. 33–75). New York, NY: McGraw-Hill. Vandermaas, M. O., Hess, T. M., & Baker-Ward, L. (1993). Does anxiety affect children’s reports of memory for a stressful event? Journal of Applied Cognitive Psychology, 7, 109–127. Wang, Q., & Ross, M. (2005). What we remember and what we tell: The effects of culture and self-priming on memory representations and narratives. Memory, 13, 594–606.

Understanding and Remembering of Stressful Experiences

59

3 Injuries, Emergency Rooms, and Children’s Memory Factors Contributing to Individual Differences Carole Peterson Kelly L. Warren

T

he relationship between children’s emotional reactions to stressful events and their subsequent memory of those events is an important one, playing a role in both clinical and forensic situations. Considerable research has shown that children’s emotional reactions during an event are major contributors to how well it is later remembered. In a review of children’s recall for autobiographical events, whether or not there was robust long-term recall of these events was highly related to emotionality, and highly negative events seem to be particularly well-remembered (Peterson, 2002). As contrasting examples, preschoolers recalled approximately 20–30% of the features of a living-room camping trip when interviewed one day or three weeks later (Boland, Haden, & Ornstein, 2003; Haden, Ornstein, Eckerman, & Didow, 2001; Ornstein, Haden, & Hedrick, 2004), but 75% of the features of facial surgery events (caused by injuries) when they were interviewed a full year later (Burgwyn-Bailes, Baker-Ward, Gordon, & Ornstein, 2001). It should be noted, however, that the latter events involved acute stress; one cannot assume that similar relationships between stress and memory hold when stress is chronic versus acute. This issue is explored by Greenhoot, Johnson, Legerski, and McCloskey (this volume). In this chapter, we will first present some recent theoretical discussion of the relationship between stress and memory, and then summarize a body 60

of research conducted in our laboratory that has explored children’s recall of naturally occurring stressful events, specifically injuries serious enough to require hospital emergency-room treatment. In particular, we will focus on potential sources of individual variation in children’s recall of real-life stressful events, both in research conducted in our laboratory and in related research conducted by other investigators. Finally, we will discuss the practical significance of this work for clinical and legal contexts.

Theoretical Considerations There has been considerable debate about the relationship between acute stress and memory in children with various studies differing in empirical findings. According to a recent meta-analytic review of this research, part of this variation may be attributable to the nature of the distressing events (Deffenbacher, Bornstein, Penrod, & McGorty, 2004). Deffenbacher et al. (2004) distinguish between events that elicit an arousal mode of attention control (i.e., an orienting response, or high level of attention focused on the event) and ones that elicit an activation mode of attention control (i.e., a defensive response such as the well-known fight or flight response). A defensive response is elicited by events that threaten bodily integrity or self-esteem, and thus involve considerably higher degrees of distress than events that elicit an orienting response. Deffenbacher et al. argue that some studies investigating the relationship between stress and memory had procedures that elicited an orienting response while others elicited a defensive response. Thus, it may be difficult to compare the effects of stress on memory when there is such variation in what constitutes the “high stress” category in different research studies. As an example of this variation, Peters (1997) exposed children to an unexpected fire alarm. Although children in his highest stress group had elevated blood pressure and pulse rates, none of the children cried or showed hysterical distress. In contrast, in a series of studies of children who suffered an unexpected and very painful injury (such as a broken bone, crushed fingers, or deep laceration), the high stress group was composed of children who suddenly began to scream in pain and were typically described by their parents as extremely upset or hysterical (Peterson, 1999; Peterson & Bell, 1996; Peterson & Whalen, 2001). In their review, Deffenbacher et al. (2004) propose a theoretical model of how stress affects memory; they suggest that as stress increases, memory for those details that are the focus of participants’ attention are increasingly recalled. However, when stress levels become very high, there is a catastrophic drop in memory performance. This model fits the data from extant studies of how accurately adults can recall event Injuries, Emergency Rooms, and Children’s Memory

61

details when the target events meet the authors’ criteria for eliciting a defensive response. Surprisingly, in the smaller number of studies that involved children, they found no relationship between stress and recall accuracy in children. The other outcome measure Deffenbacher et al. assessed was how accurately children and adults could identify faces from line-ups. Parallel to their findings on accuracy of event detail recall, they found that adults showed increases in recall as stress increased, until high levels of stress occurred. At that point, there was a similar sharp drop in the accuracy of face line-up identification. And, again, parallel to their findings for accuracy of event detail recall, there was no relationship between stress and how well children could identify faces from line-ups. However, Deffenbacher et al. did not assess the amount of information recalled about target events. The focus of this chapter is on children, and although (as those authors suggest) the accuracy of what children recall may not be compromised under high stress conditions, it is possible that how much they recall is. Overall, their model may fit data from adult studies but the relationship between stress and memory in children is still very unclear. In the next sections, we review research from our laboratory on children’s recall of high-stress events.

Children’s Recall of Injuries Requiring Emergency-Room Treatment Over the past decade, I and my colleagues have been exploring children’s recall of naturally-occurring events that are highly salient to children and that elicit a lot of distress, namely personal injuries such as bone fractures and lacerations that are serious enough to require hospital emergency-room treatment. In this research, children and their families are recruited during their emergency-room visit and, over the years, around 80% have agreed to participate. Because this research has been conducted in Canada, where medical care is paid for by the government and all children receive equivalent treatment regardless of their socioeconomic circumstances, the children represent a cross-section of their community. Although we could recruit families from the emergency room, ethically we could not interview them until they had had time to read our information and consider it at their leisure, so all child and parent interviews took place approximately a week later. And in order to get cooperation from most families, we had to go to them, in their homes, which is where all interviews have taken place. In this entire body of research on emergency-room injuries, it is clear that many of the children were extremely upset. Many were described by parents as hysterical. In the words of one child, “I never cried—I 62

STRESS AND MEMORY, EMPIRICAL EVIDENCE

just screamed.” A parent describing a three-year-old said, “He was screaming! Bawling! Lots of blood!” Many children also perceived their bodies to be threatened. To quote another child, “The blood was just flying everywhere.” So we were confident that we had an event that was highly distressing for at least some children, and as well, other children were not so upset. According to the differentiation made by Deffenbacher et al. (2004), between distressing events that elicit an orienting versus a defensive response, we are confident that for some of our children, the defensive response was elicited, on the basis of both parental and child self-descriptions of how distressed the children were. How to capture the child’s recall has been a perennial issue. The problem is that there was not a standardized event that all children experienced. Rather, each bone fracture, laceration, dog bite, or crushed finger differed in details. Probably the one that was most divergent was a four-year-old who had a wasp fly up her nose and got thoroughly lodged in the upper nasal passage. The wasp of course stung her painfully from the inside, and was sufficiently stuck that it had to be removed with forceps at the hospital. The child, not surprisingly, was hysterical the whole time. So, how to capture this variation and allow comparison across children? We devised a standardized prototype of typical injury and hospital treatment events, and children were scored on the proportion of these prototype components that they recalled with parental witnesses determining which components of the prototype were applicable. Examples of prototype components include where and when the injury took place, who was there, what led to the injury, how the child reacted, who first got to them, what that person did, and so on. Children were scored on whether or not they provided information relevant to each applicable prototype component. Thus, we could assess the completeness or exhaustiveness of children’s recall, in terms of the proportion of the relevant prototype components that they actually recalled. We could also assess the accuracy of their recall, again by comparing their recall with witness reports. Others who have studied children’s recall of naturally occurring stressful events (such as hurricanes) have looked at the absolute amount of information that children have recalled by counting the number of new or unique units of information (Bahrick, Parker, Fivush, & Levitt, 1998; Fivush, Sales, Goldberg, Bahrick, & Parker, 2004), and so in some studies we used this measure too. This has a different focus than the completeness measure described above. In assessments of completeness, each component of a prototypical experience is scored as present or not, regardless or how much detail the child provides about that component. In contrast, when the amount of new information is assessed, each new detail is separately counted. Compare “we were in the backyard” with “we were in the backyard down by that hole in the fence Injuries, Emergency Rooms, and Children’s Memory

63

where the mud is.” These are equivalent under the prototype completeness scoring (they both specify where events took place), but the latter contains considerably more new units of information. There are two things we should qualify. First, we have often checked to see if the nature of the injury, whether a bone fracture, laceration requiring sutures, or other injury, makes a difference, and it does not. The second qualification, however, does turn out to be important: in every study, children recalled significantly more about the details of their injury experience than about their hospital treatment, even though parental ratings showed that children were often as upset in the hospital as they were at the time of injury. There could be several potential explanations for this difference in how well these two events are recalled. For one thing, the injury is a unique event, whereas these children have visited this emergency room numerous times (see Peterson & Bell, 1996, for relevant data) since it serves as the after-hours clinic for all physicians in the region. As well, it is likely that children have a better understanding of the temporal and causal connections between successive injury than hospital components, since many probably do not understand

Injury

Hospital

100 90

% Components Recalled

80 70 60 50 40 30 20 10 0 Initial

5 year 12–13

8–9

Initial 5–6

3–4

5 year 2

Figure 3–1. Percentage of relevant prototype components recalled by children initially and five years later about their injury and hospital treatment. 64

STRESS AND MEMORY, EMPIRICAL EVIDENCE

why they go to one place in the hospital and wait, talk to someone there, and then get taken to another place (e.g., x-rays) and wait, then get taken somewhere else for another treatment component. In the next section, we will review the results of this series of studies, focusing on potential sources of individual variation.

Stressful Events are Well-Remembered Events A host of research demonstrates the robust nature of children’s recall for stressful events. For example, investigators who have studied children’s memory for highly stressful experiences that included painful medical procedures such as voiding cystourethrograms (VCUG) (Goodman, Quas, Batterman-Faunce, Riddlesberger, & Kuhn, 1994, 1997; Merritt, Ornstein, & Spicker, 1994; Quas et al., 1999; Salmon, Price, & Pereira, 2002) and natural disasters such as hurricanes (Bahrick et al., 1998; Fivush et al., 2004), have repeatedly shown that children have extensive memory for these stressful, highly salient events. Injuries serious enough to require hospital emergency room treatment are also well-recalled events. We have found, for completeness of recall, that children recalled on average (across all age groups) 75% of the prototype components of their injuries when interviewed a week after they occur, and 73% of those components a full five years later (Peterson & Whalen, 2001). They did not recall as many prototype components of their hospital treatment (57% initially and 50% five years later, see Figure 3–1), but recollection is still considerably better than for more mundane events that are often not recalled at all, or require considerable cuing (Peterson, 2002). For accuracy, children’s recall is quite accurate shortly after the target events occurred (averaging 94% accuracy for both events), and although accuracy deteriorates over time, accuracy of injury and hospital recall still averages 86% and 78%, respectively, five years later (see Figure 3–2). Although decreases were found five years later in the completeness and accuracy of recall, a different pattern was found for the amount of new information children provide, which increased over time. Children provided on average 52 new units of information initially about their injury and 46 new units about hospital treatment, and five years later they averaged 71 and 54 units of new information for their injury and hospital treatment, respectively (see Figure 3–3). This increase over time in the number of new units of information has been found by others as well (Fivush et al., 2004), and may reflect children’s improved vocabularies and narrative skills. As they get older, they provide more descriptive detail about the components of their experiences. In summary, stressful events are typically well-remembered events. Nevertheless, the relationship between stress and memory is more Injuries, Emergency Rooms, and Children’s Memory

65

Injury

Hospital

100 90 80

% Accuracy

70 60 50 40 30 20 10 0 Initial

5 year 12–13

Initial 8–9

5–6

5 year 3–4

2

Figure 3–2. Percentage of children’s recall about their injury and hospital treatment that is accurate, both initially and five years later.

complex than this because researchers have documented considerable individual variation between children’s recall of identical events. Because this relationship is complex, it is necessary to use a range of perspectives, from neurobiological (see Carver & Cluver, this volume, and Wiik & Gunnar, this volume) to physiological (see Wallin, Quas, & Yim, this volume, and Alexander & O’Hara, this volume), to children’s cognitive understanding (see Baker-Ward, Ornstein, & Starnes, this volume, Compas, Campbell, Robinson, & Rodriguez, this volume, and Laible & Panfile, this volume), to socioemotional perspectives (see Chae, Ogle, & Goodman, this volume, McDermott-Sales, this volume, Laible & Panfile, this volume, and Oppenheim & Koren-Karie, this volume, as well as research presented here). Below, we show via our data the well-known importance of the child’s age for how well events are recalled; we then discuss a number of other individual difference variables that have been investigated in our laboratory.

Importance of Children’s Age One factor that must be taken into account is the child’s age. Even when the target events being recalled seem equivalent, how old the child 66

STRESS AND MEMORY, EMPIRICAL EVIDENCE

is at the time of event occurrence makes a difference (Peterson, 2002; Quas et al., 1999; Salmon et al., 2002). Children less than two years of age when the events happened typically recall little or nothing of even highly stressful events, and what little they may recall tends to be fragmentary (Peterson & Rideout, 1998; Terr, 1988; see review in Peterson, 2002). Importantly, it is children’s age at the time of event occurrence rather than age at the time of event recall that is key; older children seem to retain only fragments of target events that occurred when they were very young, although some of them may weave these fragments together with memories from other related events to provide coherent narratives, even though these narrative accounts may be amalgamations of a number of separate events. We found this in our five-year follow-up interviews of injured former one- and two-year-olds (Peterson & Parsons, 2005). By the time children are three years of age, their recall of stressful events can be quite extensive and robust. However, it still becomes better with age, as researchers who have investigated a range of target events have found (Burgwyn-Bailes et al., 2001; Goodman et al., 1997; Peterson, 1999; Peterson & Bell, 1996; Peterson & Whalen, 2001; Salmon et al., 2002; Shrimpton, Oates, & Hayes, 1998). Figures 3–1 to 3–3 demonstrate age differences in how complete, accurate, and informative children’s recall of emergency-room injuries is across a wide age variation. The figures also show how recall changes over five years. For older children, it becomes less complete with age. However, this is not necessarily the case for younger children, particularly two-year-olds, who recall a greater proportion of injury components five years later. This increase is probably an artifact of how difficult it is to interview twoyear-olds. As anyone who has tried to interview children this young knows, they are much more interested in playing than sitting and talking, but five years later, they are cooperative seven-year-olds. These data on the completeness and accuracy of recall have been presented elsewhere (Peterson, 2002; Peterson & Whalen, 2001) and are reprinted here for comparison with data on the informativeness of children’s recall. As is apparent in Figure 3–3, children of all ages provide more descriptive detail five years later, particularly about injury components, even if they recall fewer of those prototype components. This increase in informativeness after delays spanning years has been found by others as well. For example, Fivush et al. (2004) found that children provided more new units of information about a destructive hurricane six years later than they had initially. These findings are consistent with the notion that memory and language are different although interacting skills. Although language can be used to encode as well as communicate aspects of memory, it is unlikely to capture the entirety of any given Injuries, Emergency Rooms, and Children’s Memory

67

Injury

Hospital

100 90 80

# Information Units

70 60 50 40 30 20 10 0 Initial

5 year 12–13

8–9

Initial 5–6

3–4

5 year 2

Figure 3–3. Number of new units of information provided initially and five years later.

memory. As children’s linguistic ability increases, the same memory may become verbally conveyed with more linguistic complexity, more descriptive detail, and so on. Thus, even though long-term recall may become less complete and less accurate, it nevertheless may contain more information.

Variation in Stress How to measure the degree of distress experienced by a child during stressful events has been a contentious issue. Some investigators advocate using only physiological measures (Peters, 1997), but these are not practical for naturally occurring events. Others have used externally validated measures such as the amount of damage done to one’s home by a destructive hurricane (Bahrick et al., 1998; Fivush et al., 2004). However, it is not clear that this is a good measure of the amount of distress experienced by an individual child. One child whose house’s damage is classified as moderate may have been terrified while another whose house fell into the same category of damage may have been much less upset. In our research with children recruited from the emergency room we have asked parents (who were witnesses to the child’s experience) 68

STRESS AND MEMORY, EMPIRICAL EVIDENCE

to rate their child’s distress on a Likert scale ranging from not at all distressed to extremely distressed. This rating was completed not only for the time of injury but also while the child was being treated at the hospital. Although the number of choices on the distress scale varied slightly, essentially we classified children’s level of stress as low, moderate, or high. But what about the child’s own version of how distressed he or she was? Steward et al., in their 1996 monograph concerning interviewing children about body touch during medical exams, stress the importance of looking at children’s own feelings about what happened to them. Unfortunately, with the exception of one study, we did not ask children for their own estimates of distress. However, all of them were asked about whether or not (and how much) they had cried. A child who stated that she had been “very very crying,” or another who replied to the question “What happened in the hospital?” with “Well, I was screeching a lot” are communicating considerable information about their emotional state at the time of event occurrence, as are children who stated that they didn’t cry. (By the way, we had parental confi rmation on crying.) We classified children’s level of distress on the basis of their selfdescriptions of crying as low, moderate, or high, and the correlations between children’s self-descriptions of crying and their parents’ ratings of children’s distress were quite high: for the injury, Pearson’s r = .58 and for hospital treatment, r = .72 (both ps < .001). Furthermore, we did regression analyses (see below) on the relationship between recall and our stress measures (both parental ratings and children’s self-descriptions of crying, independently), and the results were the same, regardless of which measure of stress we used. There was, however, one study in which Rees, Fivush, Sales and I used the Faces Pain Scale that was used by Steward and her colleagues (1996) to get children’s self-ratings of distress. There were 62 children between two and six years of age who provided this scale, and the correlations between their self-ratings of distress on the Faces Pain Scale and parental ratings of how distressed their children were, on a verbal Likert scale, were r = .24 (p < .05) for distress during the injury event and r = .30 (p < .01) for distress during hospital treatment. Although these are statistically significant, the correlations are nevertheless rather low. Furthermore, correlation and regression analyses found little relationship between children’s Faces Pain Scale judgments and their recall, while there were such relationships for both parental ratings and children’s self-descriptions of crying. Thus, having children make judgments on the basis of line drawings of faces may not be a particularly good way of measuring children’s distress, at least when children are this young. On the other hand, looking at children’s selfInjuries, Emergency Rooms, and Children’s Memory

69

descriptions of their emotional reactions, such as how much they cried, does seem to hold promise as a measure of children’s distress about real-world events that are stressful enough to elicit what Deffenbacher et al. (2004) term a defensive mode of response, when there are no research investigators around who could collect other measures. In the research in which we collected parental Likert ratings of child distress as well as children’s self-descriptions of crying, we looked at the relationship between children’s degree of distress and their recall. We had 201 children between two and 13 years of age who had both an initial and one-year follow-up interview, as well as 145 with a twoyear follow-up. Overall, for the injury, approximately 50% of the children were rated as highly upset and 40% moderately upset, with 10% in the low stress category. For hospital treatment, approximately 30% of the children were classified as highly stressed and about the same for moderately stressed, while about 40% were classified in the low stress category. Repeatedly we have found that there is no correlation between children’s distress at the time of injury and their age—older children were as likely to be highly distressed as younger children. However, stress ratings were correlated with age for hospital stress in that younger children were more likely to be highly distressed during hospital treatment than were older children. Both partial correlations (age partialled out) and stepwise regressions were calculated between children’s distress and their recall, using both children’s self-descriptions and parental ratings separately. For the accuracy of children’s recall, stress played no significant role in either their initial or follow-up interviews. This agrees with the conclusions derived by Deffenbacher and his colleagues in their review of the effects on stress on memory. In that review, they found that, although the accuracy of adults’ recall was compromised by high degrees of distress, children’s accuracy was unimpaired. In contrast, Salmon et al. (2002) found that children who cried during a voiding cystourethrogram had less accurate recall. Our findings for both the completeness of children’s recall and the amount of new information provided were quite different than for accuracy, however. During their initial interview, as stress increased, children became less complete in their accounts. For the number of new units of information, children who were least distressed consistently provided more information in their initial interview than did those who were most distressed. (See Table 3–1.) Salmon et al. (2002) also found crying associated with less information recalled by children. In contrast to the results described above, children who recalled a destructive hurricane recalled more new units of information if they were placed in the moderately distressed group rather than either the 70

STRESS AND MEMORY, EMPIRICAL EVIDENCE

Table 3–1. Percentage Completeness of Recall in Initial Interview as a Function of Distress (Child Self-Descriptions and Parent Ratings Averaged) Degree of Stress Low Event

Mean

Moderate

High

SD

Mean

SD

Mean

SD

17.4 23.2

72.6% 57.3%

18.5 19.2

71.5% 48.6%

15.7 21.2

29.3 29.9

46.5 49.2

23.6 24.4

48.3 39.9

21.4 24.4

Recall Completeness (percentage) Injury Hospital

75.0% 63.6%

Number of New Units of Information Injury Hospital

56.9 48.8

low- or high-stress groups (Bahrick et al., 1998). However, there are numerous differences between that study and ours: not only did they differ in terms of the nature of the event and the delay before an initial interview took place (several months in the hurricane study), but classification of stress level was done on the basis of damage to the children’s houses rather than how the children themselves emotionally reacted. Furthermore, children in the moderate stress group lived with house renovations and repairs caused by hurricane damage, while other children either had to move out or had little house damage. Thus, children in the moderate-stress category may have had more extended reminders of the event or may have included more information on the aftermath of the hurricane. To summarize the results of our emergency-room injury studies, children who were highly distressed had worse recall during their initial interview, both in terms of recall completeness and the amount of information provided. Thus, children’s recall seems to be compromised under conditions of high stress. This relationship is consistent with what Deffenbacher et al. (2004) found with adults, who also had poorer recall when they experienced high degrees of distress. However, what the children in our study did recall remained just as accurate regardless of how distressed they were, unlike what Deffenbacher et al. found for adults.

Variation in Language Competence Some investigators have suggested that children’s language competence may play a role in how well children recall events. As suggested by Boland et al. (2003), children with better language skills at the time of an event might be better at verbally encoding the details, which in turn might help them remember it later. Although better verbal skills have Injuries, Emergency Rooms, and Children’s Memory

71

sometimes been shown to help children recall the details of nonstressful events (Boland et al., 2003; Gordon et al., 1993; Simcock & Hayne, 2002), little research has looked at this variable when target events were stressful. An exception is Burgwyn-Bailes et al., (2001) who assessed three- to seven-year-olds’ recall of details surrounding facial surgery for lacerations. They found that children’s receptive vocabulary (as measured by the Peabody Picture Vocabulary Test) was predictive of younger children’s recall, although not that of older children. We also investigated variation in language competence using the Peabody Picture Vocabulary Test. This work involved 95 children (46 girls and 49 boys) between two and six years of age, and was done in collaboration with Rees, Fivush, and Sales. In contrast to the findings of Burgwyn-Bailes et al. (2001), this same measure of receptive language was unrelated to how complete, how informative, or how accurate children’s recall of their injury or subsequent hospital treatment was.

Variation in Temperament It has also been suggested that children’s behavioral style, or temperament, may influence how they recall a stressful event. For example, Merritt et al. (1994) found that children who had higher adaptability and were higher on the approach/withdrawal dimension had better open-ended as well as total recall about a VCUG procedure. In contrast, BurgwynBailes et al. (2001) and Greenhoot, Ornstein, Gordon, & Baker-Ward (1999) found little or no relationship between children’s recall of details of their treatment of facial lacerations or of pediatric examinations and the temperament dimensions measured by the Temperament Assessment Battery for Children. In the same study of two- to six-year-olds mentioned above, in collaborative work with Rees, Sales, and Fivush, we assessed child temperament using the Emotionality, Activity, and Sociability Temperament Survey. We found no relationship between children’s recall of injury and emergency room treatment and any of the measured dimensions of temperament.

Parental Conversational Style A socioemotional factor that is beginning to receive attention is the way that parents habitually discuss past events with their children. Considerable research has shown that parents differ in the way they talk with their children about past experiences (e.g., Fivush, 1991; Fivush & Fromhoff, 1988; Haden, Haine, & Fivush, 1997; Harley & Reese, 1999; McCabe & Peterson, 1991; Peterson & McCabe, 2004; Reese & Fivush, 1993—see Fivush, Haden, & Reese, 2006, for a review). Some parents 72

STRESS AND MEMORY, EMPIRICAL EVIDENCE

engage in much more of this “memory talk” than others (McCabe & Peterson, 1991), and the structure of these conversational exchanges differs in systematic ways. A dimension of conversational exchange differences that has been frequently studied is that of parental elaboration. When parents use an elaborative conversational style while discussing past events with their children, they elaborate on what their children say as well as encourage their children to provide more elaboration about target events in their own turns at talk. Specifically, elaborative parents (compared to nonelaborative) provide more information in their turns at talk and encourage and support their children’s contributions. They ask open-ended questions and encourage extended, dyadic discussion of target events. Other parents, in contrast, ask a few formulaic questions and engage in little of this elaborative exchange. In short, parents differ in the frequency as well as structure of conversations about events in their children’s past. These parental differences in reminiscing style have repeatedly been shown to be related to how much information children later provide in their open-ended memory conversations with both parents and researchers (see reviews in Fivush et al., 2006; Nelson & Fivush, 2004; Peterson & McCabe, 2004; and Reese, 2002). However, the increased contribution to memory conversations by children of elaborative parents may reflect only an increased willingness to keep on talking. It is another matter to suggest that memory itself for past events may be better for children whose parents engage in this elaborative style of interaction. Recently, this is exactly what some investigators are suggesting. They have found that mothers who engaged in elaborative talk while a target event was ongoing, such as museum visits, video viewings, or living-room camping trips, had children who recalled more about those experiences later (Boland et al., 2003; Low & Durkin, 2001; Tessler & Nelson, 1994). Importantly, such talk can even have a facilitative effect on children’s memory when it takes place after the event is over (Conroy & Salmon, 2006; Leichtman, Pillemer, Wang, Koreishi & Han, 2000; McGuigan & Salmon, 2004). For example, Leichtman et al. (2000) recorded the visit of a former preschool teacher who visited her class after her baby was born and engaged in a series of activities with the children. Later that day, mothers (who were not present and did not know the details of the visit) talked with their children about their former teacher’s visit. Mothers who were elaborative in these memory conversations had children who recalled more about the visit three weeks later. In summary, mothers who engage in elaborative memory conversations with their children about both everyday and distinctive nonstressful events, not only during the event but also later, have children who in turn seem to recall more detail about those events. But to our knowledge, there Injuries, Emergency Rooms, and Children’s Memory

73

has been little investigation of whether parental reminiscing style affects children’s memory for highly stressful events which by their nature have high memorability. To explore the question of whether parents’ styles of talking to their children about past experiences is related to children’s recall of a stressful event, 67 parents of two- to six-year olds were provided with a tape recorder and asked to talk with their children about the target injury/hospital treatment events in as natural a manner as possible, while the researcher left the room (Peterson, Sales, Rees, & Fivush, 2007). The elaborativeness of the parents’ talk was coded by breaking parental utterances into propositional units and then coding each unit as one of the following: (a) memory question elaborations, in which a question tried to elicit new information from the child as well as incorporated new information within the question; (b) memory question repetitions, in which a question tried to elicit new information from the child but did not itself contain any new information; (c) yes-no question elaborations, in which the child was required to confirm or negate the new information provided by the parent; (d) yes-no question repetitions, in which the child was required to confi rm or negate previously mentioned information; or (e) evaluations, which were statements that confirmed or negated the child’s previous utterance. After this, we created a composite score for each parent. This composite score was derived as follows: the number of elaborations plus evaluations was divided by the number of all utterance types to provide an elaboration ratio. This ratio was then entered in statistical analyses to assess the relationship between an elaborative parental style and their children’s memory. As seen in Table 3–2, the relationship between age and all of the memory measures except accuracy of injury recall was highly significant. For the relationship between the elaboration ratio and the memory measures (with age partialled out), virtually all partial correlations were significant for the initial interview. For the two-year follow-up interview, the completeness of both injury and hospital recall as well as the accuracy of hospital recall were still significantly related to the parental elaboration ratio calculated from parent-child talk two years previously. Next, we did a series of hierarchical regression analyses, one on each of our recall outcome measures: amount of information, completeness, and accuracy, separately for the injury and hospital events. We removed the variance attributable to age, gender, the elaboration ratio, and the age x elaboration interaction, in that order. For the two-year follow-up data, we did the same, except that variance attributable to the child’s prior recall during the initial interview was removed fi rst. Our results are displayed in Figure 3–4. 74

STRESS AND MEMORY, EMPIRICAL EVIDENCE

Table 3–2. Correlations between Children’s Memory Measures and Age, and Partial Correlations (Age Partialled Out) between Memory Measures and Parental Elaboration Ratio Measure

Event

Initial Interview Information Completeness Accuracy 2-Year Interview Information Completeness Accuracy

Age

Elaboration

Injury Hospital Injury Hospital Injury Hospital

.59*** .54*** .70*** .59*** .21 .43***

.25* .34** .25* .36** –.11 .38***

Injury Hospital Injury Hospital Injury Hospital

.57*** .54*** .48** .52*** .16 .38*

.24 .21 .32* .39* .12 .38*

* p < .05 ** p < .01 *** p < .001.

Initial Recall Regressions Information