Human-Computer Interaction. Interacting in Various Application Domains: 13th International Conference, HCI International 2009, San Diego, CA, USA, ... Applications, incl. Internet Web, and HCI)

Lecture Notes in Computer Science Commenced Publication in 1973 Founding and Former Series Editors: Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen

Editorial Board David Hutchison Lancaster University, UK Takeo Kanade Carnegie Mellon University, Pittsburgh, PA, USA Josef Kittler University of Surrey, Guildford, UK Jon M. Kleinberg Cornell University, Ithaca, NY, USA Alfred Kobsa University of California, Irvine, CA, USA Friedemann Mattern ETH Zurich, Switzerland John C. Mitchell Stanford University, CA, USA Moni Naor Weizmann Institute of Science, Rehovot, Israel Oscar Nierstrasz University of Bern, Switzerland C. Pandu Rangan Indian Institute of Technology, Madras, India Bernhard Steffen University of Dortmund, Germany Madhu Sudan Massachusetts Institute of Technology, MA, USA Demetri Terzopoulos University of California, Los Angeles, CA, USA Doug Tygar University of California, Berkeley, CA, USA Gerhard Weikum Max-Planck Institute of Computer Science, Saarbruecken, Germany

5613

Julie A. Jacko (Ed.)

Human-Computer Interaction Interacting in Various Application Domains 13th International Conference, HCI International 2009 San Diego, CA, USA, July 19-24, 2009 Proceedings, Part IV

13

Volume Editor Julie A. Jacko University of Minnesota Institute of Health Informatics MMC 912, 420 Delaware Street S.E., Minneapolis, MN 55455, USA E-mail: [email protected]

Library of Congress Control Number: 2009929048 CR Subject Classification (1998): H.5, I.3, I.7.5, I.5, I.2.10 LNCS Sublibrary: SL 3 – Information Systems and Application, incl. Internet/Web and HCI ISSN ISBN-10 ISBN-13

0302-9743 3-642-02582-X Springer Berlin Heidelberg New York 978-3-642-02582-2 Springer Berlin Heidelberg New York

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, re-use of illustrations, recitation, broadcasting, reproduction on microfilms or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable to prosecution under the German Copyright Law. springer.com © Springer-Verlag Berlin Heidelberg 2009 Printed in Germany Typesetting: Camera-ready by author, data conversion by Scientific Publishing Services, Chennai, India Printed on acid-free paper SPIN: 12707249 06/3180 543210

Foreword

The 13th International Conference on Human–Computer Interaction, HCI International 2009, was held in San Diego, California, USA, July 19–24, 2009, jointly with the Symposium on Human Interface (Japan) 2009, the 8th International Conference on Engineering Psychology and Cognitive Ergonomics, the 5th International Conference on Universal Access in Human–Computer Interaction, the Third International Conference on Virtual and Mixed Reality, the Third International Conference on Internationalization, Design and Global Development, the Third International Conference on Online Communities and Social Computing, the 5th International Conference on Augmented Cognition, the Second International Conference on Digital Human Modeling, and the First International Conference on Human Centered Design. A total of 4,348 individuals from academia, research institutes, industry and governmental agencies from 73 countries submitted contributions, and 1,397 papers that were judged to be of high scientific quality were included in the program. These papers address the latest research and development efforts and highlight the human aspects of the design and use of computing systems. The papers accepted for presentation thoroughly cover the entire field of human–computer interaction, addressing major advances in knowledge and effective use of computers in a variety of application areas. This volume, edited by Julie A. Jacko, contains papers in the thematic area of Human-Computer Interaction, addressing the following major topics: • • • •

eLearning and Education Games and Entertainment Work, Collaboration and Business Advanced Applications

The remaining volumes of the HCI International 2009 proceedings are: • • • • • •

Volume 1, LNCS 5610, Human–Computer Interaction––New Trends (Part I), edited by Julie A. Jacko Volume 2, LNCS 5611, Human–Computer Interaction––Novel Interaction Methods and Techniques (Part II), edited by Julie A. Jacko Volume 3, LNCS 5612, Human–Computer Interaction––Ambient, Ubiquitous and Intelligent Interaction (Part III), edited by Julie A. Jacko Volume 5, LNCS 5614, Universal Access in Human–Computer Interaction––Addressing Diversity (Part I), edited by Constantine Stephanidis Volume 6, LNCS 5615, Universal Access in Human–Computer Interaction––Intelligent and Ubiquitous Interaction Environments (Part II), edited by Constantine Stephanidis Volume 7, LNCS 5616, Universal Access in Human–Computer Interaction––Applications and Services (Part III), edited by Constantine Stephanidis

VI

Foreword

• • • • • • • • • •

Volume 8, LNCS 5617, Human Interface and the Management of Information––Designing Information Environments (Part I), edited by Michael J. Smith and Gavriel Salvendy Volume 9, LNCS 5618, Human Interface and the Management of Information––Information and Interaction (Part II), edited by Gavriel Salvendy and Michael J. Smith Volume 10, LNCS 5619, Human Centered Design, edited by Masaaki Kurosu Volume 11, LNCS 5620, Digital Human Modeling, edited by Vincent G. Duffy Volume 12, LNCS 5621, Online Communities and Social Computing, edited by A. Ant Ozok and Panayiotis Zaphiris Volume 13, LNCS 5622, Virtual and Mixed Reality, edited by Randall Shumaker Volume 14, LNCS 5623, Internationalization, Design and Global Development, edited by Nuray Aykin Volume 15, LNCS 5624, Ergonomics and Health Aspects of Work with Computers, edited by Ben-Tzion Karsh Volume 16, LNAI 5638, The Foundations of Augmented Cognition: Neuroergonomics and Operational Neuroscience, edited by Dylan Schmorrow, Ivy Estabrooke and Marc Grootjen Volume 17, LNAI 5639, Engineering Psychology and Cognitive Ergonomics, edited by Don Harris

I would like to thank the Program Chairs and the members of the Program Boards of all thematic areas, listed below, for their contribution to the highest scientific quality and the overall success of HCI International 2009.

Ergonomics and Health Aspects of Work with Computers Program Chair: Ben-Tzion Karsh Arne Aarås, Norway Pascale Carayon, USA Barbara G.F. Cohen, USA Wolfgang Friesdorf, Germany John Gosbee, USA Martin Helander, Singapore Ed Israelski, USA Waldemar Karwowski, USA Peter Kern, Germany Danuta Koradecka, Poland Kari Lindström, Finland

Holger Luczak, Germany Aura C. Matias, Philippines Kyung (Ken) Park, Korea Michelle M. Robertson, USA Michelle L. Rogers, USA Steven L. Sauter, USA Dominique L. Scapin, France Naomi Swanson, USA Peter Vink, The Netherlands John Wilson, UK Teresa Zayas-Cabán, USA

Foreword

Human Interface and the Management of Information Program Chair: Michael J. Smith Gunilla Bradley, Sweden Hans-Jörg Bullinger, Germany Alan Chan, Hong Kong Klaus-Peter Fähnrich, Germany Michitaka Hirose, Japan Jhilmil Jain, USA Yasufumi Kume, Japan Mark Lehto, USA Fiona Fui-Hoon Nah, USA Shogo Nishida, Japan Robert Proctor, USA Youngho Rhee, Korea

Anxo Cereijo Roibás, UK Katsunori Shimohara, Japan Dieter Spath, Germany Tsutomu Tabe, Japan Alvaro D. Taveira, USA Kim-Phuong L. Vu, USA Tomio Watanabe, Japan Sakae Yamamoto, Japan Hidekazu Yoshikawa, Japan Li Zheng, P.R. China Bernhard Zimolong, Germany

Human–Computer Interaction Program Chair: Julie A. Jacko Sebastiano Bagnara, Italy Sherry Y. Chen, UK Marvin J. Dainoff, USA Jianming Dong, USA John Eklund, Australia Xiaowen Fang, USA Ayse Gurses, USA Vicki L. Hanson, UK Sheue-Ling Hwang, Taiwan Wonil Hwang, Korea Yong Gu Ji, Korea Steven Landry, USA

Gitte Lindgaard, Canada Chen Ling, USA Yan Liu, USA Chang S. Nam, USA Celestine A. Ntuen, USA Philippe Palanque, France P.L. Patrick Rau, P.R. China Ling Rothrock, USA Guangfeng Song, USA Steffen Staab, Germany Wan Chul Yoon, Korea Wenli Zhu, P.R. China

Engineering Psychology and Cognitive Ergonomics Program Chair: Don Harris Guy A. Boy, USA John Huddlestone, UK Kenji Itoh, Japan Hung-Sying Jing, Taiwan Ron Laughery, USA Wen-Chin Li, Taiwan James T. Luxhøj, USA

Nicolas Marmaras, Greece Sundaram Narayanan, USA Mark A. Neerincx, The Netherlands Jan M. Noyes, UK Kjell Ohlsson, Sweden Axel Schulte, Germany Sarah C. Sharples, UK

VII

VIII

Foreword

Neville A. Stanton, UK Xianghong Sun, P.R. China Andrew Thatcher, South Africa

Matthew J.W. Thomas, Australia Mark Young, UK

Universal Access in Human–Computer Interaction Program Chair: Constantine Stephanidis Julio Abascal, Spain Ray Adams, UK Elisabeth André, Germany Margherita Antona, Greece Chieko Asakawa, Japan Christian Bühler, Germany Noelle Carbonell, France Jerzy Charytonowicz, Poland Pier Luigi Emiliani, Italy Michael Fairhurst, UK Dimitris Grammenos, Greece Andreas Holzinger, Austria Arthur I. Karshmer, USA Simeon Keates, Denmark Georgios Kouroupetroglou, Greece Sri Kurniawan, USA

Patrick M. Langdon, UK Seongil Lee, Korea Zhengjie Liu, P.R. China Klaus Miesenberger, Austria Helen Petrie, UK Michael Pieper, Germany Anthony Savidis, Greece Andrew Sears, USA Christian Stary, Austria Hirotada Ueda, Japan Jean Vanderdonckt, Belgium Gregg C. Vanderheiden, USA Gerhard Weber, Germany Harald Weber, Germany Toshiki Yamaoka, Japan Panayiotis Zaphiris, UK

Virtual and Mixed Reality Program Chair: Randall Shumaker Pat Banerjee, USA Mark Billinghurst, New Zealand Charles E. Hughes, USA David Kaber, USA Hirokazu Kato, Japan Robert S. Kennedy, USA Young J. Kim, Korea Ben Lawson, USA

Gordon M. Mair, UK Miguel A. Otaduy, Switzerland David Pratt, UK Albert “Skip” Rizzo, USA Lawrence Rosenblum, USA Dieter Schmalstieg, Austria Dylan Schmorrow, USA Mark Wiederhold, USA

Internationalization, Design and Global Development Program Chair: Nuray Aykin Michael L. Best, USA Ram Bishu, USA Alan Chan, Hong Kong Andy M. Dearden, UK

Susan M. Dray, USA Vanessa Evers, The Netherlands Paul Fu, USA Emilie Gould, USA

Foreword

Sung H. Han, Korea Veikko Ikonen, Finland Esin Kiris, USA Masaaki Kurosu, Japan Apala Lahiri Chavan, USA James R. Lewis, USA Ann Light, UK James J.W. Lin, USA Rungtai Lin, Taiwan Zhengjie Liu, P.R. China Aaron Marcus, USA Allen E. Milewski, USA

Elizabeth D. Mynatt, USA Oguzhan Ozcan, Turkey Girish Prabhu, India Kerstin Röse, Germany Eunice Ratna Sari, Indonesia Supriya Singh, Australia Christian Sturm, Spain Adi Tedjasaputra, Singapore Kentaro Toyama, India Alvin W. Yeo, Malaysia Chen Zhao, P.R. China Wei Zhou, P.R. China

Online Communities and Social Computing Program Chairs: A. Ant Ozok, Panayiotis Zaphiris Chadia N. Abras, USA Chee Siang Ang, UK Amy Bruckman, USA Peter Day, UK Fiorella De Cindio, Italy Michael Gurstein, Canada Tom Horan, USA Anita Komlodi, USA Piet A.M. Kommers, The Netherlands Jonathan Lazar, USA Stefanie Lindstaedt, Austria

Gabriele Meiselwitz, USA Hideyuki Nakanishi, Japan Anthony F. Norcio, USA Jennifer Preece, USA Elaine M. Raybourn, USA Douglas Schuler, USA Gilson Schwartz, Brazil Sergei Stafeev, Russia Charalambos Vrasidas, Cyprus Cheng-Yen Wang, Taiwan

Augmented Cognition Program Chair: Dylan D. Schmorrow Andy Bellenkes, USA Andrew Belyavin, UK Joseph Cohn, USA Martha E. Crosby, USA Tjerk de Greef, The Netherlands Blair Dickson, UK Traci Downs, USA Julie Drexler, USA Ivy Estabrooke, USA Cali Fidopiastis, USA Chris Forsythe, USA Wai Tat Fu, USA Henry Girolamo, USA

Marc Grootjen, The Netherlands Taro Kanno, Japan Wilhelm E. Kincses, Germany David Kobus, USA Santosh Mathan, USA Rob Matthews, Australia Dennis McBride, USA Robert McCann, USA Jeff Morrison, USA Eric Muth, USA Mark A. Neerincx, The Netherlands Denise Nicholson, USA Glenn Osga, USA

IX

X

Foreword

Dennis Proffitt, USA Leah Reeves, USA Mike Russo, USA Kay Stanney, USA Roy Stripling, USA Mike Swetnam, USA Rob Taylor, UK

Maria L.Thomas, USA Peter-Paul van Maanen, The Netherlands Karl van Orden, USA Roman Vilimek, Germany Glenn Wilson, USA Thorsten Zander, Germany

Digital Human Modeling Program Chair: Vincent G. Duffy Karim Abdel-Malek, USA Thomas J. Armstrong, USA Norm Badler, USA Kathryn Cormican, Ireland Afzal Godil, USA Ravindra Goonetilleke, Hong Kong Anand Gramopadhye, USA Sung H. Han, Korea Lars Hanson, Sweden Pheng Ann Heng, Hong Kong Tianzi Jiang, P.R. China

Kang Li, USA Zhizhong Li, P.R. China Timo J. Määttä, Finland Woojin Park, USA Matthew Parkinson, USA Jim Potvin, Canada Rajesh Subramanian, USA Xuguang Wang, France John F. Wiechel, USA Jingzhou (James) Yang, USA Xiu-gan Yuan, P.R. China

Human Centered Design Program Chair: Masaaki Kurosu Gerhard Fischer, USA Tom Gross, Germany Naotake Hirasawa, Japan Yasuhiro Horibe, Japan Minna Isomursu, Finland Mitsuhiko Karashima, Japan Tadashi Kobayashi, Japan

Kun-Pyo Lee, Korea Loïc Martínez-Normand, Spain Dominique L. Scapin, France Haruhiko Urokohara, Japan Gerrit C. van der Veer, The Netherlands Kazuhiko Yamazaki, Japan

In addition to the members of the Program Boards above, I also wish to thank the following volunteer external reviewers: Gavin Lew from the USA, Daniel Su from the UK, and Ilia Adami, Ioannis Basdekis, Yannis Georgalis, Panagiotis Karampelas, Iosif Klironomos, Alexandros Mourouzis, and Stavroula Ntoa from Greece. This conference could not have been possible without the continuous support and advice of the Conference Scientific Advisor, Prof. Gavriel Salvendy, as well as the dedicated work and outstanding efforts of the Communications Chair and Editor of HCI International News, Abbas Moallem.

Foreword

XI

I would also like to thank for their contribution toward the organization of the HCI International 2009 conference the members of the Human–Computer Interaction Laboratory of ICS-FORTH, and in particular Margherita Antona, George Paparoulis, Maria Pitsoulaki, Stavroula Ntoa, and Maria Bouhli. Constantine Stephanidis

HCI International 2011

The 14th International Conference on Human–Computer Interaction, HCI International 2011, will be held jointly with the affiliated conferences in the summer of 2011. It will cover a broad spectrum of themes related to human–computer interaction, including theoretical issues, methods, tools, processes and case studies in HCI design, as well as novel interaction techniques, interfaces and applications. The proceedings will be published by Springer. More information about the topics, as well as the venue and dates of the conference, will be announced through the HCI International Conference series website: http://www.hci-international.org/

General Chair Professor Constantine Stephanidis University of Crete and ICS-FORTH Heraklion, Crete, Greece Email: [email protected]

Table of Contents

Part I: eLearning and Education Arab Children’s Reading Preference for Different Online Fonts . . . . . . . . . Asmaa Alsumait, Asma Al-Osaimi, and Hadlaa AlFedaghi Adaptation Decisions and Profiles Exchange among Open Learning Management Systems Based on Agent Negotiations and Machine Learning Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Silvia Baldiris, Ram´ on Fabregat, Carolina Mej´ıa, and Sergio G´ omez Accessing e-Learning Systems via Screen Reader: An Example . . . . . . . . . Maria Claudia Buzzi, Marina Buzzi, and Barbara Leporini

3

12 21

Using Tablet PCs and Pen-Based Technologies to Support Engineering Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ignacio Casas, Sergio F. Ochoa, and Jaime Puente

31

Optimal Affective Conditions for Subconscious Learning in a 3D Intelligent Tutoring System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pierre Chalfoun and Claude Frasson

39

Computer-Based Learning to Improve Breast Cancer Detection Skills . . . Yan Chen, Alastair Gale, Hazel Scott, Andrew Evans, and Jonathan James Virtual Classroom and Communicability: Empathy and Interaction for All . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Francisco V. Cipolla Ficarra

49

58

Communicability for Virtual Learning: Evaluation . . . . . . . . . . . . . . . . . . . . Francisco V. Cipolla-Ficarra, Miguel Cipolla-Ficarra, and Pablo M. Vera

68

Attention and Motivation in Hypermedia Systems . . . . . . . . . . . . . . . . . . . . Francisco V. Cipolla Ficarra and Miguel Cipolla-Ficarra

78

A Web-Based, Interactive Annotation Editor for the eCampus Development Environment for SCORM Compliant E-Learning Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Benedikt Deicke, Jan-Torsten Milde, and Hans-Martin Pohl An Innovative Way of Understanding Learning Processes: Eye Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Berrin Dogusoy and Kursat Cagiltay

88

94

XVI

Table of Contents

A Set of Rules and Strategies for UNSAM Virtual Campus . . . . . . . . . . . . Jorge Fern´ andez Niello, Francisco V. Cipolla Ficarra, Mario Greco, Rodolfo Fern´ andez-Ziegler, Silvia Bernaten´e, and Maria Villarreal HCI Professional Involvement in k-12 Education: On Target or Missing the Mark? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Martin Jelin, Adrian Sudol, Jeffrey Damon, Douglas McCadden, and David Brown A Language Learning System Utilizing RFID Technology for Total Physical Response Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Harumi Kashiwagi, Yan Xue, Yi Sun, Min Kang, and Kazuhiro Ohtsuki

101

111

119

Promoting Metacognition in Immersive Cultural Learning Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H. Chad Lane

129

The Application of the Flexilevel Approach for the Assessment of Computer Science Undergraduates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mariana Lilley and Andrew Pyper

140

Development of Ubiquitous On-Demand Study Support Environment for Nursing Students . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yukie Majima, Yumiko Nakamura, Yasuko Maekawa, and Yoichiro So

149

The Effects of Prior Knowledge on the Use of Adaptive Hypermedia Learning Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Freddy Mampadi, Sherry Chen, and Gheorghita Ghinea

156

Supporting Learners in Adaptive Learning Environments through the Enhancement of the Student Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Luca Mazzola and Riccardo Mazza

166

The Concept of IMPRESSION: An Interactive Instruction System and Its Practice for Real-Time Distance Lessons between U.S. and Japan . . . Takashi Mitsuishi, Fumiko Konno, Yuki Higuchi, and Kentaro Go

176

Improving Children’s Writing Ability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Joana Pereira, Lu´ıs Carri¸co, and Carlos Duarte From Paper to Module – An Integrated Environment for Generating SCORM Compliant Moodle Courses Out of Text and Multimedia Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hans-Martin Pohl, Benedikt Deicke, and Jan-Torsten Milde Development of a Simulator of Abacus: Ancient Analog Calculator on a Mobile Phone as a Teaching Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kenta Saito, Yuki Makita, Vu Quang, and Hitoshi Sasaki

186

196

204

Table of Contents

A Proposal for a Framework for an e-Alumni Program Using SNS . . . . . . Hiroshi Sano Supporting End-User Development of Personalized Mobile Learning Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Marco de S´ a and Lu´ıs Carri¸co

XVII

209

217

Didactic Models as Design Representations . . . . . . . . . . . . . . . . . . . . . . . . . . Chris Stary

226

Interactive Learning Panels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ricardo Tesoriero, Habib Fardoun, Jos´e Gallud, Mar´ıa Lozano, and Victor Penichet

236

WebELS: A Content-Centered E-Learning Platform for Postgraduate Education in Engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Haruki Ueno, Zheng He, and Jingxia Yue A Pen-Based Teaching System for Children and Its Usability Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Danli Wang, Tingting Ying, Jinquan Xiong, Hongan Wang, and Guozhong Dai Development of a Visualised Sound Simulation Environment: An e-Approach to a Constructivist Way of Learning . . . . . . . . . . . . . . . . . . . . . Jingjing Zhang, Beau Lotto, Ilias Bergstom, Lefkothea Andreou, Youzou Miyadera, and Setsuo Yokoyama

246

256

266

Part II: Games and Entertainment Causal Links of Presence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Donghun Chung and Chae-Hwan Kim

279

Games Design Principles for Improving Social Web Applications . . . . . . . Ines Di Loreto

287

A Multiple-Level 3D-LEGO Game in Augmented Reality for Improving Spatial Ability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Trien V. Do and Jong-Weon Lee

296

An Online Survey System on Computer Game Enjoyment and Personality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Xiaowen Fang, Susy Chan, and Chitra Nair

304

Playability Testing of Web-Based Sport Games with Older Children and Teenagers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Xavier Ferre, Angelica de Antonio, Ricardo Imbert, and Nelson Medinilla

315

XVIII

Table of Contents

Exploring the Elements and Design Criteria of Massively-Multiplayer Online Role-Playing Game (MMORPG) Interfaces . . . . . . . . . . . . . . . . . . . Chun-Cheng Hsu and Elvis Chih-Hsien Chen

325

Healthcare Game Design: Behavioral Modeling of Serious Gaming Design for Children with Chronic Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . Hadi Kharrazi, Anthony Faiola, and Joseph Defazio

335

Analyzing Human Behaviors in an Interactive Art Installation . . . . . . . . . Takashi Kiriyama and Masahiko Sato The Effects of Quest Types and Gaming Motivations on Players’ Knowledge Acquisitions in an Online Role-Playing Game Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jiunde Lee and Chih-Yi Chao Self-movement Feeling Generation in Sports Watching with Screen Movement via Pan-Tilt Steerable Projector . . . . . . . . . . . . . . . . . . . . . . . . . . Hiroshi Noguchi, Kei Yoshinaka, Taketoshi Mori, and Tomomasa Sato Design of Interactive Emotional Sound Edutainment System . . . . . . . . . . Myunjin Park and Kyujung Kim

345

353

359

368

Understanding Online Game Addiction: Connection between Presence and Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SungBok Park and Ha Sung Hwang

378

The Experience of Presence in 3D Web Environment: An Analysis of Korean Second Life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SungBok Park, Ha Sung Hwang, and Myungil Choi

387

Influence of Real-World Ten-Pin Bowling Experience on Performance during First-Time Nintendo Wii Bowling Practice . . . . . . . . . . . . . . . . . . . . Kirsten A. Peters

396

Emotionally Adapted Games – An Example of a First Person Shooter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Timo Saari, Marko Turpeinen, Kai Kuikkaniemi, Ilkka Kosunen, and Niklas Ravaja DiamondTheater: A System for Reproducing Theater and Supporting Creative Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tatsushi Takeuchi, Koichiro Watanabe, Tomoo Inoue, and Ken-ichi Okada

406

416

Part III: Work, Collaboration and Business New Health Information Systems (HIS) Quality-in-Use Model Based on the GQM Approach and HCI Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reem Al-Nanih, Hana Al-Nuaim, and Olga Ormandjieva

429

Table of Contents

An Information Visualization Approach to Hospital Shifts Scheduling . . . Carmelo Ardito, Paolo Buono, Maria F. Costabile, Rosa Lanzilotti, and Adalberto L. Simeone Designed to Fit: Challenges of Interaction Design for Clothes Fitting Room Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bo Begole, Takashi Matsumoto, Wei Zhang, Nicholas Yee, Juan Liu, and Maurice Chu Usability for Poll Workers: A Voting System Usability Test Protocol . . . . Dana Chisnell, Karen Bachmann, Sharon Laskowski, and Svetlana Lowry

XIX

439

448

458

CAD and Communicability: A System That Improves the Human-Computer Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Francisco V. Cipolla Ficarra and Roc´ıo A. Rodr´ıguez

468

A Novel Visualization Tool for Evaluating Medication Side-Effects in Multi-drug Regimens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jon Duke, Anthony Faiola, and Hadi Kharrazi

478

Design of a Web Intervention to Change Youth Smoking Habits . . . . . . . . Kim Nee Goh, Yoke Yie Chen, Emy Elyanee Mustapha, Subarna Sivapalan, and Sharina Nordin

488

Smart Makeup Mirror: Computer-Augmented Mirror to Aid Makeup Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eriko Iwabuchi, Maki Nakagawa, and Itiro Siio

495

Studying Reactive, Risky, Complex, Long-Spanning, and Collaborative Work: The Case of IT Service Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eser Kandogan, Eben M. Haber, John H. Bailey, and Paul P. Maglio

504

Human Computer Interaction in Virtual Standardized Patient Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Patrick G. Kenny, Thomas D. Parsons, and Albert A. Rizzo

514

Towards Standardized Pen-Based Annotation of Breast Cancer Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Suzanne Kieffer, Annabelle Gouze, Ronald Moncarey, Christian Van Brussel, Jean-Fran¸cois De Wispelaere, Fran¸coise Kayser, and Benoˆıt Macq ImproV: A System for Improvisational Construction of Video Processing Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Atsutomo Kobayashi, Buntarou Shizuki, and Jiro Tanaka E-Assessment: A Suitable Alternative for Measuring Competences? . . . . . Martin Kr¨ oll

524

534 543

XX

Table of Contents

Green Advocate in E-Commerce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ying-Lien Lee, Fei-Hui Huang, and Sheue-Ling Hwang

551

Gesture-Based Sharing of Documents in Face-to-Face Meetings . . . . . . . . Alexander Loob and Christian Rathke

558

Developing, Deploying and Assessing Usage of a Movie Archive System among Students of Film Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nazlena Mohamad Ali, Alan F. Smeaton, Hyowon Lee, and Pat Brereton Using Activity Descriptions to Generate User Interfaces for ERP Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Timothy O’Hear and Yassin Boudjenane Developing a Nomenclature for EMR Errors . . . . . . . . . . . . . . . . . . . . . . . . . Win Phillips and Yang Gong

567

577 587

Mapping for Multi-source Visualization: Scientific Information Retrieval Service (SIRS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dario Rodighiero, Matina Halkia, and Massimiliano Gusmini

597

Client-Side Visualization of Internet Forums for Information Retrieval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Guangfeng Song

606

Social-Technical Tools for Collaborative Sensemaking and Sketching . . . . James Sullivan, Meredith Banasiak, Christopher Messick, and Raymond Rimey

614

Developing Some User Interfaces of TV under Enormous Channels Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Shumpei Tamaoki, Tomohiro Torikai, and Hirohiko Mori

624

Electronic Glassboard – Conception and Implementation of an Interactive Tele-presence Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peter Thies and Benjamin Koehne

632

A New Automatic Teller Machine (ATM) Proposal through the Analysis of ATMs of Three Banks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Serdar Yarlikas

641

Part IV: Advanced Applications Designing Usable Bio-information Architectures . . . . . . . . . . . . . . . . . . . . . . Davide Bolchini, Anthony Finkestein, and Paolo Paolini Run-Time Adaptation of a Universal User Interface for Ambient Intelligent Production Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kai Breiner, Daniel G¨ orlich, Oliver Maschino, Gerrit Meixner, and Detlef Z¨ uhlke

653

663

Table of Contents

XXI

Heuristic Evaluation of Mission-Critical Software Using a Large Team . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tim Buxton, Alvin Tarrell, and Ann Fruhling

673

Interface Development for Early Notification Warning System: Full Windshield Head-Up Display Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . Vassilis Charissis, Stylianos Papanastasiou, and George Vlachos

683

Reflections on the Interdisciplinary Collaborative Design of Mapping the Universe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chaomei Chen, Jian Zhang, and Michael S. Vogeley

693

Distilling Support Opportunities to Improve Urban Search and Rescue Missions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tjerk de Greef, A.H.J. Oomes, and Mark A. Neerincx

703

A New Approach to Design an Interactive System for Molecular Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mouna Essabbah, Samir Otmane, Joan H´erisson, and Malik Mallem

713

The Differences of Aviation Human Factors between Individualism and Collectivism Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wen-Chin Li, Don Harris, Lon-Wen Li, and Thomas Wang

723

Web-Based Training System for Improving Aviation Maintenance Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Guo-Feng Liang, Jhih-Tsong Lin, Sheue-Ling Hwang, Eric Min-yang Wang, Patrick Patterson, and Jiun-Fa Li Allocating Human-System Interfaces Functions by Levels of Automation in an Advanced Control Room . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chiuhsiang Joe Lin, Chih-Wei Yang, Tzu-Chung Yenn, and Lai-Yu Cheng Development of an Expert System as a User Interface for an RFID Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Deok Hee Nam Developing a Validation Methodology for Educational Driving Simulators and a Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hatice Sancar, Kursat Cagiltay, Veysi Isler, Gizem Tamer, Neslihan Ozmen, and Utkan Eryilmaz

731

741

751

760

Developing a Usable Mobile Flight Case Learning System in Air Traffic Control Miscommunications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kuo-Wei Su, Keh-Yeu Lee, Po-Hsin Huang, and I-Tsun Chen

770

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

779

Arab Children’s Reading Preference for Different Online Fonts Asmaa Alsumait1, Asma Al-Osaimi2, and Hadlaa AlFedaghi2 1 Computer Engineering Dep., Kuwait University, Kuwait Regional Center For Development of Educational Software, Kuwait [email protected], {alosaimi,hadlaa}@redsoft.org 2

Abstract. E-learning education plays an important role in the educational process in the Arab region. There is more demand to provide Arab students with electronic resources for knowledge now than before. The readability of such electronic resources needs to be taken into consideration. Following design guidelines in the e-learning programs’ design process improves both the reading performance and satisfaction. However, English script design guidelines cannot be directly applied to Arabic script mainly because of difference in the letters occupation and writing direction. Thus, this paper aimed to build a set of design guidelines for Arabic e-learning programs designed for seven-to-nine years old children. An electronic story is designed to achieve this goal. It is used to gather children’s reading preferences, for example, font type/size combination, screen line length, and tutoring sound characters. Results indicated that Arab students preferred the use of Simplified Arabic with 14-point font size to ease and speed the reading process. Further, 2/3 screen line length helped children in reading faster. Finally, most of children preferred to listen to a female adult tutoring sound. Keywords: Child-Computer Interfaces, E-Learning, Font Type/Size, HumanComputer Interaction, Information Interfaces and Presentation, Line Length, Tutoring Sound.

1 Introduction Ministries of education in the Arab region are moving toward adopting e-learning methods in the educational process. In fact, the public schools in the state of Kuwait are using e-learning programs for elementary and middle school students. These elearning programs are designed by the Ministry of Education and the Regional Center for Development of Educational Software (ReDSOFT). Nowadays e-learning became a necessity not a luxury as it affects the patterns of accumulating knowledge. Thus, without a sound educational model, e-learning education will fail the designer, the instructor and the learner. It is very important to design the e-learning material in a way that can keep children's concentration focused on the tasks. This can be done by using proper font type, size, and screen line length. Many researchers have considered building a set of guidelines to be used in designing e-learning software for adults and children. For example, Bernard et. al. (2002) J.A. Jacko (Ed.): Human-Computer Interaction, Part IV, HCII 2009, LNCS 5613, pp. 3–11, 2009. © Springer-Verlag Berlin Heidelberg 2009

4

A. Alsumait, A. Al-Osaimi, and H. AlFedaghi

studied adults’ preferences regarding font type and size from measuring adults’ reading efficiency, adults’ reading time, legibility, attractiveness, and adults’ font preference [1]. They found that Times and Arial font type with size 12-point were read faster than other fonts and sizes that were tested. Further, Arial and Courier font type were considered the most legible fonts. Again, Arial font type with 12-point size was the most preferable to read online. Another study carried out for adults compared the fonts used in previous editions of Windows with those new created for Windows Vista [3]. They found that the new fonts designed for Windows Vista such as Cambria and Constantia were more legible than traditional Times New Roman. While studies conducted on children indicated that the children prefer to read English text that is Arial font type with 14-point size and Comic font type with 12-point size [2]. Moreover, in English, adults prefer to read medium line length (approximately 65 to 75 CPL) and children prefer narrow line length (approximately 45 CPL) [4, 6]. On the other hand, few studies were conducted to find adult users’ preferences for Arabic script. For example, Hemayssi et. al. concluded that in order to increase legibility, it is preferred to use bold fonts, colors and clear icons [5]. Most existing research in this area is oriented to build guidelines for designing English e-learning programs. Therefore, in this paper, we examined the preferences of seven to nine years old Arab children's for five font types at sizes 12- and 14-point and their screen line length using an electronic story.

2 Usability Evaluation Process The usability evaluation of the Arabic electronic story for font type, size, and line length was conducted on July 2008. Participants were asked to spend half an hour with the e-story under the supervision of the ReDSOFT team member experimenter. At the beginning of the usability evaluation, participants were asked to complete a user profile questionnaire. The background questionnaire revealed that thirty five participants (16 males, 19 females) aged between seven and nine years old. Moreover, majority (86%) of the participants use computer at home, and almost all of them (97%) had previous experience with electronic reading. Each page of the e-story was written with a different combination of font type and size. In total, the e-story was presented in 10 pages representing the five font types (Arial Unicode MS, Courier New, Microsoft Sans Serif, Simplified Arabic, and Traditional Arabic) and two font sizes (12 and 14). Figure 1 shows sample of a page of the e-story.

Fig. 1. A sample of a page of the e-story

Arab Children’s Reading Preference for Different Online Fonts

5

After reading each page of the e-story, participants were asked to answer four questions to measure their satisfaction regarding font type and size used. 1. Is this font type and size easy to read? 2. Do you think you can read faster with this font type and size? 3. Is this font type and size attractive? 4. Would you like this font type and size to be used in your schoolbook? Text was displayed in the center of the screen. Fonts were black presented on a white background. A stopwatch was used to record the time participants took to read the paragraphs and the experimenter also noted the number of incorrectly pronounced words. To confirm findings, participants were asked to read a simple sentence with all font type and size combinations in a single page and select their preferable font type and size. Also participants were asked to choose the most comfortable screen line length while reading. The same font type and size (Simplified Arabic with size 14-pionts) script was presented to the participant in different line length conditions; 1/3 screen line length, 2/3 screen line length, and full screen line length. The participant was asked to rate his/her reading comfortance and satisfaction with the three different screen line lengths. We also explored a new direction for e-learning computer tutoring sound characters, which we believe will maximize students’ learning gains and enjoyment. The traditional scenario allows students to interact primarily with a single coach or tutor character sound on-screen. In this evaluation process, we allowed the child to select his own tutor sound to be a teacher, a boy, or a girl. The usability evaluation process measured seven main factors in the reading process: the ease of reading factor, reading faster factor, font attractiveness factor, the desire to use font combination in schoolbooks factor, the font preferences factor, the screen line length factor. And finally the tutor character sound factor. Table 1 shows different font types and sizes used in the experiment. Table 1. Font types and sizes studied Font type Arial Unicode MS Courier New Microsoft Sans Serif Simplified Arabic Traditional Arabic

12-point size

 ҭҥҫᦎᦸᦿҫᥰᦃҧ᥮ᧇҧ ΐΣ΃ Ύϧ΃ Γ˯΍ήϘϟ΍ ̸͓͙̺͙͛ͣ ͚̜͕ ̼͕̓ Γ˯΍ήϘϟ΍ ΐΣ΃ Ύϧ΃ Γ˯΍ήϘϟ΍ ΐΣ΃ Ύϧ΃

14-point size

 ҭҥҫᦎᦸᦿҫᥰᦃҧ᥮ᧇҧ Γ˯΍ήϘϟ΍ ΐΣ΃ Ύϧ΃ ̸͓͙̺͙͛ͣ ͚̜͕ ̼͕̓ Γ˯΍ήϘϟ΍ ΐΣ΃ Ύϧ΃ Γ˯΍ήϘϟ΍ ΐΣ΃ Ύϧ΃

3 Results 3.1 Ease of Reading Ease of reading measures whether a specific font type and size gave children an impression that the reading process was simple and trouble free. Analysis showed that

6

A. Alsumait, A. Al-Osaimi, and H. AlFedaghi

61% of the participants considered the Simplified Arabic font as an easiest to read font. The lowest ease of reading score was given to the Courier New and Traditional Arabic. Further analysis is done to find font sizes that help children in reading faster. The results showed that there is small difference between fonts that are sized 12-point and fonts that are sized 14-point. 57 % of the children thought that 14-point sized font are easy to read in oppose to 45% who thought that 12-point sized font is easy to read. The results revealed that using Simplified Arabic with 14-point font size was considered the easiest to read font for Arab children. Guideline 1 Use Simplified Arabic with 14-point font size to ease the reading for Arab children.

3.2 Reading Faster The second factor to be measured was reading speed. It is a variable that measures whether a specific font type and size gave children an impression that the reading process was done quicker. Analysis showed that font type has an affect on children's reading speed. Simplified Arabic and Microsoft Sans Serif had the highest participants’ score preference; 54% for reading faster. On the other hand, Courier New had the lowest preference score regarding the reading faster factor, only 34 % of the participants thought that they can read faster with it. When it comes to font size, results showed that children (47%) had preferred to read fonts that are sized 14-point. Fonts with size 12-point got a lower preference score. In addition, results showed that 60% of the participants believed that the Simplified Arabic with 14-point font size helped them read faster. In contrast, only 26% of the participants thought that Traditional Arabic with font size 12-point, will help them read faster. Guideline 2 Use Simplified Arabic with 14-point font size to speed the reading for Arab children.

3.3 Font Attractiveness Font Attractiveness measures whether children found a specific font type and size nice looking and eye catching. 66% of the participants considered the Arial Unicode MS as the most attractive font type among the types we tested. When it comes to font size, results showed that it has no difference in attractiveness scores. Font sizes 12-point and 14-point had similar attractiveness scores (around 60%). According to experimental results, Arial Unicode MS with 14-point font size got the best score when it comes to font type/size attractiveness.

Arab Children’s Reading Preference for Different Online Fonts

7

Guideline 3 Use Arial Unicode MS with 14-point font size when considering font attractiveness for children.

3.4 Desire to Use Font Type and Size Combination in Schoolbooks Desire to use font combination in schoolbooks factor measures whether children want to use a specific font type and size in their schoolbooks. The results showed that Simplified Arabic with font size 12-point and Arial Unicode MS with font size 14-point are considered the most desired font type and size combinations preferred to be used in schoolbooks. Guideline 4 Use Simplified Arabic with font size 12-point or Arial Unicode MS with font size 14-point when considering children's preferences for schoolbooks.

3.5 Confirm Font Preference 3.5.1 12-Point Font Size Figure 2 shows font type preferences percentage with size 12-point. Arial Unicode MS was selected as the most preferred (45%) font type with size 12-point.

50

Percentage (%)

40

30

20

10

0 Arial Unicode Courier New MS

Microsoft Sans Serif

Simplified Arabic

Traditional Arabic

Fig. 2. The font type preference percentage with size 12-point

3.5.2 14-Point Font Size Figure 3 shows font type preferences percentage with size 14-point. Arial Unicode MS was considered to be the most preferred (57%) font type with size 14-point.

8

A. Alsumait, A. Al-Osaimi, and H. AlFedaghi

60

50

Percentage (%)

40

30

20

10

0 Arial Unicode Courier New MS

Microsoft Sans Serif

Simplified Arabic

Traditional Arabic

Fig. 3. The font type preference percentage with size 14-point

Figure 4 shows which font size is preferred by participants. Result showed that participants preferred to read using 14-point font size. These results are in agreement with the results obtained through the usability evaluation process of reading the estory. 80 70

Percentage (%)

60 50 40 30 20 10 0 12-point

14-point

Fig. 4. Best font size selected by participants

3.6 Tutoring Sound At the end of the experiment, we asked the child to choose a character sound to tutor him/her through the final part of the survey. Figure 5 (a, b, c) shows the percentage of the selected tutoring sound character by the participants. The result showed that most children (48 %) regardless of their gender had selected the female teacher as a sound tutoring (Figure 5.a). This result is related to the fact that at this specific age (seven to nine years old) children are used to be taught by female teachers at the schools.

Arab Children’s Reading Preference for Different Online Fonts

60

50

Percentage

40

30

20

10

0 The teacher character "Hanan"

The boy character "Salem"

The girl character "Abeer"

Fig. 5a. The percentage of the selected tutoring sound character by all participants

60

50

Percentage (%)

40

30

20

10

0 The teacher "Hanan"

The boy character "Salem"

The girl character "Abeer"

Fig. 5b. The percentage of the selected tutoring sound character male participants

60

50

Percentage (%)

40

30

20

10

0 The teacher "Hanan"

The boy character "Salem"

The girl character "Abeer"

Fig. 5c. The percentage of the selected tutoring sound character by female participants

9

10

A. Alsumait, A. Al-Osaimi, and H. AlFedaghi

Interestingly, results also reveled that the male child participants have selected only the boy sound charter "Salem" (50%) along with the teacher sound character "Hanan" (50%) (Figure 5.b). Similarly, the female child participants have selected only the girl sound character "Abeer" (53%) along with the teacher character "Hanan" (47 %) (Figure 5.c). This indicates that children prefer only the teacher sound character along with the child tutor sound of the same gender. Guideline 5 To motivate both genders and gain their concentration, use neutral character (e.g. a teacher). To motive gender-specific participants' group and gain their concentration, use either a neutral character or an equivalent gender.

3.7 Line Length This research also investigated the effect of screen line length on reading performance and satisfaction. The same paragraph was presented in three different screen line lengths. Despite the fact that there were no significant differences in satisfaction scores (see Figure 6), a line length that supports faster reading could impact the overall experience of e-learning programs. Reading rates were found to be the fastest (4.59 sec/word) with the 2/3 screen line length, (5.44 sec/word) with the full screen line length and to be the slowest (6.56 sec/word) at the 1/3 screen line length. Participants reported either liking or disliking the extreme screen line lengths (1/3 and full line length). Those that liked the 1/3 screen line length indicated that the short line length helped faster reading and was easier because it required less eye movement and the paragraph seemed neat and clear. Those that liked the full screen line length stated that they liked having more information on one line and believe that it seemed to them that they are reading less information.

40

Percentage (%)

30

20

10

0 1/3 Screen Line Length 2/3 Screen Line Length Full Screen Line Length

Fig. 6. The percentage of the preferred screen line length

Arab Children’s Reading Preference for Different Online Fonts

11

Although some participants reported that they felt like they were reading faster at 1/3 screen line length, this condition actually resulted in the slowest reading speed. From this study, it may be beneficial to use 2/3 line lengths when possible. Guideline 6 Use 2/3 screen line length to speed the reading.

4 Conclusion Readable and satisfying e-learning interfaces are achieved through following e-learning design guidelines. These design guidelines need to be configured to suite some language characteristics. Therefore, this paper aimed to build a set of design guidelines for Arabic e-learning programs designed for seven-to-nine years old children. The results of evaluating the e-story indicated that children preferred to read text using Simplified Arabic or Arial Unicode MS with size 14-point. These font type and size combinations were considered, among the different font type and size combinations used in this study, to be the easiest and fastest font type and size combinations to read. Conversely, children, for font attractiveness, chose other preference. They preferred to read using Arial Unicode MS with size 14-point. Generally, font types with 14-point size were preferable over the other font types with 12-point size. In addition, participants were able to read faster with 2/3 screen line length. When it comes to the tutoring sound character, children preferred to listen to a female adult tutoring sound, which is closer to their actual life style. Tutoring sound characters can play a critical success factor in learner acceptance of e-learning programs. Larger combinations of font types and sized need to be assessed by Arab children. Further assessments are also needed to investigate the influence of other variables on the reading experience. Examples of those variables are: text color, effect of emphasizing pieces of text; bold, italic or underlined text, paragraph spacing, eye movements, and scrolling movements.

References 1. Bernard, M.L., Lida, B., Riley, S., Hackler, T., Janzen, K.: A comparison of popular online fonts: Which size and type is best?. Usability News. 4.4 (2002) 2. Bernard, M.L., Melissa, M., Talissa, F., Jan, M.: Which Fonts do Children Prefer to Read Online?. Usability News. 3.1 (2001) 3. Chaparro, B., Dawn Shaikh, A., Chaparro, A.: Examining the Legibility of Two New ClearType Fonts. Usability News. 8.1 (2006) 4. Dyson, M.C.: How physical text layout affects reading from screen. Behaviour & Information Technology 23(6), 377–393 (2004) 5. Hemayssi, H., Sanchez, E., Moll, R., Field, C.: Designing an Arabic user experience: methods and techniques to bridge cultures. In: Proceedings of the Conference on Designing for User Experiences DUX 2005 (2005) 6. McPherson, M.A., Nunes, J.M., Zafeiriou, G.: New Tutoring Skills for Online Learning: Are e-Tutors adequately prepared for e-learning delivery? In: Proceedings of EDEN 2003 The Quality Dialogue; Integrating Quality Cultures in Flexible, Distance and e-learning, Rhodes, Greece, June 15-18, 2003, pp. 347–350 (2003)

Adaptation Decisions and Profiles Exchange among Open Learning Management Systems Based on Agent Negotiations and Machine Learning Techniques Silvia Baldiris, Ramón Fabregat, Carolina Mejía, and Sergio Gómez Institute of Informatics and Aplications (IIiA), Universitat de Girona, Spain [email protected], [email protected], {carolina,sergiog}@eia.udg.edu

Abstract. We have developed some projects [1,2] for addressing the heterogeneity problem in open learning management systems (LMS). In [3], an independent adaptation platform to support competences development through personalization is presented. Three user characteristics (competences profile, learning style, and accessing context) are modeled by means of analyzing user interaction data in a LMS. This process is supported by the assigment of independent adaptation tasks to different JADE intelligent agents. In this paper we introduce some negotiation strategies among those intelligent agents in order to: 1) select the best types of adaptation through collaborative tasks, and 2) generate standards and exchangeable user profiles based on the inferred user characteristics, describing the mechanisms to mobilize these profiles between different LMSs. These profiles support the generation of specifics learning designs for each particular user. Keywords: competence development, adaptation, intelligent agents, adaptive hypermedia, machine learning.

1 Introduction Nowadays, learning management systems (LMS) are information intensive systems, which manage large amounts of knowledge in a database or in content repositories. ADAPTAPlan project (TIN 2005-08945-C06-00) deals with the conceptual modeling problem for this kind of systems providing a user modeling and adaptation processes [3,4] for delivering adjusted learning path to users. In this way a conceptual adaptation platform (ADA+) have been designed considering a set of independent agent tasks. The agent tasks are supported in the definition of several elements of the learning process in the LMS (competences, learning resources, metadata), in the user interaction monitoring and in the use of machine learning techniques in order to know the user. However, the information in the e-learning platform can be limited in order to generate the adjusted learning design for a particular user and produce unwanted results. On the other hand, use of XML-based models such as IMS specifications facilitates the interaction, information distribution and exchange among LMS. This is one of the J.A. Jacko (Ed.): Human-Computer Interaction, Part IV, HCII 2009, LNCS 5613, pp. 12–20, 2009. © Springer-Verlag Berlin Heidelberg 2009

Adaptation Decisions and Profiles Exchange among Open LMS

13

main issues to be addressed in the context in an open learning environment because of the impact that it could have in the learning design process. In this paper, we have focused on how to establish collaborative work among the agents in the ADA+ adaptation platform and we present a general schema for the IMS specification exchange among e-learning platforms. For the first problem we propose the definition of an optimization problem to relate the tasks developed by each agent in the platform and the cost associated to the development of these tasks. The second problem will be addressed by combining semantics and structural techniques for similarity analysis among XML documents. The paper is structured as follows. First, in the section 1 an approach for the state of the art is presented, and its relationships with the problems related in the paper. Next, the two problems mentioned above are described. In the section 4, a detailed description about the proposal is presented and finally some conclusions and future works are shown.

2 Related Works In this section we introduce some related work and its relationship with our research works. The first research area related with our proposal is the use of intelligent agents technologies in education. Many educational systems are known to reduce navigation effort, time to achieve the learning goal, and learner retention, and to increase quality of learning. In particular, Shaboo [5], Mas-PlanG [1], MAS-SHAAD [2] and aLFanet [6] are projects developed by Universidad Industrial de Santander, University of Girona and UNED, respectively, which show how user modelling processes and adaptation take into account different user characteristics independently to improve learning. Some of these projects use intelligent agents to develop adaptation tasks but frequently this set of agents do not collaborate among them to achieve a common objective. Our research efforts are oriented to include some characteristics from adaptive hypermedia systems in the context of LMS, taken into account the autonomy and mobility of the agents. An adaptation platform [3] was developed with this purpose but it is necessary to improve our work including collaborative task among the agents in the platform. This is the first objective in this paper. Conceptual modeling to facility the interoperability among LMS is the second research line of our interest. IMS specifications are one of the main elements in the user modeling and adaptation process. They are exchangeable XML documents with an associated schema. The analysis of the structural and semantic similarities between XML and with the problem of detecting change among these kinds of documents is another large research area. Tendencies for the analysis of the structural similarity in a XML document can be summarized into three groups. The first one considers a XML document as an ordered tree [7]. The second one is based on the tag similarity [8], and the third one consist on the interpretation of the XML document as a time series which applies the Discrete Fourier Transform [9], analyzes the frequency of this signal and infers the similarity measure.

14

S. Baldiris et al.

Some algorithms such as MH-Diff [10] and LaDiff [11] have been developed for detecting change between structured documents algorithms. For the specific XML documents, XMLTreeDiff [12], XyDiff [13], X-Diff [14] has been developed. Our intention is not propose a new algorithm but to use the best one with the purpose of generating a full user model exchange. The user modelling process defines and maintains up-to-date user models, where user models are a representation of information about an individual user that is essential for an adaptive system to provide the adaptation effect [15]. The result of this process is a user model that consists in different user characteristics and can be expressed in different formats according to the system that stores and exchanges it. User model servers or shells are systems in charge of maintaining the user models and they generally have the follows characteristics: 1) it use a particular language to represent the user characteristics, 2) require specific collaborative filtering techniques in order to create stereotypes of user groups, 3) count with a user history recorder, and also 4) apply techniques such as user models retrieval, update and validation. Some of the developments in this area are GUC [16], UM [17], BGP-MS [18], DOPPELGÄNGER [19], TAGUS [20], GUMS [21], PROTUM [22], UMT [23]. A main issue regarding the exchange of information between user model servers is the analysis of the information that it contains. This problem is related for us with the XML similarity analysis.

3 Problem Description Our research works are based on the following two hypotheses: The first one is to support the learning process with the aim to improve the learning. In this case, the knowledge about the users (user modeling) such as previous competences, individual preferences and human values has to be considered [1] and the learning process has to be adaptive according to this knowledge. The second hypothesis is that in an Table 1. Characterization of the framework elements to support competence development

Specifications Exchange Service Agents Technologies • Technological Standard Based Learning Adaptation Process Agent Technologies • Platform Independence • Technological Standard Based Decision Generation Process Machine Learning Techniques Collaborative Filtering • Platform Independence • Technological Standard Based User Modelling Process Machine Learning Learning style, competences, etc. Independent Agents Platform LEARNING PROCESS MODELLING (Competences, Course Structure, Learning Resources and Activities, Evaluation Another Necessary Models)

Adaptation Decisions and Profiles Exchange among Open LMS

15

open learning environment the possibility to interchange user profiles based on competences can facilitate the management of the learning process. Table 1 presents a general vision of different process to verify these hypotheses. In [4] we describe a user modeling and e-learning adaptation process based in the use of different IMS Specification. The user modeling process defines the method for inferring some user characteristics. In particular, the specific competence level, the collaborative competences level, the user learning style and the context. User modeling process supports a specific learning adaptation process [3]. The result of this process can be an IMS Learning Design (LD) level B, which is a course that takes into account some user features. This adapted learning design permits to deliver learning objects (LO) and activities according to the values of user characteristics presented above. The ADA+ adaptation platform was developed with this proposes and consists in different intelligent agents, which have particular adaptation tasks (see table 2). Table 2. Specific Agent Task in the adaptation platform

Specific Agent Task Planner Agent Specific Competence Agent Collaborative Competence Agent Learning Style Agent

Context Agent

Task description − Generates an adapted IMS-LD (level B) for users. − Generates the planning problem using competence definition and the LO metadata. − Defines the collaborative level of the user using a grouping behaviour. − Defines the delivering order of the learning resource type using automatic classification techniques. − Monitors the change in the user learning style according to the user interaction. − Selects equivalent LO according to the user access device and accessibility user preferences.

Actually, agents in ADA+ adaptation platform carry out some tasks independently. The reason behind this is that we are interested in creating a flexible collaborative environment among intelligent agents. This environment is meant to consider additional users’ characteristics when needed. However, we have analysed that the problem of context limitations in the virtual environment can produce undesired adaptation results. For these reason, one of the purposes of this paper is to introduce an approach based on the definition of an adaptation decision process in order to identify the types of adaptation that can be offered in a specific moment according to the learning context limitations. On another hand, one of the main purpose for applying specifications and standards in education technologies is to facilitate the information exchange among LMS.

16

S. Baldiris et al.

IMS consortium has worked in the generation of different kind of specifications with the objective of creating a standard e-learning language and increase the systems interoperability. These languages permit to specify products and sub-process of the learning process such as competence, learning designs, and assessments among others. We present the Specification Exchange Service (SES) as a mechanism for exchange information among e-learning platforms, in particular, IMS specifications. It proposes a general framework supported in an adjusted version of X-Diff algorithm [24]. SES could support the decision process by increasing the quantity of resource than can be used by the agents in ADA+.

4 Proposal Description In this section, we present our proposal to establish a collaborative work among the agents in the adaptation platform, and present a general schema for the IMS specification exchange among e-learning platforms. 4.1 Adaptation Decision through Detection of Context Constrain and Negotiation When the adaptation idea is introduced in the context of a virtual learning process, the starting point is still the modelling of the teaching-learning process as in the traditional (face to face) process. However, the initial work for the teacher increases because she needs to perform new activities using a LMS and adjust the environment according to the requirements. If the teacher does not provide the necessary information to the systems, multiples context limitations can be detected in the environment. These limitations impact negatively the automatic generation of the adaptive learning design. Table 3 summarizes the different limitations that can be found in the learning environment. Table 3. Agent tasks vs.context limitations

Specific Agent Task Specific Competence Agent Collaborative Agent Learning Style Agent

Context Agent

Context Limitations − Existence and quality of the Competence definition. − Existence and quality of the LO metadata according to the needed model. − Amount and quality of users interaction in collaborative tools. − Amount of learning objects according to the different LO metadata resource types. − Amount and quality of users interaction with the different LOM resources types. − Existence of the equivalent LO for the topics.

Adaptation Decisions and Profiles Exchange among Open LMS

17

Table 4. Adaptation Decision Variables

Context Constraint Variables competence_definition_exist competence_definition_complete collaborative_interaction_exist collaborative_interaction_enough learning_objects_type_exist learning_objects_type_enough equivalen_learning_objects _exist equivalen_learning_objects_enough objects_metadata_exist objects_metadata_complete objects_metadata_enough

Values boolean integer boolean integer boolean integer boolean integer boolean integer integer

Description 1 if competence definition exist % 1 if collaborative interaction exist % 1 if learning objects exist % 1 if equivalent learning objects exist % 1 if objects metadata exist % %

These constraints can be expressed in terms of the specific variables. Table 4 shows some of them. In ADAPTAPlan project each particular agent can define the value of specific variables and they use them as the input of the decision process. The natures of the variables permit us model the decision problem as an optimization problem, where the objective is to reduce the cost of accessing distributed resource repositories in order to obtain the necessary information to generate the adaptation task. In this way, the decision process can be supported in a cost matrix where different source and its cost are related in terms of different characteristics as network traffic needed, purchase cost, availability, and importance, among other. Taken into account the cost associated to the necessary resources for each particular adaptation task and also the cost associated to combinations of adaptation tasks, the decision is defined. As a result of this collaboration process based on the cost of the different resources, an adapted IMS-LD is generated. 4.2 Specification Exchange Service (SES) Consider two different e-learning platforms where the corresponding students developed different learning activities. Each particular platform is associated to a specific user modeling system and user model shell system. As was described above, the user modeling system obtains user model for the users through the analysis of user interaction and the user model shell system maintain the model (see figure 1). The question is: how can the information in these systems be shared? In last years, XML have become the most common language to represent information to be shared about the learning process and the people. Indeed, IMS specifications use XML as a representation language and schemas for the structure definition and validation.

18

S. Baldiris et al.

Fig. 1. Exchange Problem

For this reason we have tested some algorithms for the analysis of the similarities between XML documents, algorithms were shown in the section 2. We have selected diffX algorithm [24] in order to detect the structural changes in the XML documents and for supporting the generation of a new user model which is a mix of previous two documents. diffX algorithm was selected for including move operation that another tested algorithm does not implement. The process of the new user model generation is shown in the figure 2.

Fig. 2. Complete User Model Generation

The first part of the process, matching, considers the XML documents as an ordered tree, identifying isolated tree fragment mapping (i.e. an iterative top-down mapping technique). The result of this mapping is the definition of the set of sub trees that are the same in both trees. This identification permits to determinate the set of necessary operations to transform one document into another. The set of operations are applied to the document of the LMS that has generated the request. Then the document is sent to the requester LMS. Semantic analysis of the XML document in the matching process it is supported in the schemas associated to different IMS specifications.

5 Conclusion and Future Works In this paper we deal with two problems associated to the use of adaptive e-learning platforms and we propose an approach for each of them. The first one is about how the incomplete learning process modeling can produce undesired adaptation results in the LMS. We propose a strategy based on the collaboration among the intelligent agents in order to generate an adequate adaptation decision. The decision strategy is supported by a cost matrix where different source and its cost are related in terms of

Adaptation Decisions and Profiles Exchange among Open LMS

19

different characteristics as network traffic needed, purchase cost, availability, and importance. The second problem is related to the need of information exchange among elearning platforms, user profiles exchange, competence definition or any other type of information, in order to address this problem one approach supported in diffX algorithm was introduced. Proposals presented in the paper are under construction in the context of A2UN@ project (TIN2008-06862-C04-02/TSI). In particular, different types of repositories for supporting the agent collaboration are under construction. Testing process for comparing different algorithms have been developed and the integration of different syntactic and semantic algorithms is been implemented.

References 1. Peña, C.I.: PhD Thesis: Intelligent agents to improve adaptivity in a web-based learning environment. Universidad de Girona (2004) 2. Mérida, D., Cannataro, M., Fabregat, R., Arteaga, C.: MAS-SHAAD a Multiagent System Proposal for an Adaptive Hypermedia System. In: Proceedings of IJCEELL journal Special issue: Adaptivity in Web and Mobile Learning Services (2004) 3. Baldiris, S., Santos, O., Huerva, D., Fabregat, R., Boticario, J.G.: Multidimensional Adaptations for Open Learning Management Systems. Accepted at TUMASA Workshop of 2008 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology, Sydney (Australia), December 9-12 (2008) 4. Baldiris, S., Santos, O.C., Barrera, C., Boticario, J.G., Velez, J., Fabregat, R.: Integration of educational specifications and standards to support adaptive learning scenarios in ADAPTAPlan. International Journal of Computer Science and Applications (IJCSA). Special Issue on New Trends on AI techniques for Educational Technologies 5, 1 (2008) 5. Moreno, G.D., Baldiris, S.M.: Degree project memories: Adaptive Hypermedia System for Teaching Object Oriented Programming. Universidad Industrial de Santander (2003) 6. Santos, O.C., Boticario, J.G.: Meaningful pedagogy via covering the entire life cycle of adaptive eLearning in terms of a pervasive use of educational standards: the aLFanet experience. In: Nejdl, W., Tochtermann, K. (eds.) EC-TEL 2006. LNCS, vol. 4227, pp. 691–696. Springer, Heidelberg (2006) 7. Shasha, D., Zhang, K.: Fast algorithms for the unit cost editing distance between trees. Journal of Algorithms 11, 581–621 (1990) 8. Buttler, D.: A Short Survey of Document Structure Similarity Algorithms. In: International Conference on Internet Computing 2004, pp. 3–9 (2004) 9. Flesca, S., Manco, G., Masciari, E., Pontieri, L., Pugliese, A.: Fast Detection of XML Structural Similarity. IEEE Transaction on Knowledge and Data Engineering 17(2), 160–175 (2005) 10. Chawathe, S., Garcia-Molina, H.: Meaningful change detection in structured data. In: Proceedings of the ACM SIGMOD International Conference on Management of Data, Tucson, Arizona, USA, May 13-15 (1997) 11. Chawathe, S., Rajaraman, A., Garcia-Molina, H., Widom, J.: Change detection in hierarchically structured information. In: Proceedings of the ACM SIGMOD International Conference on Management of Data, Montreal, Quebec, Canada, June 4-6 (1996) 12. Cubera, F., Epstein, D.: Fast Difference and Update of XML Documents, March 1999. Xtech, San Jose (1999)

20

S. Baldiris et al.

13. Cobena, G., Abiteboul, S., Marian, A.: Detecting changes in XML documents. In: Proceedings of the 18th International Conference on Data Engineering, San Jose, California, USA, February 26 - March 1 (2002) 14. Wang, Y., DeWitt, D., Cai, J.: X-Diff: An effective change detection algorithm for XML documents. In: Proceedings of the 19th International Conference on Data Engineering, Bangalore, India, March 5-8 (2003) 15. Brusilovsky, P., Millán, E.: User Models for Adaptive Hypermedia and Adaptive Educational Systems. In: Brusilovsky, P., Kobsa, A., Nejdl, W. (eds.) Adaptive Web 2007. LNCS, vol. 4321, pp. 3–53. Springer, Heidelberg (2007) 16. Van Der Sluijs, K., Houben, G.: A generic component for exchanging user models between web-based systems. International Journal of Continuing Engineering Education and Life-Long Learning 16(1/2), 64–76 (2006) 17. Kay, J.: The um Toolkit for reusable, long term user models. User Modeling and UserAdapted Interaction 4(3), 149–196 (1995) 18. Kobsa, A.: Modeling the user’s conceptual knowledge in BGP-MS, a user modeling shell system. Comput. Intelligence 6, 193–208 (1990) 19. Orwant, J.: Heterogenous learning in the Doppelgänger user modeling system. User Model. User-Adapted Interact. J. Personal. Res. 4(2), 107–130 (1995) 20. Paiva, A., Self, J.: TAGUS: A user and learner modeling workbench. User Model. UserAdapted Interact. J. Personal. Res. 4(3), 197–226 (1995) 21. Finin, T.W., Drager, D.: GUMS1: A general user modeling system. In: Sixth Canadian Conference on Artificial Intelligence, Montreal, Canada, pp. 24–29 (1986) 22. Vergara, H.: PROTUM: a prolog based tool for user modeling. WIS-Report 10, WG Knowledge-Based Information Systems, Department of Information Science, University of Konstanz, Germany (1994) 23. Brajnik, G., Tasso, C.: A shell for developing non-monotonic user modeling systems. Int. J. Human-Computer Studies 40, 31–62 (1994) 24. Al-Ekram, R., Adma, A., Baysal, O.: diffX: An Algorithm to Detect Changes in MultiVersion XML Documents. School of Computer Science, University of Waterloo (2005)

Accessing e-Learning Systems via Screen Reader: An Example Maria Claudia Buzzi1, Marina Buzzi1, and Barbara Leporini2 1

CNR-IIT, via Moruzzi 1, 56124 Pisa, Italy [email protected], [email protected] 2 CNR-ISTI, via Moruzzi 1, 56124, Pisa, Italy [email protected]

Abstract. The evolution of the Information and Communication Technology (ICT) and the rapid growth of the Internet have impelled the pervasive diffusion of e-Learning systems. This is a great opportunity for visually-disabled people provided that both the interactive environment, created by the Learning Management Systems, and the Learning Objects, created by teachers, are properly designed and delivered. In this paper we investigate interaction of the blind user with an open source Virtual Learning Environment (Moodle) and discuss how the use of the W3C Accessible Rich Internet Applications (ARIA) suite may improve the experience of navigation via screen reader. Keywords: e-Learning, accessibility, usability, blind, ARIA.

1 Introduction Today Learning Management Systems (LMSs) offer many educational tools from a single environment: web courses, exercises, chats, wikis, self-assessment SW, surveys, forums, podcasts and even more. User Interfaces (UIs) then become richer with embedded videos, text, sounds, customized widgets. Accessibility and usability of eLearning systems and objects are crucial for providing an easy and satisfying experience to all. Specifically, Virtual Learning Environments (VLEs) should be friendly and simple to use for everyone, in order to eliminate any “technical barriers” to the learning process and allow students to concentrate on content. However, despite considerable research focus in this field, interacting with a virtual environment and using learning objects is still difficult for a blind user who cannot see the screen and is unable to use a mouse. Furthermore, interaction requires the aid of assistive technology (i.e., screen reader and voice synthesizer), which adds another degree of complexity. Designers of e-Learning systems must consider three crucial factors: usability, accessibility, and educational effectiveness. Consequently, the challenge is to design systems that are simple to use and accessible to all, while maintaining pedagogical and educational efficacy. In particular, blind students may fruitfully utilize e-Learning systems if educational materials are accessible and learning paths can be tuned to the “rhythm” of the individual student. J.A. Jacko (Ed.): Human-Computer Interaction, Part IV, HCII 2009, LNCS 5613, pp. 21–30, 2009. © Springer-Verlag Berlin Heidelberg 2009

22

M.C. Buzzi, M. Buzzi, and B. Leporini

When designing for blind users, it is necessary to consider the three main interacting subsystems of the Human Processor Model: the perceptual, motor and cognitive systems [5]. Sightless persons perceive page content aurally and navigate via keyboard. This makes the “reading process” time-consuming and sometimes difficult and frustrating, if the contents are not designed with special attention to their needs. The cognition part of the interaction is important, since many learning techniques are only relevant to people with good vision and may not apply to someone with a visual impairment. Thus, alternative ways to deliver the same content should be provided. Furthermore, a blind person may develop a different mental model of both the interaction and the learning processes, so it is crucial to provide an easy overview of the system and contents. Non-visual perception can lead to: 1. Content serialization. The screen reader reads the contents sequentially, as they appear in the HTML code. This process is time-consuming and annoying when part of the interface (such as menu and navigation bar) is repeated in every page. As a consequence, blind users often have to stop the screen reading at the beginning, and they prefer to navigate by Tab Keys, from link to link, or explore the content row by row, via arrow keys. 2. Content and structure mixing. The screen reader announces the most important interface elements such as links, images, and window objects as they appear in the code. For the blind user, these elements are important for figuring out the page structure, but require additional cognitive effort. 3. Table. If the table’s content is organized by columns the screen reader (which reads by rows) announces the content of the page out-of-order, and consequently the information might be confusing or misleading for the user. 4. Lack of context. When navigating by screen reader the user can access only small portions of text and may lose the overall context of the page; thus it may be necessary to reiterate the reading process. 5. Lack of interface overview. Blind persons do not perceive the overall structure of the interface, so they can navigate for a long time without finding the most relevant contents. 6. Difficulty understanding UI elements. Links, content, and button labels should be context-independent and self-explanatory. 7. Difficulty of working with form control elements. The new JAWS version (v. 10) simplifies the interaction with forms since it can automatically activate the editing modality (for text input) when the virtual focus arrives at the text box (for instance when the user presses the tab key). However, with previous screen reader versions the user may have great difficulties since switching between exploration and editing modalities is required (i.e. form mode on/off). 8. A blind person is unable to access multimedia content such as video streaming, video conferencing, and captioning. If an alternative description is not present the user may lose content. In this paper we analyze accessibility and usability for the blind of two demo courses offered by Moodle (http://www.moodle.org/), a very popular open source eLMS. Then we illustrate how ARIA, the suite developed by the Web Accessibility Initiative (WAI) group of W3C, may facilitate interaction for the blind. Lastly, the paper presents our conclusions.

Accessing e-Learning Systems via Screen Reader: An Example

23

2 Related Works E-learning systems pose new challenges with respect to classic user-centered product design, where the target is a set of homogeneous users. Learner-centered Design must answer to the needs of multiple learner categories due to differences in learning strategies, know-how, experiences, motivation to learn and, not least, user age and ability. If appropriately designed and implemented, e-Learning systems are more effective and useful than classroom learning [4]. Various studies focus on the usability of e-Learning systems and some also include a general discussion on accessibility, but to our knowledge only a few focus on totally blind persons. Fifteen years ago Nielsen proposed an informal method for evaluating systems and design usability based on verifying the conformance to a set of principles of usable software design (heuristics) performed by experts [8]. This approach, which detects a high percentage of problems, is cost-effective and easy to implement compared to usability testing. However, to be effectively applied to the e-Learning domain these general principles needed to be further refined. A few years later, Squires and Preece embedded usability heuristics in the socio-constructive theory and specified criteria ad hoc for e-Learning [12]. However, even today various researchers criticize the lack of accurate studies in this field. In [1] the authors take the first steps in defining a methodology for the rigorous evaluation of e-Learning applications, but accessibility for special needs students is not analyzed. Furthermore, Zaharias critically examined the usability of e-Learning applications and proposed a new usability measure: the student’s intrinsic motivation to learn [15]. Developing a usability evaluation method based on a questionnaire, he carried out two large empirical studies showing the reliability of this approach. For Sloan et al. the goal of universal accessibility on the Web is inappropriate and instead it is necessary to explore multiple routes to provide equivalent experiences [11]. As Kelly et al. argued, rather than demanding that an individual learning resource be universally accessible, it is the learning outcome that needs to be accessible [7]. Based on user profiles, metadata and dynamic connection to resources, the user’s experience can be customized to match his/her abilities. Then an appropriate design is crucial for improving the accessibility and usability of eLearning Systems. De Marsico et al. [5] defined methodological guidelines involving users with disabilities as well as pedagogical experts in the development process, believing that input of different know-how may enrich the quality of e-Learning applications, and provide a more satisfying learning experience. They also include two examples of building and providing learning objects accessible respectively to visually- and hearing- impaired students. Rodriguez et al. describe a project for improving e-Learning experience for the visually impaired, based on ethnomethodology and taking into account psychosocial issues, the user context and experience [9]. Next they created different learning object formats suitable for the blind, including DAISY (Digital Accessible Information SYstem). However, although authors describe the methodology used to improve learning materials no general guidelines are offered to the reader. Within the framework of a project aimed at providing an accessible e-Learning platform for disabled and adult learners, Santos et al. [10] illustrate a methodology for developing standard based accessible courses which use two-step evaluations. However for the totally blind, more specific UI features are necessary than those provided

24

M.C. Buzzi, M. Buzzi, and B. Leporini

in this study, such as providing a page overview, full control of interface elements and easy and rapid navigation via keyboard. E-Learning is a great opportunity for visually-disabled people, provided that both the interactive environment (created by the LMS) and the learning objects are properly designed and delivered. In this paper we mainly focus on accessibility and usability of the virtual environment for the blind, taking as an example the environment offered by Moodle. Concerning Moodle, Debevc et al. [4] compared usability of two LMSs, i.e. Moodle vs a proprietary system, when delivering an identical package of educational materials. This usability study was carried out with a SUMI (Software Usability Measurement Inventory) evaluation that is not specific for e-Learning environments. However indications derive from this study, that point out Moodle’s weaknesses (vs the proprietary system) in efficiency as well as learnability.

3 The Moodle Virtual Learning Environment To evaluate accessibility and usability of a Virtual Learning Environment (VLE) via screen reader, we chose Moodle, an Open Source LMS which offers a rich VLE. Thanks to its large spectrum of available tools and options it has become very popular worldwide. Specifically it offers (http://www.moodle.org): Assignment modules, Blogs, Chat, Choices with multiple response options, Course resources (Moodle pages, uploaded files or web links), Databases, Forums, Glossaries, Interaction, Lessons, SCORM packages, Surveys, Quiz module, Wikis and Workshop modules. The richness of the VLE is Moodle’s strength but the environment‘s complexity can also create difficulties in interaction via screen readers. 3.1 Evaluation Methodology In the first step of our research, we analyzed two demo courses available on the website (http://demo.moodle.org) by using the screen reader JAWS for Windows (http://www.freedomscientific.com) v. 9.0 and 10. We used both the MS IE version 7.0 and the Mozilla Firefox version 3.0.5 browsers. The test was carried out by all the authors independently; afterwards, outcomes were compared and integrated. One author has been totally blind since childhood and uses the JAWS screen reader every day; thus she knows this tool’s functions very well and is able to use advanced commands. By analyzing the test results we noticed that in spite of her great expertise using JAWS, she was unable to perceive the exact structure of the layout. This paper’s sighted authors carried out the test using only JAWS basic commands, but viewing the Moodle UI allowed them to rapidly understand the origin of obstacles encountered when interacting via screen reader. Therefore, integrating both these outcomes led to a more accurate analysis. The different experiences of the authors when using JAWS allowed us to cover a variety of interaction modalities: i.e. both basic commands, which simulate the use of beginner users, and advanced screen reader functions. 3.2 Exploring the Moodle Demo To illustrate how a blind user interacts with the Moodle environment, we explore several pages related to two selected courses, according to specific features to be

Accessing e-Learning Systems via Screen Reader: An Example

25

tested: main page functions, topic outline of a selected course, interacting environment, and so on. When opening the demo page, before doing the log-in, JAWS announces four headings. Through the special command “Insert+F6” JAWS shows the headings list. Although different levels are assigned to these headings, the levels order used is unsuitable. In fact, level 1 is associated with the course title; thus, by pressing the key “1” the virtual focus moves to the beginning of the page. In this case, it would be more appropriate to assign the level 1 to the “Log in” section, which logically is the first part of the interface the user must navigate to enter the system. After the login, the list of available courses is shown. Once the chosen course is clicked, its page is loaded and opened. The screen reader informs the user about headings as well as the number of links. Basically, this page seems to be accessible, because several accessibility features are detected (for instance, a hidden link pointing to the “main content”) by the screen reader. Also, headings are immediately identified and announced by JAWS. But, are all these features suitable for comfortable navigation by blind users? Let us consider the “Higher Education Film Studies Module” (in the following referred to as Course A), available as a demo course.

Fig. 1. Moodle demo - Course A

After selecting this course the following elements are detected by JAWS: • Headings: 32 headings are announced. By exploring the heading list (Insert+F6 command) we note that most of them refer to the course, but the others are related to the e-learning tools (e.g. chat, blog, and so on). Those headings are too numerous to be navigated comfortably. Although headings are used, the page is too long to be read by keyboard via screen reader. Furthermore, it is not very clear which headings refer to the course modules and which ones to the interacting platform functions (see below).

26

M.C. Buzzi, M. Buzzi, and B. Leporini

• Layout table: to show the Topic outline a table has been used for the layout. Specifically two tables are used, the second nested in the first cell of the first table, but JAWS announces only one table. Although headings are used to split the long content, a table should not be used for the rendering. Moreover, the “summary” attribute of the table is “table layout”, totally useless. A more appropriate summary value could be “table of contents”; in this way, by just pressing the letter “t” JAWS will announce immediately “Table of contents”. If a layout table is used, at least a meaningful summary should be applied. • Number of links. The page contains too many links (in our case 115 links). • Images. Many of the images in the pages of the course are not perceived by the screen reader since no alternative text is provided or descriptions are negligible. After the log-in as teacher, we tried to insert an image and found that the alternative text is now mandatory. However there is no control on the number of characters inserted, so it is possible to skip this edit field with only a space press. In order to understand whether the issues encountered in the selected course could be generalized, we attempted to navigate another demo course. Specifically, we picked up the course “Moodle Features Demo” (in the following referred to as Course B). Also the main page of the second course includes a great number of links, two tables used for visual layout, and 25 headings. However, the level assigned to each heading as well as their arrangement differ from Course A. In this case, first the headings for interacting with the Moodle environment are available, and then the Topic outline. Consequently, it seems the UI is not consistent among the courses even if they are very similar (probably due to changes in the Moodle versions used to create these courses). However, headings arranged in this way might make the page structure somewhat unclear. In fact, it is difficult to know how many modules compose the course and their titles and contents, since the module number has not been included in the tag and thus does not appear in the headings lists (Fig. 3); to obtain this information the user must read the page in a sequential way, as shown in the Fig. 2 – left section. All headings are mixed and not well grouped by clearly indicating which are modules (See Fig. 3), and which are associated with the virtual environment (on the right side in Fig. 1). In other words, the main regions: “menu”, “topic outline” and “interacting actions” should be better indicated. For example, by opening the page related to “eXe SCORM package”, the screen reader recognizes some unclear elements: • various graphical icons; the image tags have the ALT attribute, but the alternative descriptions used are not clear for the type of the icon. For instance, several links are preceded by an icon labelled “Completed” (See Fig. 2 – right section). It is not clear what this means. • Frame: in the page a frame called “scoframe1 frame” is identified by JAWS, which is not particularly clear. Even if frame content is explored in a sequential way, is unclear what the student. According to the last version of accessibility guidelines [WCAG 2.0] frames should be avoided (deprecated). The following figure reports the page content as it is interpreted by the screen reader JAWS.

Accessing e-Learning Systems via Screen Reader: An Example

COURSE A Heading level 2 Topic outline Summary: Layout table table with 3 columns and 18 rows Heading level 2 Out of the Digital Dungeon: Exploring Gender and Technology in Sci Fi, … list of 15 items All Art work on this Course Page: Alex... ...... This topic 1 list of 9 items Heading level 2 Getting Started ... Heading level 4 Ice Breaker Activity ... 2 list of 15 items Heading level 2 Genres, Gender … in context Art work: Alex Ronald, Judge Dredd & 2000 AD Copyright © Rebellion A/S 2004. All Rights Reserved. ... 8 Heading level 2 Learning Support Space ... Heading level 4 Continued Support for your … … List end ... Table end Heading level 2 Online Users ... Heading level 2 Activities Link Chats Link Choices ...

27

COURSE B Features Demo: eXe SCORM package Heading level 1 Moodle Features Demo You are logged in as Link Sam Student (Link Logout) Heading level 2 You are here list of 4 items Link Moodle Demo/ ... ? eXe SCORM package list end … eXe Scorm Package Link Expand/Collide Graphic Completed Link eXe Scorm Package list of 5 items nesting level 1 Graphic Completed Adding an image Graphic Completed ... list end nesting level 1 list end Review Mode scoframe1 frame Adding an image Graphic img6_2 The eXe logo In eXe it is easy to include images with your web content. ... scoframe1 frame end

Fig. 2. Page segments sequentially read by JAWS (up/down arrows): on the left Course A main page, on the right a SCORM module function of Course B

Fig. 3. List of headings of the demo course A (“Insert+F6” JAWS command)

28

M.C. Buzzi, M. Buzzi, and B. Leporini

In short, although the pages seem to be accessible, our initial interaction using the screen reader JAWS shows some usability issues when interacting via keyboard and in a sequential manner. The few pages considered in our preliminary evaluation highlighted some difficulties orienting oneself among different information (i.e. headings and links), as well as in handling conceptual information. For instance, in some cases pictures or graphical representations are used to provide information on modules or on intersecting concepts. In a learning system it is also important to make those concepts accessible and usable by alternative modalities.

4 Applying ARIA When designing a web interface it is essential to keep in mind both usability and accessibility principles. Accessibility is a basic pre-requisite for allowing access to page content, while usability guarantees an easy, simple, efficient, rapid and satisfactory navigation and interaction. Navigation is vital for special-needs persons, and in particular for the blind, because it is crucial for them to be aware of their current location on the webpage and how to return to the beginning, or how to reach a certain point in the materials [4]. The use of ARIA would enhance Moodle’s usability in many ways, as explained and exhaustively illustrated by the ARIA best practises document [13]: • Reduce the amount of unnecessary text announced. Specifically the use of Table as layout may be silently ignored by the screen reader if the table is tagged with the presentation role : … • Definition of regions to allow the user get a page overview. Moodle utilizes heading levels to structure the page and to allow the user to perceive the whole structure. Using only standard (X)HTML code, the usability of keyboard navigation is reduced so blind users are forced to use tabbing for accessing active elements (form elements and links). To simplify interaction and allow easy jumping to main interface regions, developer usually relay on creating a link to the main content or use heading levels to structure the page (since the screen reader gives a table of headings). However the use of headings to mark sections is not consistent across web sites [13]. ARIA allows marking sections with standard specifying XHTML landmarks or defining customized regions: … • Simplifying keyboard navigation. The use of landmarks/regions also allows to simplifying navigation via keyboard since the user may jump from one region to the next by pressing a key (in JAWS v.10 the “;”). Furthermore, the developer using the attribute “flowto” defines the order in which regions should be visited: These are basic considerations to improve the interaction via screen reader with the VLE offered by the LMS, that should be integrated on the basis of all observations reported below. Further improvements are possible in the specific learning object

Accessing e-Learning Systems via Screen Reader: An Example

29

prepared by the teacher; thus it is very important to provide automatic tools to support the easy creation of accessible and usable learning objects.

5 Conclusion In this paper we analyze the usability of the Moodle Learning Environment for the blind. Specifically we analyzed two demo courses provided as examples by the system, highlighting features that could be improved. We mainly focus on the navigability of the virtual environment provided by Moodle, which by offering several educational tools integrated into one system, may create complex interfaces. It is important to notice that LMSs may greatly favours the student learning process since the same educational material may be transmitted anywhere, anytime, at any learning rhythm, in a format suited to each individual's ability. On the other hand, since LMSs automatically add a virtual environment to the educational material, if the virtual environment layout is not appropriately designed with a thorough knowledge of accessibility and usability issues, it may induce problems that could be spread to the learning objects themselves. This highlights the importance of considering usability issues from the beginning of the development of every LMS. Making a VLE suitable for the abilities and skills of all users offers many challenges. When defining the graphical UI it is fundamental to consider the needs of sighted users but the needs of the blind should also be kept in mind when writing the UI code. Specifically, the same information should be provided through both visual and auditory channels, the design should be optimized for reading via screen reader, the UIs should be easy to use via keyboard and no additional cognitive effort should be required of the blind user. In conclusion, we believe that our findings could have general applications and that applying ARIA would enhance usability via screen reader in any Virtual Learning Environment.

References 1. Ardito, C., Costabile, M., De Marsico, M., Lanzilotti, R., Levialdi, S., Roselli, T., Rossano, V.: An Approach to Usability Evaluation of e-Learning Applications. Universal Access In the Information Society 4(3), 270–283 (2005) 2. Card, S.K., Moran, A., Newell, T.P.: The Psychology of Human-Computer Interaction. Lawrence Erlbaum Associates Inc., New Jersey (1983) 3. Debevc, M., Bele, J.L.: Usability testing of e-learning content as used in two learning management systems. European Journal of Open, Distance and E-Learning (2008), http://www.eurodl.org/materials/contrib/2008/Debevc_Bele.htm 4. Debevc, M., Verlic, M., Kosec, P., Stjepanovic, Z.: How Can HCI Factors Improve Accessibility of m-Learning for Persons with Special Needs? In: Stephanidis, C. (ed.) HCI 2007. LNCS, vol. 4556, pp. 539–548. Springer, Heidelberg (2007) 5. De Marsico, M., Kimani, S., Mirabella, V., Norman, K.L., Catarci, T.: A proposal toward the development of accessible e-Learning content by human involvement. UAIS Journal 5(2), 150–169 (2006)

30

M.C. Buzzi, M. Buzzi, and B. Leporini

6. Eijl, P., Pilot, A., Voogd, P.: Effects of Collaborative and Individual Learning in a Blended Learning Environment. Education and Information Technologies 10(1-2), 51–65 (2005) 7. Kelly, B., Phipps, L., Howell, C.: Implementing a holistic approach to e-Learning accessibility (Retrieved), http://www.ukoln.ac.uk/web-focus/papers/alt-c2005/accessibility-elearning-paper.doc 8. Nielsen, J.: Usability inspection methods. In: Heuristic evaluation, pp. 25–62. John Wiley & Sons, Inc., New York (1994) 9. Rodriguez, E.P.G., Domingo, M.G., Ribera, J.P., Hill, M.A., Jardi, L.S.: Usability for All: Towards Improving the E-Learning Experience for Visually Impaired Users. LNCS, pp. 1313–1317. Springer, Heidelberg (2006) 10. Santos, O.C.B., del Viso, J.G., de la Cámara, A.F., Sánchez, S.P., Gutiérrez, C.R., Restrepo, E.: HPCN-Europe 1994. LNCS, pp. 796–805. Springer, Heidelberg (2007) 11. Sloan, D., Heath, A., Hamilton, F., Kelly, B., Petrie, H., Phipp, L.: Contextual web accessibility - maximizing the benefit of accessibility guidelines. In: Proceedings of the 2006 international cross-disciplinary workshop on Web accessibility (2006) 12. Squires, D., Preece, J.: Predicting quality in educational software: Evaluating for learning, usability and the synergy between them. Interacting with Computers 11(5), 467–483 (1999) 13. W3C. WAI-ARIA Best Practices. W3C Working Draft 4 February (2008), http://www.w3.org/TR/wai-aria-practices/ 14. Wilson, R., Landoni, M., Gibb, F.: A user-centered approach to e-book design. The Electronic Library 20(4) (2002) 15. Zaharias, P.: A usability evaluation method for e-learning: focus on motivation to learn. In: Proceedings of CHI 2006 extended abstracts on Human factors in computing systems (2006)

Using Tablet PCs and Pen-Based Technologies to Support Engineering Education Ignacio Casas1, Sergio F. Ochoa2, and Jaime Puente3 1

Department of Computer Science, Pontificia Universidad Catolica de Chile, Chile 2 Department of Computer Science, Universidad de Chile, Chile 3 Microsoft Research, Redmond, WA, USA [email protected], [email protected], [email protected]

Abstract. Several experiences and results of the Tablet PC adoption have been reported, mainly in American universities. Although the benefits seem to be highly interesting, it is not clear if they are replicable in developing countries. In order to try to understand the impact of Tablet PCs on engineering education in Chile, the authors conducted several experiments at the two traditional Chilean universities. This paper reports the experiences and the obtained results, comparing them with those obtained in American universities. Keywords: Tablet PCs, Pen-based Technologies, Engineering Education, Mobile Computing.

1 Introduction Every day more and more persons are adopting Tablet PCs as a way to support educational activities. Several researchers have highlighted the contributions of these mobile computing devices and pen-based technologies as a support of the teachinglearning process [1, 8, 11]. Some of the envisioned benefits are the possibility to do hand-writing annotations keeping the metaphor of a paper notebook, and the capability to share resources, support students mobility and collaboration, and count on a mobile repository of educational material accessible and replicable. However the advantages recognized by the users depend on the features of the teaching-learning scenario. After four years using Tablet PCs in computer science [9] and science education programs [5], we have seen several differences between the benefits identified by instructors and students from developed countries (such as USA) and from a developing country such as Chile. This article presents a preliminary study about the impact the use of Tablet PCs and pen-based technologies have on undergraduate computer science programs at the two main universities in Chile. The study covers the perspective of the students and the instructors. It also analyzes the benefits, challenges and limitations perceived by students and instructors, from a social and technological viewpoint. This study was conducted based on the recurrent use of three applications: Classroom Presenter [2], MOCET [9] and an extended version of MS OneNote [10]. Next section presents the related works. Section 3 lists the benefits of using Tablet PCs for both, students and instructors. Section 4 shows the empirical study carried out J.A. Jacko (Ed.): Human-Computer Interaction, Part IV, HCII 2009, LNCS 5613, pp. 31–38, 2009. © Springer-Verlag Berlin Heidelberg 2009

32

I. Casas, S.F. Ochoa, and J. Puente

by the authors in order to determine the impact that pen-based computing have in engineering education in the two main Chilean universities. Section 5 discusses the obtained results. Finally, section 6 shows the conclusions and the future work.

2 Tablet PCs in High Education During the last years several researchers have been doing experiments in order to determine the impact of using Tablet PCs in education. Some of them believe that a “Tablet PC has the potential to dramatically alter the educational process. This new technology significantly changes the way students and teachers interact. It adds completely new dimensions to classroom interaction by providing digital ink and drawing tools for writing, sketching, and drawing; and for real-time collaboration” [1]. At the moment we have seen reports of particular experiences where Tablet PCs have became interesting tools in certain scenarios. Now the challenge is to identify in which scenario the Tablet PCs can effectively add value. For example, Anderson et al. show how Tablet PCs can be used to improve the knowledge delivery and facilitate the interaction between students and the instructor during a distributed lecture [2]. Berque et al. report the results of using Tablet PC and DyKnow tools to support collaborative problem solving [3]. Davis et al. and Anderson et al. report experiences of using Tablet PC to take notes during a class [2, 4]. In addition, the authors have shown, in previous works, how these mobile devices can be used to improve the evaluation process in computer science courses [9] and also to support the activities of communities of practice in science education [5]. There are also several articles describing educational experiences on physics, mathematics and agronomy, in which Tablet PCs have been used as the supporting platform for the teaching-learning process. Considering the experiences reported in recent literature, and focusing just on the main features of the Tablet PCs, we can say that the main activities to be supported with these devices are the following ones: 1. Taking notes. Students can take notes on the slides the professor shares with them, or just on a digital blank sheet. These digital notes can be organized, shared, searched, emailed and linked to other resources [6]. In that sense, this functionality seems to be highly useful, and better than the paper notebook or even a laptop. 2. Enrich the lecture presentation. Instructors can use Tablet PCs to enrich the information they are presenting during a lecture. It could be used, for example, to do marks or make annotations on the slides, or use the Tablet PC as a replacement for (or supplement to) the black/whiteboard [2]. Since digital annotations made on these devices can be shared, the cognitive load of the attendees is reduced. 3. Support group work. During work meetings the group members (i.e. students and/or instructors) usually carry out brainstorming, discuss proposals/alternatives to carry out the job, and validate ideas or the work done. Usually a black/whiteboard is used to support these activities; however Tablet PCs could also be used for this purpose [3]. A benefit of using these devices instead of whiteboards is that notes written on the space shared by the attendees are easier to record and share. In addition, it is not longer needed to erase the whiteboard because its physical space is completely used.

Using Tablet PCs and Pen-Based Technologies to Support Engineering Education

33

In this list we have excluded the activities that can be performed with a laptop. In other words, we included just the activities requiring the special features of a Tablet PC. Please note there are activities that are variants of the listed ones, which were not included in the list. Summarizing, it seems the use of these computing devices could replace the use of the paper notebook and also the black/whiteboard.

3 Impact on the Students and Instructors Experiences reported in the literature show the students are usually comfortable using Tablet PCs to take notes and perform problem-solving activities [13, 14, 15]. Something similar happens with the instructors’ opinion [5, 15]. In addition, most of the researchers that explore the use of these devices in educational scenarios state that Tablet PCs facilitate the active learning [12, 16][Roschelle et al., 2007]. Considering these previous works, some of the main contributions reported about the use of Tablet PCs, are the following ones: • Benefits for students. Students recognize the possibility to take notes is an important contribution of Tablet PCs usage. The free-style handwriting possibility makes the students more comfortable to express their ideas in sketches or annotations [14]. The possibility to exchange digital resources among the students (or with the instructors) was recognized as an important feature [16]. Finally, students find valuable the possibility to store and manage their courses information in digital format [16]. • Benefits for instructors. Many instructors have found that Tablet PCs not only can replace the black/whiteboard, but also extend the resources pool to be used during a lecture [6]. It makes the lectures more dynamic and interactive [Roschelle et al., 2007]. In addition, the annotations the instructor performs during a lecture can be easily stored and shared with the students. Several researchers report an increment of the students’ interest during the classes and also an improvement in the way used by the instructors to deliver knowledge [13, 16]. Based on the authors’ experience, many of the envisioned benefits are inherent to the use of laptops during lectures. Other benefits, for example the change of students’ and instructors’ attitude during the lectures, could be the result of the redesign of the teaching-learning process performed to take advantage of the Tablet PCs features. Of course, there are several benefits that really are a consequence of these devices usage, for example the possibility to make handwriting annotations on the lecture slides. Next section describes a preliminary study carried out in two traditional Chilean universities, which tries to determine the impact of Tablet PCs in computer science and engineering courses. The study also discusses the obtained results and compares them with the findings reported by other researchers in USA.

4 Empirical Study In order to try to understand the impact that the use of Tablet PCs could have in engineering education in Chile, the authors conducted an empirical study in two traditional

34

I. Casas, S.F. Ochoa, and J. Puente

Chilean universities: Universidad de Chile and Pontificia Universidad Catolica de Chile. This study was performed in computer science courses at the engineering school and it involved several types of Tablet PCs and also software products. 4.1 Experience in the Universidad de Chile This experience involved students of a computer science course (CC51A: Software Engineering) during two semesters: fall 2008 (29 students) and spring 2008 (20 students). Ten Tablet PCs were delivered among the students in order to support two main activities: course examinations using MOCET [9] (Figure 1), and group problem-solving using MS OneNote [10]. In addition, the instructor used Classroom Presenter [2] during half of the course lectures, and students using Tablet PCs took notes during lectures using this software. When the course was finishing, the instructor and the students filled a survey that evaluated the experience of using these devices. Some of the issues identified by the participants were the following:

Fig. 1. Use of MOCET at University of Chile

• The hardware matters. Students using particular models of Tablet PCs (DELL Latitude XT and HP Pavilion TX1000) consistently reported problems to write/erase annotations, regardless of the software they were using. In addition, these students also reported precision problems when these devices recognize the stylus location on the screen. • The working conditions matters. Students and the instructor identified the importance of avoiding situations where only some of the students had Tablet PCs. Some educational activities designed to take advantage of the Tablet PC features, for example the group design, were difficult to perform for students that did not have available these devices or a whiteboard. Therefore, if you are going to perform an activity that takes advantage of these devices’ features, be sure that all the students count with Tablet PCs. • Training is required. Students and the instructor reported an important improvement in the usefulness of the devices and the effectiveness of their work, once they were trained in the use of Tablet PCs. • The instructor mobility becomes reduced. Because the Tablet PC must be connected to the projector through a physical cable during a lecture, the mobility of the

Using Tablet PCs and Pen-Based Technologies to Support Engineering Education

•

•

•

• •

•

35

instructor becomes reduced. After some sessions the instructor started to use MaxiVista [7] as intermediary to recover the mobility. Although the solution was good, it required to have two computers in the classroom to deliver the lecture. Tablet PCs do not replace the blackboard. Although the Tablet PC’s functionalities are comparable to those of a whiteboard (and even better in some aspects), the information shown to the attendees through these devices is replaced when the instructor change the slides. In the case of the whiteboard, it acts as an extended screen of the instructor presentation; therefore it keeps visible the information written on it, even when the instructor changes the slides. In other words, students and the instructor think the whiteboard and the Tablet PC are complementary. Useful to make free-style annotations. Students and the instructor highlighted the capability of these devices to make free-style handwriting annotations. Several software products were used to carry out this activity; all of them were useful and comfortable for the users. They are similar to laptops in several aspects. Students and the instructor identified as an important contribution the capability of mobile computing devices to record annotations and manage information in digital format. They also highlighted the capabilities of these devices to share information and communicate among them in a simple way, even when the users are on the move. However, these functionalities are not only available in Tablet PCs, but also in laptops. Something similar occurs with the possibility to perform simulations, or execute or compile programs during the lectures. In other words, several of the advantages reported as result of the Tablet PCs usage are also present when using laptops. The instructor and the students feel comfortable using Tablet PCs. Users felt comfortable using these devices during the experiences. Although some training was required, the learning effort was worthy. Computing devices could be distracters during lectures. Students and the instructor agreed that mobile computing devices can be a distracting element during the lectures. If they do not have a clear role to play during the lecture, their use compete with the instructor’s speech. The redesign of the lecture style makes them more active. The participants in these experiences recognized the lectures became more active and participative. However, it seems to be a consequence of the lectures style redesign more than a result of the use of Tablet PCs.

4.2 Experience in Pontificia Universidad Catolica de Chile This section describes the experiences of introducing Tablet PCs in two courses at the Pontificia Universidad Catolica de Chile during 2007: “Computing Software Tools for Engineering” which is a workshop course (iic2100) and “Introduction to Programming” (iic1102). The main objective of the experiment was to motivate engineering students to learn and enjoy the art of problem solving and project team work supported by computers. In order to reach such a goal, we started to use Tablet PCs in the classroom and we modified the dynamic of the teaching-learning process in these courses. Tablet PCs were used in the classroom for modeling, problem solving and programming.

36

I. Casas, S.F. Ochoa, and J. Puente

Fig. 2. Use of OneNote at Pontificia Universidad Catolica de Chile

The course iic2100 was the subject of this experimentation during two terms in 2007. Forty students participated in each semester. Twenty one Tablet PCs (HP Compact TC4200) were delivered among the students. These devices were shared by the students and they were used to support problem solving during lectures (Figure 2). The impact of the Tablet PCs usage was measured based on opinion surveys from students, teaching assistants and professors. A similar experimentation process was conducted in the course iic1102: Introduction to Programming, during the Fall and Spring terms in 2007. One hundred students participated each semester. Twenty one Tablet PCs (HP Compact TC4200) were delivered among the students in order to support group problem solving during lectures. In spite of all the positive aspects of the use of technologies in the classroom, a few problems were observed; for instance, annoying delays at the beginning of some activities due to instability of the wireless network and software server (a technical assistant was present in the classroom to solve this kind of incidents). Problems also occurred with software tools and their different versions. It was also observed that at the beginning of the class some students had difficulties to turn on the Tablet PC or manage the stylus-pen, due to simple lack of knowledge (this problem was easily solved). Importantly, some students became distracted with the technology and engaged in Internet searching, chatting, playing or e-mailing, but after several sessions this distraction diminished. The students highlighted, as a positive factor, the possibility to solve problems during classes. They considered the use of Tablet PCs helped to make the lectures more active and interesting. However, they also identified limitations in the problemsolving process because not all of the students counted with one of these devices. Most of the students agreed that the use of technology motivated them. It is also important to note that in both courses there were no students considering Tablet PCs as an inhibitor factor; however they recognized that some training is required to take advantage of these devices features. They considered that Tablet PCs facilitated the discussions and the group work during lectures.

5 Discussion At the moment, the use of Tablet PCs in computer science courses seems to have positive effects. However, there is not a comprehensive and scientific study that

Using Tablet PCs and Pen-Based Technologies to Support Engineering Education

37

shows which are the real benefits of including this kind of technology in our courses. Although using mobile computing devices during lectures brings several benefits, it does not seem to be consequence of using just Tablet PCs. Our experiments have also identified that computing devices can become a distracter factor during lectures, if there is not a clear role assigned to them. These experiments also have shown that the use of Tablet PCs motivates to students and professors, however it is not clear if the use of laptops generates the same effect. Provided the courses involved in the experiments tried to take advantage of the Tablet PCs features, and following the recommendations of other researchers in the area, we redesigned their teaching-learning process. Now these courses are more active and there is more interactive work during the lectures, which produced a positive impact among the participants. However this result seems to be a consequence of the redesign process more than the inclusion of Tablet PCs. The experiences help us to identify that not all of the Tablet PCs provide an adequate support for these activities, and the mix of technology and paper-based work is not a good combination; particularly, if students have to use paper notebooks because there are not enough computers for everybody. Although the students recognized the importance of Tablet PCs for taking notes and facilitating group work, the use of these devices requires a training process. Instructors were happy when using Tablet PCs. They think these devices help to improve the quality of the lectures; however they are not able to replace the black/whiteboard. In one of the experiences, a lack of the instructor mobility was identified because of the use of Tablet PC to deliver the lecture. Such problem was then solved using additional software technology.

6 Conclusions and Future Work The use of Tablet PCs in engineering education seems to be growing more and more every day. Several researchers have reported experiences of the Tablet PCs adoption, mainly in USA. They have identified a list of benefits derived from the use of these devices. In order to try to understand the impact of this type of computers t on engineering education in Chile, two experiences were conducted at two traditional Chilean universities. The obtained results show some similarity with those obtained in American Universities. However there are others that are a bit different. The lack of resources for every student opens new challenges for the adoption of these technologies in developing countries. The experiments have also shown that several benefits associated to the Tablet PCs adoption are more a consequence of others activities around this situation. At the moment it is not clear which benefits are the direct consequences of the Tablet PCs adoption, however the balance seems to be positive. More experimentation and scientific work is required to understand its impact of this emerging technology.

Acknowledgements This work was partially supported by Fondecyt (Chile), grant Nº 11060467 and also by Hewlett Packard through the Technology for Teaching Higher Education Grant Initiative.

38

I. Casas, S.F. Ochoa, and J. Puente

References 1. Alvarado, C., Anderson, R., Prey, J., Simon, B., Tront, J., Wolfman, S., Tablet, P.C.: Computing Curriculum. In: Summary of the Tablet PC and Computing Curriculum Workshop, Seattle, USA, August 4 (2004) 2. Anderson, R., Davis, P., Linnell, N., Prince, C., Razmov, V., Videon, F.: Classroom Presenter: Enhancing Interactive Education with Digital Ink. IEEE Computer 9(40), 56–61 (2007) 3. Berque, D., Bonebright, T., Dart, J., Koch, Z., O’Banion, S.: Using DyKnow Software to Support Group Work: A Mixed-Method Evaluation. In: Prey, Reed, Berque (eds.) The Impact of Tablet PCs and Pen-based Technology on Education, pp. 11–20. Purdue University Press (2007) 4. Davis, K.M., Kelly, M., Malani, R., Griswold, W.G., Simon, B.: Preliminary Evaluation of NoteBlogger: Public Note Taking in the Classroom. In: Prey, Reed, Berque (eds.) The Impact of Tablet PCs and Pen-based Technology on Education, pp. 33–42. Purdue University Press (2007) 5. Herrera, O., Ochoa, S., Neyem, A., Betti, M., Fuller, D., Aldunate, R.: Mobile Portfolio to Support Communities of Practice in Science Education. In: Proceedings of HCI International 2007, Beijing, China (July 2007) 6. Huettel, L.G., Forbes, J., Franzoni, L., Malkin, R., Nadeau, J., Nightingale, K., Ybarra, G.A.: Transcending the traditional: Using tablet PCs to enhance engineering and computer science instruction. In: Frontiers in education conference (FIE 2007) (2007) 7. Bartels Media GmbH. MaxiVista (Last visit, December 2008), http://www.maxivista.com/ 8. Nguyen, H., Bilen, S., Devon, R., Wise, J.: Adopting Tablet PCs in Design Education: Student Use of Tablet PCs and Lessons Learned. In: Richards, G. (ed.) Proceedings of World Conference on E-Learning in Corporate, Government, Healthcare, and Higher Education 2007, pp. 1172–1177. AACE, Chesapeake (2007) 9. Ochoa, S., Neyem, A., Bravo, G., Ormeño, E.: : MOCET: a MObile Collaborative Examination Tool. In: Proc. of HCII 2007, Beijing, China (July 2007) 10. Microsoft Corp. OneNote (2007), http://office.microsoft.com/es-es/ onenote/FX100487703082.aspx (Last visit December 15, 2008) 11. Prey, C., Reed, R.H., Berque, D.A. (eds.): The impact of Tablet PCs and Pen-based Technology on Education. Purdue University Press (2007) 12. Prey, C., Weaver, A.: Tablet PC Technology: The Next Generation. IEEE Computer 9(40), 32–33 (2007) 13. Price, E., Simon, B.: A Survey to Assess the Impact of Tablet PC-based Active Learning: Preliminary Report and Lessons Learned. In: Prey, Reed, Berque (eds.) The Impact of Tablet PCs and Pen-based Technology on Education, pp. 97–105. Purdue University Press (2007) 14. Sommerich, C., Collura, K.: Learning with Mobile Technology in high School: A HumanFactors Perspective. In: Prey, Reed, Berque (eds.) The Impact of Tablet PCs and Penbased Technology on Education, pp. 127–136. Purdue University Press (2007) 15. Toto, R., Lim, K., Wise, J.: Supporting Innovation: The Diffusion and Adoption of Tablet PCs in College of Engineering. In: Prey, Reed, Berque (eds.) The Impact of Tablet PCs and Pen-based Technology on Education, pp. 147–155. Purdue University Press (2007) 16. Tront, J.G.: Facilitating Pedagogical Practices through a Large-Scale Tablet PC Development. IEEE Computer 9(40), 62–68 (2007)

Optimal Affective Conditions for Subconscious Learning in a 3D Intelligent Tutoring System Pierre Chalfoun and Claude Frasson Département d’informatique et de recherche opérationnel, Université de Montréal, Montréal, Canada {chalfoun,frasson}@iro.umontreal.ca

Abstract. In this paper we take a closer and in-depth look at initial results obtained from a previous novel experiment conducted with a 3D subliminal teaching Intelligent Tutoring System. Subliminal priming is a technique used to project information to a learner outside of his perceptual field. Initial results showed great promise by illustrating the positive impact of the subliminal module on the overall emotional state of the learners as well as their learning performances. Indeed, since emotion monitoring is critical in any learning context, we monitored the physiological reactions of the user while they learned and while they answered questions. We present a detailed and precise look at the optimal affective conditions that set the best learners apart. We will also explain a most surprising finding: the positive long term impact of subliminal priming on the entire learning process. Keywords: optimal affective conditions, HCI, subconscious learning, 3D ITS.

1 Introduction In recent years, researchers in human-computer interfaces (HCI) as well as in various fields such as Intelligent Tutoring Systems (ITS) have taken advantage of adaptive and customizable HCI to record and analyze emotions [1]. This is not surprising since emotions, especially motivation and engagement, are widely related in various cognitive tasks [2]. Moreover, the importance of measuring emotions as well as consider them has become the focus of much growing research. The availability, ease of use and affordability of physiological devices helped in their integration into the tutoring systems. That data is then used to model the learner’s emotional and physiological profile in order to better adjust and adapt learning accordingly [3]. Learning in virtual worlds has taken a very important part in the HCI community for recent evidence has shown the relevance of using such virtual ITS for affective feedback and adaptation [3, 4]. Nevertheless, the current learning strategies have a limitation when it comes to processing complex information. Indeed, cognitive learning theories base mostly their intervention on attention to the specified task at hand. Complex information is broken down into pieces to gradually enable the learner to concentrate on one small part of the puzzle at a time. However, a large body of work in neuroscience and other fields lead us to believe that learning simple to complex information can be done without J.A. Jacko (Ed.): Human-Computer Interaction, Part IV, HCII 2009, LNCS 5613, pp. 39–48, 2009. © Springer-Verlag Berlin Heidelberg 2009

40

P. Chalfoun and C. Frasson

perception or complete awareness to the task at hand [5-8]. In fact, the existence of perceptual learning without perception has been neurologically proven and accepted [9]. Furthermore, recent work has put forth the performance increase in performance when using a subliminally teaching Intelligent Tutoring System [10]. Yet, subconscious learning systems are still widely absent in the HCI community. We intend to investigate in this paper the optimal emotional state of learners when using a subliminal teaching ITS by stating two research questions. First, in learning to solve a problem in a 3D virtual system, is there a significant emotional state in which the best learners are that sets them apart from the rest? Second, in answering questions following a learning session, what significant relationship can we establish between learners’ emotional state and subliminal projections? The organization of this paper is as follow: In the next section, we will present and discuss the previous work related to various aspects of our research. The following section describes the experiment setup and depicts the various aspects related to subliminal stimuli in a virtual 3D tutoring system. The obtained results will follow the experiment section leading to the last section where we conclude and present future work.

2 Related Work To the best of our knowledge, only a handful of papers in various fields have claimed the use of subliminal priming as a support for memory in the HCI community. The first and most referred to is Wallace’s text editor program [11]. In this experiment, Wallace and colleagues put forward two important findings: (1) the projected stimuli must take into account the specifications of the computer such as screen resolution and refresh rate (2) that the frequency at which subjects requested help was much lower when the requested information was projected subliminally. The Memory Glasses by [5] used wearable glasses that projects subliminal cues as a strategy for just-in time memory support. The objective was to investigate the effect of various subliminal cues (correct and misleading) on retention in a word-face learning paradigm and compare recall performance. Another use of priming for memory support can be found in the thesis of [12] where the author assesses the effects of brief subliminal priming on memory retention during an interference task. Finally, our most recent work showed the positive impact of subliminal stimuli on the learner’s performance [10]. Besides seeming to impact memory, subliminal priming can also have an emotional consequence on learners. Indeed, subliminal priming can have an emotional impact on the self-attribution of authorship of events [13]. Subjects were asked to compete against a computer in removing non words such as “gewxs” from a computer screen in the fastest time possible. However, after a determined amount of time, the computer would remove the word. Subliminal primes of self-associated words like “I” and “me” before an action increased the personal feeling that it was the participant that eliminated the non word and not the computer, thus increasing the feeling of selfauthorship of events. Furthermore, visual subliminal stimulus has been neurologically proven to have an impact in many physiological signals, namely the galvanic skin response (correlated to arousal) [14].

Optimal Affective Conditions for Subconscious Learning in a 3D ITS

41

Since we also use physiological sensors to monitor the emotional reactions of the learner, it would be relevant to sum some of the work related to using physiological sensors to record and analyze emotions that can occur in a learning environment. Physiological signals are generally correlated with emotions by associating specific signals, such as skin conductance and heart rate, to valance and/or arousal [15]. Indeed, the Empathic Companion is a good example where multiple physiological sensors, namely galvanic skin response (also referred to as skin conductance), heart rate and respiration were taken in real-time to analyze and adapt the tutor to the emotional reactions of the learner in a virtual 3D ITS [16]. Further research has analyzed a more detailed and relevant emotional significance of physiological signals, either in complex learning or gaming [17-19].

3 Experiment The current experiment uses precise and timed subliminal projections in a 3D intelligent tutoring system while monitoring the physiological reactions of the learner. In the same time we record the actions on the screen as well as the facial movements of the learners. Those visual recording are crucial to remove noise and identify events of special interest. Moreover, we constructed the subliminal cues in a way which would accelerate the learning process by triggering and enhancing an already possessed knowledge. 3.1 Design of the Experiment Indeed, the focus of the experiment is to visually teach, in a virtual 3D environment, the construction of an odd magic square of any order with the use of neither a calculator nor one mental arithmetic operation. A magic square of order n is a square containing n2 distinct integers disposed in a way such as all the n numbers contained in all rows, columns or diagonals sum to the same constant. The first part of Fig. 1. below depicts such a square. Magic Square

Trick #1

Trick #2

Trick #3

Fig. 1. Experiment design : Magic square and the three tricks taught

To construct the following square, one must successively apply three simple tricks. These tricks are illustrated in Fig. 1 and labelled trick 1 to 3 respectively. We decided to show the learners multiple examples of each trick without explaining how the trick works. As an example, the first trick to construct any magic square is to place the following number one square above and two squares to the right of the previous one (exactly like a knight’s move in chess). If we look at the second picture of Fig. 1, we notice that number 15 is placed one square above and two squares to the right of

42

P. Chalfoun and C. Frasson

number 14. The same logic applies to numbers 1 and 2, 4 and 5 and so forth. Instead of giving away the answer to the first trick, we ask the subjects to deduce the rule by themselves. This is where the subliminal stimulus comes into play. We will have two groups, one group will take part of the experiment without subliminal stimuli (control group) and the tutor will subliminally send the answer to the other group. We will then compare performances, trick completion time, question completion time as well as physiological signal variations. The teaching material is separated into parts, or PowerPoint-like slides, and displayed at a slow rate to give every learner an equal chance at fully reading each “slide”. The subliminal stimuli and threshold were carefully chosen following the neural bases of subliminal priming [9]. Each stimulus was preceded by a 271 ms pre-mask of random geometrical figures, a 29 ms prime and a 271 post-mask of random geometrical figures. The subliminal stimuli that will be presented to one of the two groups will be displayed at significant places before and after specific slides. The experiment intends to “boost” learning by priming the answer before showing the corresponding slide. Fig. 2. shows a diagram of the way subliminal priming will take place between slide 1 and slide 2 when learning to deduce the inner working of the first trick.

271 ms

mask

29 ms prime 271 ms Slide 1

mask

Priming between slides

Slide 2

Fig. 2. Subliminal priming of the solution between 2 slides

The learners were instructed to answer a series of two to three questions following each learned trick to test their knowledge. The learners were instructed to finish the experiment as quickly and efficiently as possible. No time limit was imposed. A base line for the physiological signals preceded all monitored activities. A questionnaire preceded the experiment aiming at collecting demographical data as well as the gaming experience of the subjects. Another series of questions were asked at the end of the experiment to evaluate the learner’s appreciation and more importantly their overall appreciation of the system. 3.2 The 3D Virtual Environment Learning takes place in a game-like environment called MOCAS [20] as show in Fig. 3. The experiment has three rooms like the one illustrated on Fig. 3. Each room teaches one trick. MOCAS takes place in full-screen mode for a better immersion and less window distracting events. Furthermore, the system clock is hidden so users don’t get distracted by continuously monitoring the time they have spent on each lesson. The interactions between the avatar’s learner and the pedagogical agents are done via mouse clicks. This interaction is important because learners have a time

Optimal Affective Conditions for Subconscious Learning in a 3D ITS

43

window of 30 seconds to answer each question. If they feel that time was not enough, they can simply re-click on the agent and the question restarts. This re-click factor was important in distinguishing good from bad learners. The learners are instructed to continue once they are convinced they have discovered the inner working of each trick. They are then asked to answer a series of questions (two to three) by another set of visually different pedagogical avatars. Each question is related to the last trick learned. The agent asks the user to correctly place a number in a magic square. The learner responds by choosing the path that correctly answers the question. Physiological signals of the learners were also monitored in real-time and saved for further analysis. The used signals were heart rate, galvanic skin response, respiration rate and skin temperature. The signals are managed by the ProComp Infinity encoder [21].

Path leading to the Question Agents

Avatar of a pedagogical agent

Clicking on the Pedagogical agent starts the lesson.

Teaching material

Avatar of the learner

Fig. 3. 3D virtual learning environment

3.3 Learners Tested A total of 31 healthy volunteers, 16 men and 15 women, took part of the experiment. One participant had to be removed because of a major recording issue in the videosignal synchronisation module. The sample’s mean age was 26 (SD = 4.51). Only two volunteers had extensive video gaming experience. All the others gaming experience ranged equally anywhere from weak to moderate high. A repartition of the learners can be found in table 1. Table 1. Participants’ distribution

Group A : no subliminal stimuli Group B : primed with subliminal stimuli Total

Men 8 7 15

Women 7 8 15 30

44

P. Chalfoun and C. Frasson

4 Results The first aspect we wanted to examine was the existence, if any, of a relationship between affective variations and subliminal projections while learning the tricks. Fig. 4 shows the average quantitative affective variations of learners with regards to valence and arousal when learning all three tricks with and without the subliminal module. The signal used and correlated with valence is the heart’s inter-beat interval (IBI) and galvanic skin response was used and correlated with arousal (GSR) [15]. These signal values are normalized by mean-shifting, that is subtracting each signal’s value from the signal’s baseline mean then dividing the result by the signal’s standard deviation. This widely accepted technique enables us to compare learners’ results for it solves the problem of extra-individual physiological differences. Fig. 4 shows the average affective values for a period of 4 second following every subliminal stimulus. The full brown bars represent the average value of the signal for all subliminal stimuli at the precise moment the stimulus was projected (t=0s, s is for seconds). The horizontal dashed bars represent the same averaged value except for it’s computed for the 4 seconds following that projected stimulus (T=t + 4s). Since group A was not primed with subliminal stimuli, we placed markers for each learner at the precise moment where subliminal cues would have been projected if these learners would have been taking the course with the subliminal module. For example, the first four numbers from the left (-0.7, -0.7, 0.3 and 0.6) represent the following situation: on average, all learners in the experiment have had a normalized valence change of zero (-0.7 at moment t=0 and -0.7 after 4 seconds) when learning without the subliminal module compared to a normalized valence increase of +0.3 (0.3 at moment t=0 and 0.6 after 4 seconds) when learning with the subliminal module. Average Quantitative Affective Variations Of Learners While Learning The Tricks Group A

Normalized Mean-Shifted Values With Regards To Baseline

8.0

6.8 6.7

7.0

Group B

Group A

6.0 5.0

4.4 4.4

5.5

Group B

4.0

Group Group B A

3.6 3.2 2.9

3.0

Group B

2.0 1.0

Group A 0.3

Group B

1.7

Group A

0.6

Group B

2.5 2.5

1.0

0.8

Group A

1.1 0.5

0.4

0.0 Valence

-1.0

-0.7 -0.7

Arousal

Valence

All Learners

Arousal Best Learners

-0.9 -0.9

Valence -0.6 -0.6

Arousal Worst Learners

All 3 Tricks

-2.0 Valence And Arousal For All Tricks Broken Down By Best, All and Worst Learners

Without Subliminal Module - Average Value At t=0s

Without Subliminal Module - Average Value At T=t+4s

Fig. 4. Quantitative affective variations when subjected to subliminal stimuli while learning

Optimal Affective Conditions for Subconscious Learning in a 3D ITS

45

The results shown in Fig. 4 are not only statistically significant (p