CULTURAL ERGONOMICS ADVANCES IN HUMAN PERFORMANCE AND COGNITIVE ENGINEERING RESEARCH Series Editor: Eduardo Salas

Associate Editors:

CLINT A. BOWERS Department of Psychology, University of Central Florida, USA

NANCY COOKE Department of Psychology, New Mexico State University, USA

JAMES E. DRISKELL Florida Maxima, USA

ANDERS ERICSSON Florida State University, USA

DIANNA STONE Department of Management, University of Central Florida, USA

Volume 1: Advances in Human Performance and Cognitive Engineering Research, edited by Eduardo Salas Volume 2: Advances in Human Performance and Cognitive Engineering Research: Automation, edited by Eduardo Salas Volume 3: Advances in Human Performance and Cognitive Engineering Research, edited by Dianna Stone ADVANCES IN HUMAN PERFORMANCE AND COGNITIVE ENGINEERING RESEARCH VOLUME 4

CULTURAL ERGONOMICS

EDITED BY

MICHAEL KAPLAN Southwestern University, Texas, USA

2004

Amsterdam – Boston – Heidelberg – London – New York – Oxford – Paris San Diego – San Francisco – Singapore – Sydney – Tokyo ELSEVIER B.V. ELSEVIER Inc. ELSEVIER Ltd ELSEVIER Ltd Sara Burgerhartstraat 25 525 B Street, Suite 1900 The Boulevard, Langford 84 Theobalds Road P.O. Box 211, San Diego, Lane Kidlington, London 1000 AE Amsterdam CA 92101-4495 Oxford OX5 1GB WC1X 8RR The Netherlands USA UK UK © 2004 Elsevier Ltd. All rights reserved. This work is protected under copyright by Elsevier Ltd, and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier’s Rights Department in Oxford, UK; phone: (+44) 1865 843830, fax: (+44) 1865 853333, e-mail: [email protected]. Requests may also be completed on-line via the Elsevier homepage (http://www.elsevier.com/locate/permissions). In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (+1) (978) 7508400, fax: (+1) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (+44) 20 7631 5555; fax: (+44) 20 7631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of the Publisher is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier’s Rights Department, at the fax and e-mail addresses noted above. Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made. First edition 2004 Library of Congress Cataloging in Publication Data A catalog record is available from the Library of Congress. British Library Cataloguing in Publication Data A catalogue record is available from the British Library. ISBN: 0-7623-1049-9 ISSN: 1479-3601 (Series)
∞ The paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Printed in The Netherlands. CONTENTS

LIST OF CONTRIBUTORS vii

ACKNOWLEDGMENTS ix

INTRODUCTION: ADDING A CULTURAL DIMENSION TO HUMAN FACTORS Michael Kaplan xi

1. NATIONAL AND CULTURAL VARIABLES IN ERGONOMICS Alphonse Chapanis 1

2. CULTURE, CONTEXT, AND PERFORMANCE Neville Moray 31

3. THE DEVELOPMENT OF CULTURE-ORIENTED HUMAN MACHINE SYSTEMS: SPECIFICATION, ANALYSIS, AND INTEGRATION OF RELEVANT INTERCULTURAL VARIABLES Kerstin Röse 61

4. HUMAN FACTORS ENGINEERING (HFE) AND CULTURAL CALIBRATION FOR VESSEL AND OFFSHORE INSTALLATION DESIGN Denise B. McCafferty, E. Johan Hendrikse, and Gerry E. Miller 105

5. CROSS-CULTURAL FACTORS IN AVIATION SAFETY Ashleigh Merritt and Daniel Maurino 147

v vi

6. THE ROLE OF NATIONAL CULTURE IN ENHANCING TRAINING EFFECTIVENESS: A FRAMEWORK Shreya Sarkar-Barney 183

7. TOWARDS AN ANTHROPOTECHNOLOGY. X. A NEW ACTIVITY FOR THE UNITED NATIONS IN THE SERVICE OF ECONOMIC DEVELOPMENT: SPECIFYING REQUIREMENTS FOR TECHNOLOGY TRANSFERS IN GIVEN GEOGRAPHICAL AND ANTHROPOLOGICAL LOCATIONS [Translated from the French by Kendra Francisco, Southwestern University] Alain Wisner 215

8. ERGONOMICS IN INDUSTRIALLY DEVELOPING COUNTRIES (IDCs): SOCIO-CULTURAL PERSPECTIVES P. A. Scott and J. Charteris 223

9. COGNITION IN NATURAL SETTINGS: THE CULTURAL LENS MODEL Helen Altman Klein 249

10. DESIGNING THE USER EXPERIENCE FOR INTERNATIONAL WEB USERS Monica Chong 281

11. ‘A FRENCHMAN, A GERMAN, AND AN ENGLISHMAN ...’: THE IMPACT OF CULTURAL HETEROGENEITY ON TEAMS Florian Jentsch, Raegen M. Hoeft, Stephen M. Fiore, and Clint A. Bowers 317

12. CHALLENGES AND APPROACHES TO UNDERSTANDING LEADERSHIP EFFICACY IN MULTI-CULTURAL TEAMS Eduardo Salas, C. Shawn Burke, Jennifer E. Fowlkes, and Katherine A. Wilson 341 LIST OF CONTRIBUTORS

Clint A. Bowers University of Central Florida, USA

C. Shawn Burke University of Central Florida, USA

Alphonse Chapanis The Johns Hopkins University, USA

Jack Charteris Rhodes University, South Africa

Monica Chong Yahoo! Inc., United Kingdom

Stephen M. Fiore University of Central Florida, USA

Jennifer E. Fowlkes University of Central Florida, USA

E. Johan Hendrikse Paragon Engineering Services, Inc., USA

Raegen M. Hoeft University of Central Florida, USA

Florian Jentsch University of Central Florida, USA

Michael Kaplan Southwestern University, USA

Helen Altman Klein Wright State University, USA

Daniel Maurino International Civil Aviation Organization, Canada

Denise B. McCafferty American Bureau of Shipping, USA

Ashleigh Merritt University of Texas at Austin, USA

Gerry E. Miller Paragon Engineering Services, USA

vii viii

Neville Moray ergoNM, Magagnosc, France

Kerstin R¨ose University of Kaiserslautern, Germany

Eduardo Salas University of Central Florida, USA

Shreya Sarkar-Barney Illinois Institute of Technology, USA

Pat A. Scott Rhodes University, South Africa

Katherine A. Wilson University of Central Florida, USA

Alain Wisner Conservatoire National des Arts et Metiers, France ACKNOWLEDGMENTS

My warm thanks to Dean James Hunt, Provost, and Professor Jacqueline Muir-Broaddus, Chair of the Psychology Department, for making a home at Southwestern University, Georgetown, Texas, for cultural ergonomics and the International Center of Cultural Ergonomics, and for facilitating preparation of this book. Southwestern students Kendra Francisco, Staci Benson, and Ellen Gass contributed helpful assistance. At Elsevier, Fiona Barron, Publishing Editor, has been extraordinarily helpful, and the consideration and support there from Becky Lewsey and Deborah Raven have been particularly noteworthy. Dr. Pierre Falzon, Conservatoire des Arts et Metiers in Paris, made possible the acquisition of documents written by Professor Alain Wisner, who died recently. Computer advice and assistance provided by Richard H. Troxell have been invaluable. Communication and interchange of documents and information with Dr. Eduardo Salas at the University of Central Florida were facilitated by Marcella Maresco and Diana Furman. Dean Ben B. Morgan, Jr. at the University of Central Florida and Professors Margaret Thomas and Jack McGuire, successive Chairs of the Psychology Depart- ment there, enabled initial establishment of the International Center of Cultural Ergonomics. Original encouragement for the concept of cultural ergonomics and the center came from Captain Paul R. Chatelier, USN (Retired), Dr. Earl Alluisi, and Professors John A. Wise and Anthony Debons. Early on, Professors Wise and Debons provided generous support. I am also especially indebted to Professor Houshang Shahnavaz, formerly of Lulea University in Sweden, for his generosity, unfailing helpfulness, loyal friendship and kindness. For special interest and useful suggestions, I also thank Captain Daniel Maurino, International Civil Aviation Organization, Professors Alain Wisner, Hal W. Hendrick and Najmedin Meshkati, and Drs. Stuart O. Parsons, Donald Weitzman, and Harold Van Cott. To all of them, my former assistants Janet Poole, Adam Payne, Shreya Sarkar-Barney, Jessie Y. C. Chen, and Jason Kring, and especially my wife, Eugenie, my most helpful and loyal supporter, I send deepest thanks and every good wish.

Michael Kaplan Georgetown, Texas 1 March 2004 Editor ix This Page Intentionally Left Blank INTRODUCTION: ADDING A CULTURAL DIMENSION TO HUMAN FACTORS

Michael Kaplan

ABSTRACT

Briefly highlighting aspects of the evolution of ergonomics, this introduction to the present volume emphasizes the breadth of the field and acquaints the reader with the emergence of cultural ergonomics as a new sub-discipline. It interweaves in the account comments by the editor on the book’s twelve chapters and some notes relating to the new addition.

The essential fact of culture as we live it and experience it, as we can observe it scientifically, is the organization of human beings into ...groups. Bronislaw Malinowski (1944) ...ergonomics has an anthropological dimension, both in terms of its methodology and its fields of application. Alain Wisner (1995) ...to internationalize ergonomics ...requires that we approach our research, education and applications with broader perspectives [editor’s italics], perspectives so broad that they are capable of encompassing the whole world and all the peoples on it Alphonse Chapanis (1974)

Scientists working in the domain of human factors and ergonomics (the terms are used interchangeably) more often than not devote themselves to relatively cir- cumscribed specialties and, appropriately, are widely recognized for worthwhile contributions to them. Not infrequently, observers tend to identify the entire field of ergonomics itself primarily with the comparatively limited activities of one or more of these scientists and their colleagues who work in the same sub-disciplinary area. But this view fails to take account of the breadth and complexity of the field. Indeed, it was asserted at the Global Ergonomics Conference in Cape Town (Kaplan, 1998) that “(T)he world of ergonomics is THE WORLD,” although one could add “and beyond that, sites in outer space.” The point is that, if one stands back to gain some perspective on the field, one can see extensive xi xii MICHAEL KAPLAN examples (1) requiring ergonomic attention in a vast array of geographic locations, and (2) involving enormous varieties of human work – both simple and complex. Such breadth has not always been the case. However, from sparse beginnings, ergonomics has grown and continues to do so. Although its recognition as a discipline was a long time in coming, we occasionally find discrete instances of what we would label as ergonomics concerns and activities throughout history – and even pre-history (see, for example, the development of tools described by Childe, 1944). A rather unique example has been recounted by Metz (1987): in the City Hall of Strasbourg, France, records are to be found describing the construction of its great cathedral, including – in the area of performance-time functions – details of schedules and hours of work of its laborers. The road to recognition may have begun in 1857. Karwowski (1991) tells us that Wojciech Jastrzebowski, writing in a Polish newspaper that year, was the first to use the term ergonomics. And Hywell Murrell, in the 1950s, introduced the term in Britain – referring to “the natural laws of work” – and gave impetus to much research and application in the field (see Murrell, 1965). Ergonomics, then, is a continuously evolving field. We owe to Martin Helander (1997) a fine discussion of its development, which, among other things, calls our attention to some of its major specializations and topics, especially key focal areas characterized from the 1950s by Brian Shackel (and reported by Hendrick, 1993). Figure 1 is an adaptation of that characterization, with an extension to the 2000s which contains predictions by Helander on global communication and this writer on cultural ergonomics. Notwithstanding the importance and necessity of the areas and sub-disciplines represented in Fig. 1 and elsewhere in the realm of human factors, it is time for that “broader perspective” of which Professor Chapanis spoke (1974). Many of its elements are apparent in the influences and interactions of cultural variables relating to instances of human performance and human interfacing, particularly those which we find in work environments world-wide. So it is, that in recognition of a need for research, systematization, and application in the realm of cultural variables, there has been added to existing sub-disciplines of human factors a new focus – a cultural dimension, extending into the 2000s and beyond. We may speak of this sub-discipline as cultural ergonomics. Figure 2 presents some examples of culture and cultural milieux that affect work and human interfacing. It has been argued that relevant milieux should always be specified and their effects considered in all cases of ergonomic tasking (Kaplan, 1998). In the present volume, we explore aspects of cultural ergonomics through the eyes of distinguished scientists who are identified with some of its problem areas and have contributed unique insights and useful approaches and findings. Introduction xiii

Fig. 1. An Historical View of Developments in Ergonomics.

Fig. 2. Examples of Culture and Cultural Settings. xiv MICHAEL KAPLAN

The lead chapter comes to us from the innovative thinking of the late Professor Alphonse Chapanis (1974), who gave us a comprehensive treatment of ergonomic problems and variables as he found them throughout the world when he wrote. His account applies today, although it is appropriate to add new topics and revisions of earlier material generated by developments in today’s world. We may say that Professor Chapanis played a key role in starting us on the road that led to cultural ergonomics, and it is fitting to treat his chapter as an introduction to that sub- discipline and the papers that comprise this volume. Helping us to better appreciate and encompass some of the world’s ergonomic problems is the chapter by Scott and Charteris, which tells of relevant issues in industrially developing countries. In a dramatic account, they argue for the importance of confronting the needs of our fellow human beings in those countries through their eyes and circumstances rather than the sophisticated approaches of industrially advanced societies. A related set of concerns was dealt with by Professor Wisner, who was seeking ways to improve economic well-being in industrially underdeveloped countries. Turning to the potential of technology transfer, he proposed that agencies of the United Nations be brought into play to facilitate the transfer process. That process, he argued, should take account of anthropological and geographical considerations occurring in the recipient and source countries. In numerous other writings and industrial research, he pressed for the importance of culture and, in so doing, developed the approach that he labeled anthropotechnology. With his vigorous and extensive efforts relating anthropological data to ergonomic issues, he, like Professor Chapanis, played a substantial role in leading us to develop cultural ergonomics. Work issues arising from organizational groupings – see Fig. 2 – can be serious. In the United States, for example, it was recently argued that “flawed NASA culture” contributed to the Columbia shuttle disaster (Sanger, 2003). The role of culture in the performance of groups and teams is a topic calling for more analysis and research. We see some of that in the chapter by Jentsch, Hoeft, Fiore, and Bowers, who provide a thorough examination of the influence of cultural heterogeneity on teams and find that the answer is not simple. Salas, Burke, Fowlkes, and Wilson take us into new waters with a particularly detailed chapter that seeks to understand how leadership efficacy in multicultural teams can be attained and improve their functioning. It is a daunting task, and their work sets the stage for new research that is likely to prove useful. Another aspect of organizational grouping problems arises in the increasingly frequent multinational collaborations in varied settings that are with us today (e.g. military, business, aviation, government). Cognitive differences – largely culturally determined – on the part of team leaders (cf. Salas et al. above) and members, as in Introduction xv the case of perceptions of others, judgments, and decision making, can influence action and, in some instances, lead to deleterious results. In her chapter, Klein discusses these and numerous related issues in informative detail. She then presents and discusses the Cultural Lens Model, a novel and promising approach for dealing with difference problems. In brief, it enables individuals to “see” the world as others do and, as she put it, “avoid the barriers and problems introduced by national differences in cognition” and utilize “the skills and resources of people around the world.” The performance of human-machine systems is closely aligned with our more traditional views of ergonomics, with Moray pointing out that anatomy, physi- ology, and cognitive and behavioral limits are seen as constraining factors. He asserts, however, that context and culture are equally important in this regard, and throughout his chapter develops this theme in an incisive manner, offering cogent examples. In her chapter, Rose¨ provides a further development. She gives us a kind of “marriage” of cultural variables and systems engineering principles, providing an exceptionally thorough, extensive, and detailed step-by-step treatment for combining such variables in systems designed for culturally-oriented product markets. A feature of cultural ergonomics is the interaction of cultural variables and traditional areas of human factors engineering. This is portrayed in Fig. 3. We now turn to some of those areas. Regarding training for the international arena, Sarkar-Barney has raised a ques- tion that has not been previously considered. When a training system is developed in a specific cultural context, what procedures should be used so that it may be deployed globally? Her answer builds on a framework by Baldwin and Ford (1988) to transfer skills from one cultural setting to another. When we consider cultural influences on human-computer interaction, we are confronted with many issues. One of these deals with the world-wide web and the internet. Chong has focused on the process of designing Web sites that can be successfully used both by international and local audiences. Her chapter discusses in detail the ramifications and requirements of such an effort and argues for a user-centered approach. Safety is a major concern of ergonomics, and the role of culture affecting it demands considerable attention. Hence, the importance of the chapter by Merritt and Maurino on aviation safety. They see the interfaces between one culture and another as the source of serious aviation safety issues, and they recommend the use of cultural mediators as an approach likely to be effective. What they call “cultural calibration” is discussed by McCafferty, Hendrikse, and Miller in considering aspects of marine safety. Their chapter examines the joint role of human factors engineering and culture in designing vessels and offshore xvi MICHAEL KAPLAN

Fig. 3. Examples of Traditional Human Factors Areas Influenced by Cultural Variables. installations. They place special emphasis on how culture should be addressed while identifying “user” needs and requirements. A full account of cultural ergonomics does not end with the topics presented herein. Material comprising this sub-discipline would fill additional volumes. Among such items, of particular concern is the issue of culture and nuclear safety. In this area, there is an instructive paper by Najmedin Meshkati (1997) that examines the influence of local culture on the safety culture in nuclear power plants. Other topics worth considering include: aerospace safety, standardization, communication issues, architectural design, information science, manufacturing, agriculture, military activities, medicine, teaching approaches, disability prob- lems, aging. Still other listings may be found in Kaplan (1995, see Appendix) and Kaplan, Morgan and Kring (2001). Clearly, there is much work to be done.

REFERENCES

Baldwin, T. T., & Ford, K. K. (1988). Transfer of training: A review and direction for future research. Personnel Psychology, 41, 63–105. Introduction xvii

Chapanis, A. (1974). National and cultural variables in ergonomics. Ergonomics, 17, 153–175. Childe, V. G. (1944). The story of tools. London: Cobbot. Helander, M. G. (1997). Forty years of IEA: Some reflections on the evolution of ergonomics. Ergonomics, 40, 952–961. Hendrick, H. W. (1993). The IEA and international ergonomics: Past, present and future. In: Proceedings of the IEA/Russian Ergonomics Society Conference – Ergonomics in Russia, the Other Independent States and Around the World – Past, Present and Future (Russian Ergonomics Society, St. Petersburg). Kaplan, M. (1995). The culture at work: Cultural ergonomics. Ergonomics, 38, 606–615. Kaplan, M. (1998). The cultural milieux of ergonomics, Invited address, Global Ergonomics Conference, University of Cape Town, Cape Town, South Africa. Kaplan, M., Morgan, B. B., Jr., & Kring, J. P. (2001). Cultural ergonomics. In: International Encyclopaedia of Human Factors and Ergonomics. London: Taylor & Francis. Karwowski, W. (1991). Complexity, fuzziness, and ergonomic incompatibility issues in the control of dynamic work environments. Ergonomics, 34, 671–686. Malinowski, B. (1944). A scientific theory of culture and other essays. Chapel Hill: University of North Carolina Press. Meshkati, N. (1997). Cultural context of the safety culture: A conceptual model and experimental study. In: M. Mouloua & J. M. Koonce (Eds), Human-Automation Interaction: Research and Practice (pp. 286–297). Mahwah, NJ: Erlbaum. Metz, B. (1987). Personal communication. Murrell, K. F. H. (1965). Ergonomics: Man in his working environment. London: Chapman & Hall. Sanger, D. E. (August 27, 2003). Flawed NASA culture blamed in tragedy. Austin American-Statesman. Austin Texas. Wisner, A. (1995). Situated cognition and action: Implications for ergonomic work analysis and anthropotechnology. Ergonomics, 38, 1542–1557. This Page Intentionally Left Blank 1. NATIONAL AND CULTURAL VARIABLES IN ERGONOMICS

Alphonse Chapanis

ABSTRACT

Ergonomics has so far been largely an American and Western European discipline. For that reason, present-day ergonomics is not robust enough to cope with the many large and important differences that one finds among the people of the earth. These differences are discussed under five major headings: anthropometry, language, physiology, psychology, and customs and practices. In each case, the differences are so great that they require substantial modifications in commonly accepted ergonomics principles and practices. In some cases, the technical design problems involved in adapting equipment to non-Western societies have required the develop- ment of new ergonomics principles. Failure to take account of national and cultural variables may nullify the gains that one might reasonably expect to follow from the application of ergonomics in many areas of the world.

EDITOR’S NOTE

This paper was an invited address delivered at the Fifth International Congress on Ergonomics, June 8, 1973, Amsterdam, The Netherlands. It originally appeared in the journal ERGONOMICS, 1974, Volume 17, Number 2, pages 153–175, and is reprinted here by permission of the publisher, Taylor and Francis.

Cultural Ergonomics Advances in Human Performance and Cognitive Engineering Research, Volume 4, 1–29 © 2004 Published by Elsevier Ltd. ISSN: 1479-3601/doi:10.1016/S1479-3601(03)04001-3 1 2 ALPHONSE CHAPANIS

With the passing of Professor Chapanis on October 3, 2002, the world lost one of its most distinguished workers in the field of human factors and ergonomics. He was brilliant, caring, innovative, and a pioneer. Indeed, this paper was a pioneering work. It may be regarded as having set the stage for the field of cultural ergonomics, and while it originally appeared in 1974, it is as if it burst upon the scene today. For in the hiatus between then and now, the field has only recently begun to advance.

INTRODUCTION

How international is ergonomics? When I first thought about this question years ago, my initial reaction was that science and technology are universal and that they know no geographical boundaries. Over the years, however, I have come to realize that such an idealistic view does not quite agree with the realities of life. Ergonomics deals with people, and people around the globe differ greatly and in many important ways. So from my present vantage point, I conclude that the answer to my question is much more complex than I had at first imagined. Indeed, I now think that one of the most challenging problems we face today is how we can make ergonomics robust enough to cope with national and cultural variables. What I want to do in my paper is to give you some feel for the importance and the dimensions of that problem.

THE DATA BASE

Everyone here knows that a substantial number of ergonomists, perhaps as many as half of all of them, do not call themselves ergonomists. They call themselves human engineers, or human factors engineers. Although ergonomics and human factors engineering differ in some subtle ways (Chapanis, 1970), the differences are more apparent than real. For all practical purposes, both human factors engineering and ergonomics have common aims, methodologies, and principles. I emphasize the essential identity between the two fields, because whatever I say about either one of the disciplines applies equally well to the other.

Present-day Ergonomics is an American and Western European Discipline

Although I do not have any quantitative data to support my contention, I doubt whether anyone would dispute that up to the present time, ergonomics has been largely an American and Western European discipline. Until a few years ago, National and Cultural Variables in Ergonomics 3 almost all the research in the field had been done in America and Western Europe. As a result, our findings apply to large-boned people, people who were born to or use the English language, and people who have Western customs, habits and ways of life. That is a pretty small percentage of the world’s population. In the meantime, there have been some scattered reports, mostly in the anthropological literature, about problems that had been encountered when Western technology was introduced into the less developed countries of the world. In addition, various agencies of the United Nations have long been concerned about the problems of industrialization in under-developed countries. Indeed, the eminent anthropologist Margaret Mead edited a manual on some of these problems as long ago as 1953 (Mead, 1953). Still, the full implications of the wide range of national and cultural differences that we find around the world have not permeated into science and technology. In particular, the implications of these differences for ergonomics are not yet fully appreciated.

THE INTERNATIONALIZATION OF ERGONOMICS

This is a good time for all of us to be concerned about the internationalization of ergonomics. Let me tell you some reasons why I think this is the case.

The Growth of Ergonomics Around the World

Although America and Britain may still dominate the field, human factors engineering has at least outgrown their borders. The International Ergonomics Association now contains over three thousand members distributed throughout 30 countries of the world. It contains 12 member societies and if we exclude the American Human Factors Society and the British Ergonomics Research Society, the remaining 10 contain more than 1300 members. These figures remind us that we have indeed become world-wide. The more important implication, however, is that if our members are scattered over the wide researches of this globe, then our technology should be broad enough to provide the principles and practices that can be applied by our colleagues no matter where they may be.

The Development of International Standards

Not only has the IEA been growing, but it has also begun to take a more active role in spreading ergonomics throughout the world. One significant step 4 ALPHONSE CHAPANIS in that direction was the organization under the sponsorship of the IEA and the International Standards Organization, of an international symposium on ergonomics and standards that took place earlier this year in England. This meeting brought together 33 human factors and ergonomics experts from 11 countries to explore the problem involved in the preparation of international ergonomics standards for the design of equipment, work, working conditions and consumer items. At that symposium, however, I argued that international standards cannot really be developed until we understand the influence of national and cultural variables on the principles and practices of ergonomics. As more and more findings about national and cultural variables become available, we will be able to formulate better international standards to benefit all of us in our work.

Our World Has Shrunk

A third reason why I think it is timely for us to consider the internationalization of ergonomics is that our world has shrunk greatly in the past few centuries. We are all much closer together now than we have ever been. Let me remind you of a couple of milestones in the history of our civilization. When Ferdinand Magellan began the first recorded circumnavigation of our globe, he left the port of Salcar de Barrameda, Spain, on September 20, 1519. Magellan himself never completed the trip, but his ship, the Victoria, returned to Spain on September 8, 1522, after a voyage of almost exactly three years. Four hundred years later, the first aerial circumnavigation of the globe took place in 1924 when three U.S. Army planes took off from Seattle, Washington, on April 6, and returned to the same airfield on September 28. That was a trip of almost six months. Today, less than 50 years later, a commercial airliner travels around the earth in approximately 67 hours and that time includes all its normal stops at a dozen or so cities. Of course, if you are an astronaut you can make the same trip in about 90 minutes! However impressive and symbolic such comparisons may be, the consequences are even more impressive. Today, Americans buy automobiles, typewriters, tape- recorders and calculating machines made in France, Germany, Britain and Japan. We, in turn, sell our computers, aircraft and farm machinery to Europe, Africa, South America, Western Europe, and perhaps soon to the Peoples Republic of China. Similarly, the European common market has resulted in a free interchange, not only of products, but of workers among the several countries belonging to that confederation. These are only a few examples of the internationalization of business and commerce that has become an important fact of modern life. National and Cultural Variables in Ergonomics 5

Modern world trade has not been without its problems. The Stanford Research Institute (1970) recently compiled an impressive collection of multi-million dollar failures that could be traced to ignorance of behavioral factors in the international marketing of products. Incidentally, it gives me some small comfort to add that the failures were not all made by American manufacturers. The list involved products as diverse as foods, automobiles, home appliances, beauty products, recreational equipment and home furniture. Learning lessons this way is costly, and manufacturers the world over are eager to have whatever information we can give them about national and cultural variables in the design and marketing of their products. Another reason for internationalizing ergonomics, therefore, is that in so doing, we will produce and supply some of the information that is needed for more effective world trade.

The Need for Technology in the Under-Developed Areas of the World

The final reason why I think it is timely to consider the internationalization of ergonomics is that we must respond to the needs of under-developed countries. In 1965, the developed countries of the world contained a little more than 21% of the population but nearly 3/4ths of the gross national products of the world. These figures are certainly going to change. For years now, the developing countries have had a much higher average annual rate of growth in manufacturing, rail traffic and electric energy consumption than have the industrialized countries (Commission on International Development, 1969). As a result, United Nations sources predict that by 1980, the percentage of the world’s gross national products contributed by the developed countries will have decreased to 65%. There is a reason to expect that this trend will continue in the decades that follow. Although under-developed countries greatly need the benefits of modern technology, the fact is that they themselves cannot fill that need. They have neither the trained manpower nor the knowledge to do the job themselves. They have to depend on the products, the methods, and the technical skills of their more advanced neighbours. And that, of course, is where we ergonomists come into the picture again. We have an unusual opportunity to introduce modern technologies, agricultural methods, manufacturing methods, systems of communication, systems of transport and housing in places where some of these things have never existed heretofore. Here is an opportunity for us to do a first-rate job of fitting modern technology to man and to do it properly from the start. How well we do that will depend to a large extent on the viability and the universality of our principles. 6 ALPHONSE CHAPANIS

A Symposium on National and Cultural Variables

Considerations such as those I have enumerated above led me in 1970 to propose to the Advisory Group on Human Factors of the North Atlantic Treaty Organization that they support a symposium on this topic. NATO agreed and with Dr. John de Jong, I organized such a conference. The principal aim of our symposium was to try to find out the extent to which the principles and practices of ergonomics have to be modified to take account of national and cultural variables. Basic studies of national and cultural differences were deliberately excluded unless they could be shown to have an unmistakable connection with ergonomics. The symposium was held in Oosterbeek, The Netherlands, from 19 to 23 June, 1972, and was attended by 44 persons from 15 countries. The 26 contributed papers revealed a richness of data and findings that far exceeded my original expectations. Although this article draws heavily on those contributions, it is impossible for me to give you more than the barest glimpse of their highlights. I am editing the papers presented at the symposium for publication by the Johns Hopkins Press (Chapanis, 1974). That publication will reveal the full extent of the problem to which I can only allude here.

ANTHROPOMETRIC DIFFERENCES

You do not have to be a scientist or an ergonomist to observe that the peoples of the world come in vastly different sizes. This is probably the most obvious and impressive variable that one encounters in trying to apply ergonomics around the world. Still, I do not think that the full extent of these anthropometric differences are understood by ergonomists.

Differences in Stature

In our textbooks and design guides, it is customary to say that equipment should fit approximately the middle 90% of a user population. To accommodate the statures of the middle 90% of an American population, we have to use a design range of about 20.4 cm, that is from 167.3 to 187.7 cm. That range is shown on the left in Fig. 1. On the same chart are shown relevant statistics for anthropometric measures made on populations in Germany, France, Italy, Japan, Thailand and Vietnam. These populations are arranged in order of decreasing means. You can see from the figure that the American design standard fits very nearly 90% of Germans, but only about 80% of Frenchmen, 65% of Italians, 45% of Japanese, 25% of Thai, and 10% of the largest Vietnamese. Equipment built to American standards is hopelessly over-sized for large segments of the world’s population. National and Cultural Variables in Ergonomics 7

Fig. 1. Statistics on the Statures of American, German, French, Italian, Japanese, Thai, and Vietnamese Males. (After Kennedy, 1974.) 8 ALPHONSE CHAPANIS

Fig. 2. Statistics on the Statures of Pygmy, American, and Nilote Males. (After Roberts, 1974.)

Figure 1 still does not reveal the full extent of the problem, since it shows only populations for which there are reasonably good anthropometric data and in which there has been some interest by American equipment manufacturers. The tallest peoples of the world are Nilotes from the region of the Sudan. The shortest are the pygmies from central Africa. Figure 2 compares those two populations with American civilians. The outline figures represent males corresponding to the mean statures for each of the three populations. Beside each figure, I have drawn lines showing the mean, the mean plus and minus one standard deviation, and the range from the 5th to the 95th percentiles. The full extent of normal differences in stature revealed here is truly remarkable. There is no overlap at all between the 90%-ranges of the pygmy and American males, and the difference between the pygmies and Nilotes is even greater. I do not believe that any ergonomist has yet tried to deal with differences of this magnitude.

Differences in Body Proportions

Anthropometric differences among the peoples of the world are not confined to stature, of course. Consideration of other anthropometric dimensions reveals that National and Cultural Variables in Ergonomics 9

Table 1. Ratios of Mean Sitting Height to Mean Stature for Various National Groups.

Men Women Men Women

Germany (military) 0.514 France (military) 0.526 Bulgaria (civilian) 0.522a 0.528a United Kingdom (military) 0.526 United States (civilian) 0.522 0.529a Korea (military) 0.528a Norway (civilian) 0.522 Thailand (military) 0.529a Canada (military) 0.525 Greece (military) 0.529 United States (military) 0.525 0.528a Vietnam (military) 0.530a France (civilian) 0.526 Turkey (military) 0.530a Italy (military) 0.526 Japan (military) 0.544a aIndicates those populations for which the mean stature is less than the 20th percentile of the U.S. military population. the problem of fitting equipment to people is more complex than appears at first glance. The problem is not solved simply by taking a piece of equipment built for one population and scaling it up or down for some other population. The difficulty is that body proportions also differ among the peoples of the earth. Table 1, for example, shows the sitting height to total height ratios for 15 populations. The data fall naturally into three groups. At the top are the Germans who, relative to other populations of the world, have much longer legs in proportion to their heights. Most populations fall in the central group, with body proportions not very dissimilar from those of Americans. At the bottom, and distinctly separate from all the others, are the Japanese. By contrast with other national groups, the Japanese have much shorter legs in proportion to their statures. The design implications of these data are extremely important. If one takes an American jetliner, automobile, or farm tractor, and scales it down to the stature of the average Japanese, the pedals and other foot controls will be completely beyond his reach. In our NATO symposium, Roberts pointed out similar problems encountered in standards for furniture that is supposed to accommodate both German and British dimensions.

Some Implications for Ergonomic Design

In the past, human factors engineers have typically tried to design for what we now recognize is a relatively narrowly defined population – Americans. British and other European ergonomists have done the same for their own nationals. In designing equipment, vehicles, factories, furniture and housing to accommodate all the peoples of the world, we face some formidable problems. In fact, it may 10 ALPHONSE CHAPANIS be necessary for us to change some of our conventional ways of thinking. For example, I am not sure that a commercial aircraft should have a single size of seat that tries to accommodate all passengers. Why should a 747, or a Concorde, not have several different sizes of seat, with different spacings between the seats, to accommodate different groups of people more adequately, rather than trying to make one size of seat fit all passengers? However, I leave it to future congresses of the IEA to explore how best to deal with the problems I have raised here.

THE BABEL OF TONGUES

Perhaps the second most obvious and impressive national and cultural variable that we have to face in internationalizing ergonomics is abundantly clear to all of us here in this Congress. It is that the peoples of the world do not speak a common language. Americans and the British sometimes say chauvinistically that English is the international language of science, technology and commerce. That may very well be. But only a foolhardy manufacturer would design equipment to be sold around the world with the expectation that he could use only English labels, instructions and maintenance manuals. No one really knows how many different languages are spoken on the face of the earth, but realistic estimates place that number at about 3000. If we focus only on the major languages of the world, we find that 130 are used by more than one million people each, and 70 of these by more than 5,000,000, 60% of the people on the face of the earth speak some language other than the five official languages of the United Nations – English, French, Russian, Spanish and Chinese. This diversity in languages creates problems that ergonomists and human factors engineers have only recently become aware of. The problems may be divided into two major categories: those dealing with oral or spoken communication, and those dealing with written communication.

Oral Communication

Problems of oral communication are common in man-machine systems. Two examples are in international air traffic and in international telephony. The problems are not as simply resolved as one might suppose at first glance. A number of years ago, for example, English was adopted as the international language for international air traffic and flight control. So by the stroke of an administrator’s pen, the problems seemed to have been solved. In actual fact, they were not. National and Cultural Variables in Ergonomics 11

English spoken by an Englishman is not at all the same as English spoken by an American. Even within a single country one can find marked regional accents. The English of a Yorkshireman is hardly comprehensible to a Londoner and the English of a Texan may be almost as incomprehensible to a Bostonian. When one considers English as it is spoken by foreign nationals, the problems increase in severity and complexity. An air traffic controller at Heathrow Airport in London, or Kennedy International Airport in New York, may, in the space of a few minutes, have to listen to English spoken by an Englishman, German, Frenchman, Russian, Israeli, and Swede. Although the speakers may all be using English, the listener might well wonder whether they are in fact speaking the same language. At our NATO symposium Ruffell-Smith reported a study of such language problems in international aviation. His conclusion was that these problems have by no means been solved and that variations in spoken English are responsible for a great deal of confusion, uncertainty and perhaps even accidents in air traffic. Recently I have been acting as a consultant to a project sponsored by our United States Coast Guard aimed at the reduction of ship accidents in American waters. As part of the project, I have been studying accident reports of ship-to-ship collisions and ship groundings. Language is clearly implicated in some of these accidents. Although I would not say that language is a major contributing cause of accidents, it turns up with sufficient frequency that we are proposing to take some positive action about it. The situation typically occurs when an American harbour pilot gives commands to an officer or a seaman on a foreign ship. In one case, for example, testimony showed that the command “Hard port !” from an American pilot was heard and interpreted as “Starboard” by the third mate of a Brazilian ship.

Basic Data on Spoken Languages If you pick up the second edition of the American Human Engineering Guide to Equipment Design (Van Cott & Kincade, 1972) and turn to Chapter 5, you will find there an impressive collection of information about speech communication. Measurements of the R.M.S. pressure of fundamental speech sounds, the dynamic range that has to be provided in speech communication systems, the measurement of speech intelligibility, the masking of speech by noise, peak clipping, reverbera- tion, the confusability of speech sounds, and many other important characteristics of speech are all discussed in relation to the design of auditory environments, systems and procedures. One difficulty with the chapter, however, is that it is entirely about male American voices speaking American English. I know of no comparably detailed data for languages other than English, French, Swedish and possibly German and Russian. In other words, we have essentially no information about the speech characteristics of the great majority of the spoken languages of 12 ALPHONSE CHAPANIS the world. To be sure that I am not misunderstood, I want to make clear that I am not referring to linguistic information about the spoken languages of the world. We have lots of that. I am talking about engineering information, the kind that would enable us, for example, to design effective speech communication systems for those languages.

Written Communication: Alphabet Differences

Language problems dealing with written communication are at least three in number. One of these arises from differences in alphabets and the implications of those differences for the design of typewriters, accounting machines and data entry devices of all kinds. Consider, for example, the differences among such similar European languages as French, German, Danish, Spanish and English, with their umlauts, accent marks, and other diacritical symbols. You can even find differences between English English and American English. For example, in America we use the sign Ø whereas the British use p. In America our principal monetary unit is the $; in Britain it is the £. At the NATO symposium, Hanes stated that the National Cash Register 399 accounting system provides standard keyboards for any of 17 countries or regions and that the company will manufacture other variations if they are ordered specially by customers. Figures 3–6 show NCR keyboards for the United Kingdom, France- Belgium-Italy-Portugal, Germany-Austria-Switzerland and Latin America. Close examination of even these few keyboards will reveal differences in the arrangement of the basic letters (compare the keyboards in Figs 3 and 4), symbols (compare all the keyboards), and in the words used for standard operations (for example, “load” on the U.K. keyboard is “charg” on the French, “laden” on the German and “car prog” on the Latin American versions). These examples do not, of course, reveal the full magnitude of the problem. In trying to cope with languages such as Arabic, Hebrew or Japanese, still further and more complex variations are encountered.

Conflicting Ergonomics Requirements The different alphabets of the world bring us face-to-face with conflicting ergonomics requirements. On the one hand, it is advantageous to provide a specially designed keyboard for each country or region where users have a common language. Such customized keyboards reduce training time, errors and operator fatigue. On the other hand, one has to keep in mind that operators may, in fact, move frequently from one geographical region to another, as, for example, within the Common Market countries. If keyboards are customized, operators will find different keyboards in use and so will face all the vexing problems National and Cultural Variables in Ergonomics 13

Fig. 3. National Cash Register 399 Accounting System Keyboard for the United Kingdom. (After Hanes, 1974.) associated with the transfer of training. Then, too, manufacturing costs cannot be ignored. Each different kind of keyboard means additional costs for its tooling, manufacture, inventory, shipping and training and maintenance manuals. For these several reasons, manufacturers have to decide exactly how much it is worth to customize a keyboard arrangement, especially for countries that have only limited sales potentials. These are problems not easy to solve from either an ergonomics or an economic point of view.

A Japanese Data Entry Device Human factors problems multiply when one deals with languages that are markedly different from English. Some of these problems are illustrated dramatically by the development of an IBM Japanese language key-entry device, a development that Brown described in detail at the NATO symposium. By far the biggest problem in designing a Japanese keyboard is that it has to accommodate the Japanese language. The basic written language of Japan, called Kanji, was adopted from Chinese approximately 1,700 years ago. It consists of a large, indeterminate number of pictographic characters. For most purposes, however, 2,000 to 4,000 14 ALPHONSE CHAPANIS

Fig. 4. National Cash Register 399 Accounting System Keyboard for France, Belgium, Italy, and Portugal. (After Hanes, 1974.)

Kanji characters generally suffice. In addition to the Kanji characters, several other classes of characters have to be included on a Japanese keyboard. These are:
Hiragana, a phonetic subset of about 50 characters used in Japanese syntax;
Katakana, a phonetic subset of about 50 characters used to transcribe foreign expressions;
Japanese numerals, the equivalent of 0 through 9, plus a few other symbols for large quantities such as 100 and 1000;
Arabic numerals, 10 in all;
the Roman alphabet, consisting of 26 characters;
miscellaneous characters consisting of such things as special numerals for financial transactions. Imagine trying to accommodate all these characters with conventional European typewriter keys. With one character per key, the keyboard itself would be unmanageably large. Designing a keyboard for a language as complex as Japanese clearly requires principles that are different from those used on European National and Cultural Variables in Ergonomics 15

Fig. 5. National Cash Register 399 Accounting System Keyboard for Germany, Austria and Switzerland. (After Hanes, 1974.) keyboards. The IBM key-entry device makes use of two such principles. The first has to do with the size of the keys and the method used to actuate them. Keys, or character positions, are just about 4 mm in size. Since this is clearly too small to be depressed with a finger, a stylus is used. Figure 7 shows the stylus keyboard mounted on an IBM 029 card punch table. The second principle has to do with the organization of the characters on the keyboard. In essence, this principle makes use of some statistical properties of the Japanese language: the most frequently used characters are put in a special high-usage region on the keyboard. For example, about 70% of ordinary Japanese can be written with 600 of the most frequently used characters. Grouping these characters into a special region decreases the average area that has to be searched, and the average distance that has to be reached, in keying ordinary text. By using these two principles, Brown was able to show that skilled Japanese operators can key about the same average number of words per minute, and about as accurately, as skilled American card-punch operators. 16 ALPHONSE CHAPANIS

Fig. 6. National Cash Register 399 Accounting System Keyboard for Latin America. (After Hanes, 1974.)

For all of its inherent interest as a problem in applied ergonomics, this example is particularly useful to the purpose of my paper. The new American Human Engineering Guide to Equipment Design (Van Cott & Kincade, 1972) has an entire chapter devoted to “Data Entry Devices and Procedures.” The chapter concludes with over 110 references. For all practical purposes, however, that chapter might just as well have been labeled “English-Language Data Entry Devices and Procedures.” Nothing in it helped, or could have helped, to design the IBM Japanese keyboard that I have just described. Nor will that chapter help in the design of Hebrew, Arabic or Russian keyboards. This part of human factors engineering has clearly still not escaped its parochial American boundaries.

Legibility of Various Alphabets The legibility of numerals and letters is an old problem in ergonomics, even predating the field of ergonomics itself. The Cornog and Rose handbook (1967) National and Cultural Variables in Ergonomics 17

Fig. 7. IBM Japanese Language Stylus Keyboard on an IBM 029 Card Punch Table. (After Brown, 1974.) on alphanumeric characters and other symbols has references dating back to 1898. However, to repeat a theme that has by now become familiar, the handbook might just as well have been titled, “Legibility of Alphanumeric Characters for English-Speaking Persons.” In view of the hundreds of different alphabets in the world, it is strange that so little is known about them in a human engineering sense. For example, virtually nothing is known about the comparative legibilities of different alphabets. One of the two studies of this kind that I know about was reported at our NATO symposium by Daftuar who compared the legibilities of Roman, Devanagari and Bengali alphabets. His subjects were Bengali-speaking students who also knew both Devanagari and English. His results, in Fig. 8, show that the Roman alphabet was significantly more legible than either Devanagari or Bengali. This is particularly interesting in view of the fact that the subjects in his experiment were not born to the English tongue. Many more studies of this type are 18 ALPHONSE CHAPANIS

Fig. 8. Comparative Legibilities of the Roman, Devanagari and Bengali Alphabets. (After Daftuar, 1974.) needed before we can speak with assurance about alphanumeric displays for international purposes.

Written Communication: Content Problems

A second major class of problems arising from language differences is content problems: problems associated with meanings and with the content words available in a particular language. These problems are, of course, important in the preparation of instructions, maintenance manuals, signs and technical written materials of all kinds, and they are much greater than many ergonomists realize (Chapanis, 1965). Even within the same language, regional differences in meaning may produce great confusion in the use of man-machine systems. To illustrate that point, I recommend National and Cultural Variables in Ergonomics 19 a book called British Self-Taught: With Comments in American (Schur, 1973). This witty lexicon is a compilation of words that mean different things in Britain and in America. Perhaps the most impressive thing about the book is its sheer size and the number of entries in it. A few examples will have to suffice. The word “tube” means “underground” to an Englishman, and “television screen” to an American. Screen demist, wing, bonnet and boot in Britain are almost never interpreted correctly by Americans. In America the corresponding terms are defroster, fender, hood and trunk. I will not go on multiplying examples of this kind. Glance through the book if you want a glimpse of some practical problems of international ergonomics. Problems of meaning become much more serious when we turn to languages that are quite different from English. For one thing, many languages have a surprising paucity of technical terms. Vietnamese, for example, is rich in content dealing with literature, history, philosophy and the arts, but it is deficient in technical vocabulary. As a result, translators who translate technical material into Vietnamese have to use various stratagems to overcome the lack of appropriate words. They may, for example, invent words as required. An example is the American “piston” which, on occasion, has been translated into “pittong” in Vietnamese. Another stratagem is to adopt words from other languages that are richer in technical content, for example, Chinese or French. Still a third stratagem is to use circumlocutions that can be translated into the target language. An example is “tachometer” which has occasionally been translated into “rotation measuring device” in Vietnamese. Such problems of the scarcity of technical vocabulary occur in many languages. They have typically been ignored by ergonomists and human factors specialists.

Written Communication: Translations

The “obvious” solution to the problems raised by the number of different languages in the world, and the solution that occurs to most people when they first encounter these problems, is to call for translations of technical materials from one language into another. Although this appears to be both “obvious” and a “solution,” I have already said enough to suggest that the difficulties are much greater than most people realize. Still another problem, for example, is that translation by human translators is very slow. Measurements by Sinaiko and Brislin (1970) give an upper limit of about 400 to 450 words per hour for translations between almost any pair of languages. Machine translation, once thought to be the solution to all our translation problems, is still far from good. In one of the very few good empirical evaluations of translation by computer, Sinaiko and Klare (1971a, b, 1972) found that the best machine translations were invariably worse than the best translations that could be made by skilled humans, and that the difference between 20 ALPHONSE CHAPANIS the quality of human and machine translations increased as the difficulty of the material increased. Finally, skilled professional translators are often in very short supply. This is particularly true of technical translators. It is dangerous to assume that any bilingual person can do translations at all or that he can handle unfamiliar subject matter. In view of these problems, it is not surprising to find many technical translations not done well. Sinaiko reported at our NATO symposium what may be the only empirical test of the quality of translated material (Sinaiko & Brislin, 1970). Vietnamese and American maintenance men, with identical training and experience, were asked to perform a twelve-step maintenance task on a helicopter. The Vietnamese subjects used one of four maintenance manuals: a high quality translation produced under the control of the experimenters; a lesser quality translation produced by a “free lance” translator over whom the experimenters had no control; a poor quality translation purchased through a commercial translating service; or the original English text. The findings show that it makes a substantial difference whether or not one works in his native tongue or in a second language. The Vietnamese had learned English reasonably well and had been trained as technicians in that language. When they were forced to use English maintenance manuals, however, they did much worse than the Americans. Second, the quality of the translated material had a marked effect on performance. The Vietnamese who used the best translations equaled the performance of the Americans. The poor quality commercial translation, however, resulted in performance that was about seven times worse than the performance turned in by the technicians who used the high quality translation. Findings such as these have some important implications for ergonomics. It is easy to say that one can translate ergonomics standards, technical materials, instructions, signs and other written material into other languages. That does not assure that the translated materials will in fact serve their intended purpose when they are used by the target population for whom they have been translated.

PHYSIOLOGICAL DIFFERENCES

Most ergonomists seem to take it for granted that physiological differences among the various peoples of the earth are small or non-existent. Nonetheless, some data seem to show that such differences do in fact exist. Guthrie et al. (1970), for example, tested the strength of grip of Vietnamese and American soldiers using a simple hand dynamometer. They found surprising differences between the two populations. The 75th percentile Vietnamese had a strength of grip that corresponded to about the 10th percentile of the Americans. National and Cultural Variables in Ergonomics 21

Table 2. Muscles of the Hand and Forearm Whose Dimensions, Proportions and Insertions Vary Significantly Among Various Races. (After Roberts, 1974.)

Pronator teres Brachioradialis Flexor carpi radialis Extensores carpi radialis longus et brevis Palmaris longus Extensor digitorum manus communis longus Flexor carpi ulnaris Extensor indicis proprius Flexor digitorum sublimis Extensor digiti V manus proprius Flexor pollicis longus Abductor pollicis longus Pronator quadratus Extensor pollicis brevis

A study of Swedish and Indian forestry workers carried out jointly by the Institute of Work Physiology, Stockholm, and the Forest Research Institute, Dehradun (Hansson et al., 1966), has revealed striking differences between the maximal aerobic capacity of Indian and Swedish forestry workers. The work capacity of the former was about 60% that of the latter. Similarly, Wyndham (1968) reports marked differences between the work capacities of Bantu and European mine workers. These differences are so great that work standards for Bantus have to be adjusted downward as compared with those for Europeans. Finally, Roberts has compiled a list of 16 muscles of the hand and forearm whose dimensions, proportions and insertions vary significantly among various races (Table 2). Although Roberts has not drawn out the ergonomic implica- tions of these differences, it is reasonable to assume that such implications do exist. These are a few examples of physiological and biomechanical differences that have been found among certain selected groups. All of them bear on man’s capacity to work. Although such differences are of somewhat lesser importance in the highly automated and mechanized industries of America, they might very well be of considerable importance in regions where heavy physical labour has not yet been supplanted by machines.

PSYCHOLOGICAL DIFFERENCES

The belief that different ethnic groups differ psychologically (in intelligence, temperament, and emotion) must certainly go back to prehistory. Such stereo- types, whether real or imagined, can be found in the folklore, literature, religious writings and laws of virtually all societies from the days of the ancients to the present. Scientific studies of race differences date from the time of Francis Galton and the literature on this subject since Galton’s time has become voluminous 22 ALPHONSE CHAPANIS and controversial. Virtually none of that literature is useful for our purposes. Psychologists, anthropologists and sociologists have been most concerned with trying to decide what proportion of the differences they observed could be attributed to heredity and what proportion to learning. Never have they been concerned with the impact of such differences, whether innate or learned, on man at work.

The Ability to Adapt to Technology

Meanwhile, there is some evidence that man everywhere seems capable of adapting to a highly technological society. At our symposium, Sinaiko reported on the results of approximately 100 interviews with people who had had extensive experience introducing technology in the Philippines, China, Korea, Afghanistan, Turkey, Pakistan, India, Nigeria and Thailand. His informants were responsible for computers, jet transports, heavy construction equipment, communication systems and military equipment. Their practical experience suggests that probably any people in the world can learn to operate and maintain highly sophisticated equipment. The Boeing Company, for example, reports that Vietnamese pilots and maintenance men trained in regular courses in Seattle perform as well as their counterparts who work for U.S. airlines. In the 1950s, Turkish labourers who were illiterate in their own language were sent to U.S. Air Force Schools in America, where, after language and technical training, they qualified as jet engine maintenance men. At least two of the men later became managers of engine repair centres in Turkey. Other industrial contractors can cite experiences with skilled heavy equipment operators whose origins were the rice paddies of Vietnam or Thailand only a few years earlier. In organizing new airlines, Pan American World Airways has trained jet pilots and mechanics from among men who were less than a generation removed from camel drivers in Afghanistan, Iran or Pakistan. IBM World Trade Corporation reports that previously unskilled Nigerians have been trained to handle the installation and maintenance of computers. These examples of dramatic training results were, however, not easy to find. In addition, Sinaiko stated that in almost every one of these cases, training had to be preceded by rigorous preparation in the English language. To Americans, some foreign nationals appear to acquire new knowledge very slowly and the attrition rates of foreign trainees are often high. It may be, however, that these problems are reflections more of the difficulties of working in an unfamiliar language than of any inherent learning deficiency. At any rate, as far as we can tell at the present time, the psychological capacity to absorb technical training is not limited by race or ethnic origin. National and Cultural Variables in Ergonomics 23

Perceptual Differences

For all that, there do appear to be some genuine differences among certain racial groups in the way they perceive things, and these do have some ergonomic implications. For example, Dart (1972) made some anthropological studies on the Nepalese in the Himalayas. He found these people unable to use maps or diagrams representing geographical space. Sinaiko also reports that the Vietnamese found it extremely difficult to understand the interpretation of technical drawings and diagrams associated with equipment. Similar findings have been reported by Winter (1963) among the Bantus of South Africa. Winter’s experiences are particularly relevant to ergonomics. He reports that safety posters whose intent and meaning could be grasped immediately by Europeans were completely unintelligible, or even completely misunderstood, by the Bantus. If these findings are correct, and they do seem to be trustworthy, it would appear that we need to take special precautions in the design of visual displays of all kinds (dials, indicators, signs and diagrams) for selected populations of the earth. With these observations I leave the matter of psychological differences among various national and cultural groups, not because they are uninteresting or unimportant, but only because there is so little else I can say about them with any assurance. Perhaps it is time for ergonomists, human factors engineers, and other biological scientists with applied orientations to take a fresh look at this old problem.

PRACTICES AND CUSTOMS

The last class of variable that I shall discuss is those associated with habitual or traditional ways of doing things. In my view, these fall into two major categories.

Differences in Practices Due to Difference in Technology

In some cases, differences in practices that we find in various regions of the world are due to variations in the way technology was introduced into those areas. Through years, and sometimes generations, of familiarity with different technology, different national groups have come to react almost “instinctively” but in highly distinctive ways to essentially the same machine system. So long as world-wide travel and commerce were only a relatively small fraction of any country’s gross national product, such differences were relatively unimportant. In modern times, however, such differences can cause confusion, inefficiency and 24 ALPHONSE CHAPANIS accidents. Differences between the rules of the road in Britain and some of the Commonwealth countries on the one hand, and those in most other countries on the other, are an obvious example. Americans who have traveled to Britain are full of stories about the near accidents they have had with British traffic because they automatically look and move the wrong way. Englishmen traveling in America or on the Continent have experienced the same problems. A somewhat more subtle example of the same kind of problem occurs in telephony. The various national and regional telephone systems of the world developed rather independently of one another. As a result, differences among the various telephone systems of the world have for years provided amusing stories for travelers. Now that international telephony has become so commonplace, however, what used to be merely amusing has become a matter of serious ergonomics concern. People now want, and expect to get, international service that is just as familiar, understandable and efficient as they would get on domestic calls. Some problems of international telephony are being handled by a working group on human factors through the Consultative Committee on International Telegraphy and Telephony. Archbold reported on some of that work at the NATO symposium. To illustrate only one facet of the work of this group, I show in Table 3 the percentage of experimental subjects in each of nine countries who judged certain tones as ringing tones. It is especially interesting that while Britain and France are, for all practical purposes, neighbours, each has a ringing tone that is not generally recognized as such in the other country. Note also that the German ringing tone fares badly both in Australia and the U.K. The two best overall tones are the American Precise Tone Plan (PTP) ringing tone and the Japanese ringing tone. On the average they are recognized as ringing tones by about 95% of people in the nine countries shown here.

Culturally-Based Customs

Another source of important differences among the various peoples of the earth can be traced to ways of thinking, of behaving and of living that have gradually evolved through generations and have become accepted as natural. Indeed, our own customs are so natural to each of us that we are oftentimes startled to discover that other people do things differently.

Working Postures Consider the fact that we are all sitting on chairs. To us, this is the natural posture to assume at work, in the office or at home. Indeed, every handbook of human factors or ergonomics that I know about assumes that industrial workers will use National and Cultural Variables in Ergonomics 25 Kingdom States Republic of Germany as a “Ringing” Tone (After Archbold, 1974). Percent of Experimental Subjects in Each of Nine Countries Judging Each Tone in the Left-Hand Column 1. American PTP2. German3. Irish and British4. Japanese5. Czech, French, Dutch, etc.6. American “Line busy” 977. American 44 PTP “Line busy”8. 100 American PTP “Reorder” 38 76 81 199. French 26 transfer 100 47 6 84 21 97 28 87 94 74 14 100 100 100 100 4 3 4 69 100 100 0 69 27 91 100 50 0 0 75 100 100 0 100 37 100 100 4 0 0 97 63 100 9 85 0 42 0 100 13 100 0 37 97 7 87 0 34 90 55 0 81 33 0 91 58 30 3 0 0 3 18 6 10. CCITT special 31 46 29 38 56 48 33 53 44 Table 3. Busy Tones Special Tones Ringing Tones Tone Australia France Federal Holland Japan Spain Sweden United United 26 ALPHONSE CHAPANIS either a sitting or a standing posture. Many peoples of the earth, however, work in squatting posture or while sitting on the ground. This is true of most Asians and Africans. I know of no human factors work that deals with such working postures. In India, for example, the Indane Company has designed cooking stoves that can be used on the ground in Indian kitchens. Sewing machines throughout vast segments of the earth have to be designed to be used on the ground. The same is true of a great variety of tools and implements such as those used in carpentry, plumbing, and metal working. Whether such customs will change over the years is hard to predict. Suffice it to say, the complete international ergonomist today needs to know about working postures that may be quite different from those that he himself may accept as natural.

Cultural Differences in Learning Sometimes, variations in customs have more subtle effects. At the NATO symposium, for example, Sinaiko reported differences in the way peoples from various cultures learn. Learning in Vietnam and elsewhere in the Orient appears to be largely accomplished through imitation, memorization and rote. This is true of theoretical or conceptual material, as well as of more practical skills. There is also a tendency to believe that, having learned the words, one has thereby acquired understanding and knowledge. Finally, because of their strong need to “save face,” Asian students rarely ask questions. To do so implies ignorance. For these several reasons, careful attention needs to be paid to training, and to testing the effects of that training in technical subjects.

Attitudes Towards Work In some parts of the world, there are strong biases against tasks or occupations associated with “dirty hands.” These attitudes are also reflected in lower average pay scales for people in such occupations. One reason for the scarcity of profes- sional engineers among the educated in many Asian countries is that engineering is a low status occupation. Liberal arts, languages and the humanities are more popular courses of study in universities. In addition, top-level managers in some countries may be appointed because of political influence or family status rather than because of their experience or knowledge. In technical enterprises, such managers often do not perform effectively. Their lack of expertise, together with their “no dirty hands” attitudes, make for inadequate supervision. Similar problems were discussed at the NATO symposium by Wyndham who reported that conservative, tribally-oriented Bantu mine workers are not generally interested in learning new skills or in seeking promotions. They cling to the jobs that their tribe has by tradition always had. Their resistance to innovations of any sort has been an obstacle to the introduction of even simple ergonomics principles National and Cultural Variables in Ergonomics 27 in the work situation. For example, more ergonomic methods of shoveling rock in mines have been a failure even though the new methods would have led to a reduction of energy expenditure, greater production of rock from the slopes, and higher bonuses. Related problems were reported by Daftuar. Rural India is a closely knit society with commonly shared work and family responsibilities operating through two social institutions: the family system and the castes. Although there are indications lately that the family is showing some signs of disintegration, it, and not the individual, is still the basic social unit. Village life is still tradition-bound and jobs in villages are generally allocated and designed on the basis of family and caste. The son of a carpenter is most likely to become a carpenter and he will generally inherit from his father not only his job, but his tools and postural patterns. In traditional India, a job is generally regarded as a family responsibility to such an extent that its performance is shared by all the members of a family. Moreover, attempts to bring in anyone from outside the family are likely to be resented. This kind of thinking was responsible for a widespread resistance movement in the Tata Iron and Steel Company (TISCO), Jamshedpur, some time ago. In the end, management acceded to worker demands that, in filling all future vacancies, the company would absorb at least one descendant of currently employed workers. This practice is still observed in TISCO and certain other companies. The caste system in India may have equally throttling effects. Fraser (1966) states that the failure of a weavers’ cooperative started at Barpali Village Service was due in part to the fact that the project technicians, interested only in selecting the best workers, had drawn weavers from different caste groups that could not by tradition work together.

Ergonomics and Social Change Underlying these differences in practices and customs is another problem that may be more complex than any I have yet mentioned. Almost every society in the world today is in a state of ferment. Customs, habits and styles of life are changing everywhere and changing rapidly. Should ergonomics design for societies as they are today? Should we design for societies as they are likely to become at some time in the future? Or should ergonomics design for societies as they ought to be? Ergonomics has, I believe, operated on the premise that we should design for people as we find them. My thought is that ergonomics, through careful design, should deliberately become an active force in producing societal change. Some of you may regard this as a radical idea. That may very well be. At the very least, it involves us in questions of an ethical nature: in questions of what ought to be. These questions are far too philosophical for me to address here. I leave their resolution to future congresses of the IEA. 28 ALPHONSE CHAPANIS

CONCLUSION

By now each of you should have formulated your own answer to the question I asked at the very beginning of my talk, “How international is ergonomics?” My own conclusion is that a good American human factors engineer or a good European ergonomist might very quickly become disillusioned and frustrated if he tried to apply his principles innocently in large regions of the earth. Rectifying that situation means that we need to internationalize ergonomics. That in turn requires that we approach our research, education and applications with broader perspectives, perspectives so broad that they are capable of encompassing the whole world and all the peoples on it. When we have done this, we will have given new depth to the name of our organization: the International Ergonomics Association.

ACKNOWLEDGMENT

I am indebted to the Advisory Group on Human Factors of the Scientific Affairs Division, North Atlantic Treaty Organization, for its financial support of the symposium on which this paper is largely based, and so of this paper itself.

REFERENCES

Archbold, R. B. (1974). Multinational psychological research in telephony. In: A. Chapanis (Ed.), Ethnic Variables in Human Factors Engineering. Baltimore: Johns Hopkins Press (in press). Brown, C. R. (1974). Human factors problems in the design and evaluation of key-entry devices for the Japanese language. In: A. Chapanis (Ed.), Ethnic Variables in Human Factors Engineering. Baltimore: Johns Hopkins Press (in press). Chapanis, A. (1965). Words, words, words. Human Factors, 7, 1–17. Chapanis, A. (1970). Relevance of physiological and psychological criteria to man-machine systems: The present state of the art. Ergonomics, 13, 337–346. Chapanis, A. (Ed.) (1974). Ethnic variables in human factors engineering. Baltimore: Johns Hopkins Press (in press). Commission on International Development (1969). Partners in development. New York: Praeger. Cornog, D. Y., & Rose, F. C. (1967). Legibility of alphanumeric characters and other symbols: II. A reference handbook. Washington, DC: U.S. Government Printing Office. Daftuar, C. N. (1974). The role of human factors engineering in underdeveloped countries with reference to India. In: A. Chapanis (Ed.), Ethnic Variables in Human Factors Engineering. Baltimore: Johns Hopkins Press (in press). Dart, F. (1972). Science and the worldview. Physics Today, 25(6), 48–54. Fraser, T. M., Jr. (1966). Sociocultural parameters in directed change. In: A. H. Neihoft (Ed.), A Case Book of Social Change. Chicago: Aldine. National and Cultural Variables in Ergonomics 29

Guthrie, G. M., Brislin, R. W., & Sinaiko, H. W. (1970). Some aptitudes and abilities of Vietnamese technicians: Implications for training. Arlington: Institute for Defense analyses. Research Paper P-659. Hanes, R. M. (1974). Human factors in international keyboard arrangement. In: A. Chapanis (Ed.), Ethnic Variables in Human Factors Engineering. Baltimore: Johns Hopkins Press (in press). Hansson, J. E., Lindholm, A., & Birath, H. (1966). Men and tools in Indian logging operations. A pilot study in ergonomics. Institutionen for Skogsteknik, Department of Operational Efficiency, Royal College of Forestry, Stockholm. Report No. 29. Kennedy, K. W. (1974). International anthropometric variability and its effect on aircraft cockpit de- sign. In: A. Chapanis (Ed.), Ethnic Variables in Human Factors Engineering. Baltimore: Johns Hopkins Press (in press). Mead, M. (Ed.) (1953). Cultural patterns and technical change. Paris: United Nations Educational, Scientific and Cultural Organization. Roberts, D. F. (1974). Population differences in dimensions, their genetic basis and their relevance to practical problems of design. In: A. Chapanis (Ed.), Ethnic Variables in Human Factors Engineering. Baltimore: Johns Hopkins Press (in press). Schur, N. W. (1973). British self-taught: With comments in American. New York: Macmillan. Sinaiko, H. W., & Brislin, R. W. (1970). Experiments in language translation: Technical English-to- Vietnamese. Arlington: Institute For Defense Analyses. Research Paper P-634. Sinaiko, H. W., & Klare, G. R. (1971a). Further experiments in language translation: A second eval- uation of the readability of computer translations. Arlington: Institute for Defense Analyses. Research Paper P-805. Sinaiko, H. W., & Klare, G. R. (1971b). Further experiments in language translation: Readability of computer translations. ITL, 15, 1–29. Van Cott, H. P., & Kincade, R. G. (Eds) (1972). Human engineering guide to equipment design (2nd ed.). Washington, DC: U.S. Government Printing Office. Winter, W. (1963). The perception of safety posters of Bantu industrial workers. Psychologica Africana, 10, 127–135. Wyndham, C. H. (1968). An examination of the methods of physical classification of African labourers for manual work. South African Medical Journal, 40, 275–278. This Page Intentionally Left Blank 2. CULTURE, CONTEXT AND PERFORMANCE

Neville Moray

ABSTRACT

Constraints on human-machine systems’ performance are generally treated as due to anatomy, physiology, and cognitive or behavioral limits. It is assumed that research findings can be universally applied to the design of such systems. It is now clear that social and cultural constraints are equally important, even in simple work systems. Context and culture are at least as important as limits of cognitive ability, and in many situations social and cultural factors are the dominant constraints on performance. This is particularly true in the cross-cultural transfer of advanced technological systems. A particularly clear example is given by population stereotypes of stimulus-response relations.

INTRODUCTION: CULTURE, ERGONOMICS AND PSYCHOLOGICAL LAWS

Are psychological “laws” of behavior universal and context independent, or are the social and cultural contexts of a human-machine system as great a constraint on behavior as the structure of the nervous system? Which alternative is true should have a profound effect on the design of human-machine systems (HMSs) of all kinds, from the simplest to the most technologically advanced and complex.

Cultural Ergonomics Advances in Human Performance and Cognitive Engineering Research, Volume 4, 31–59 Copyright © 2004 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1479-3601/doi:10.1016/S1479-3601(03)04002-5 31 32 NEVILLE MORAY

Most research on behavioral and cognitive psychology uses paradigms taken from physical science and engineering, with an implicit (and sometimes explicit) belief that the findings represent universal laws of behavior. Similar beliefs apply to human factors and ergonomics.1 We find statements about the magnitude of just noticeable differences (jnds) in psychophysics that imply that jnds are a char- acteristic of all humans. Simple reaction times to a light or to a sound are treated as if they are also universals, and it is implied that the only important variability is stochastic variability about the mean. In more complex situations we find “Hick’s Law” and “Fitts’s Law,” Shannon’s Mathematical Theory of Communication (Shannon & Weaver, 1947), the Theory of Signal Detection (Swets, 1996), and the modeling of human performance by the laws of classical (McRuer & Krendel, 1976) or optimal (Kleinman et al., 1970) Control Theory. In some cases caveats are issued: the human must be well practised and well motivated, and must know (implicitly) the statistical structure of the stimulus sets and the relative value and costs associated with behavioral outcomes. Nevertheless an implicit claim for universal laws is never far. Such laws would be of enormous value to applied psychology, particularly human factors and ergonomics, because they would provide a firm scientific basis for the design of HMSs in the widest sense. But we will see that in applied behavioral science belief in such context-free laws is seriously misplaced. To some extent, of course, the truth of such laws does not matter in applied science and engineering. For the theorist, concerned to use behavioral experiments to de- termine the ultimate structure of the mind and the brain, the truth and universality of research findings is crucial. For the practitioner whose concern is above all to predict the performance of humans in their daily environments, the need is really for applicable and useful approximations, not for truthful laws. But even so, I will argue that context, social factors, and culture have a major role as constraints on human-machine system performance. Psychological laws are not even sufficiently general for applied systems design, unless modified by social and cultural factors. The idea that culture2 is extremely important in ergonomics is not new. As long ago as the mid-1970s Chapanis organized a conference on this topic and wrote several papers to draw attention to the problem (Chapanis, 1974, 1975). There is also long tradition in European ergonomics of using anthropology and sociology in ergonomics (see Wisner, 1995, for an introduction). The very notion of a “human-machine system” (HMS) implies that behavioral laws are constrained by the context provided by machines, as does the increasing tendency to reject the “Fitts’s List” approach to allocation of function, and to think in terms of the overall properties of joint HMSs, see for example (Hollnagel, 1998; Woods et al., 1994). The way in which social factors envelop ergonomics can be summed up in a diagram such as Fig. 1. Culture, Context and Performance 33

Fig. 1. The General Structure of Socio-Technical Systems. A Physical Device can be Anything from a Hand-Held Camera to a Nuclear Power Plant or a Railway.

A contrary view, expressed by workers such as Senders (Senders, 1983; Senders et al., 1965), is worth examining. He has argued that in the limit, or “at the margin” (Senders, 2002; personal communication), behavior is necessarily constrained by the physical laws governing the system, and hence such laws can be treated as universal. In a series of classical experiments on dynamic visual attention (following the seminal work of Fitts and his co-workers (Jones et al., 1949, for example) he claimed that the way in which the eyes fixate dynamic displays is governed necessarily by Shannon’s Sampling Theorem and the duration of fixation by the entropy of the signal, as predicted by Information Theory (Shannon & Weaver, 1947). But while his experiments did indeed show the predicted linear relation between the bandwidth and the sampling rate, those experiments showed that it was not the whole story. In the first place, the operators had to be highly practised, sufficiently to have internally modelled the bandwidth of each signal. 34 NEVILLE MORAY

As Senders (1983) put it, “No beginners need apply.” Further more, at very low bandwidths observers sampled more frequently, and at high frequencies less frequently, than the Sampling Theorem predicted. The second effect seems to be due to the ability to detect rates of change as well as instantaneous values of variables (Senders, 1983), while the first is probably due to forgetting (Moray et al., 1973). That is, both these departures from the predicted relation were due to inherent properties of the nervous system, and hence, in the sense explored here, “universal.” In contrast the level of practice is a contextual factor, differing from situation to situation, and unless it is known, it will not be possible reliably to predict the observers’ behavior. Senders’s experiments had several special characteristics. The bandwidths were very low, so that the probability that attention had to be paid to two signals at the same time was almost zero, and there were no differential values among the displays. Now the value of a pay-off to an observer is a subjective, not an objective, variable, and it is intuitively obvious that if the failure to detect the value of one display were to result in a very high penalty, or its detection to a very high reward, there would be bias to observe the display with the most valuable (costly) outcome. This in turn means that the sampling bias of observers in experiments like those of Senders will be socially and culturally determined by work habits, social pressures, cultural values, etc. The very fact that people volunteer to act as “subjects” in experiments, whether paid or unpaid, is a social factor, as is how hard they will try to reach their performance limits once in the experiment. In an experiment on subjective effort conducted by the present writer, one participant said that he had tried to increase his effort as the task became harder, while another said that he did not bother – after all he was being paid for an hour’s participation, not to exhaust himself! Senders states that the physical laws that he believes predict behavior apply “at the margin,” where people are constrained only by the physics of the situation, and for such situations he may well be correct; but people may choose not to work at the margin, and settle for a lower level of performance with lower reward. We noted that in Senders’s experiments the bandwidth was sufficiently low that there was almost never a need to sample more than one display at any one instant, but in general in the real world that is not the case. A car or train driver has many sources of information to sample – speedometers, people beside the track or road, signals, other vehicles, and so on – and there will frequently be a need to look at more than one thing at a time. Because of the sequential sampling nature of visual attention to dynamic displays this will mean that a choice will have to be made frequently between two sources of information that require simultaneous attention, and in such a case again some kind of subjective bias will determine the choice, not merely the physics of the situation (Moray & Inagaki, 2001). Culture, Context and Performance 35

Thus even in a relatively simple behavioral laboratory experiment there are almost always factors other than the physical parameters of the situation that constrain behavior, and many of these are learned, social, or cultural. When experimenters offer a bonus for the best performance by participants they are invoking cultural factors. Not all cultures view competition to be “best” as a desirable end. Some years ago the present writer repeated an experiment originally performed at RAND on competitive/cooperative decision-making. A group of students had to choose which of several very different objects would be given to the group, and then decide which person should be given the object. The objects included chocolate, money, a book, etc., and had very different (monetary) values. At RAND there had been intense discussion and bargaining, with offers of side payments by those who particularly wanted an object, and the bargaining continued for nearly half an hour. By contrast my English students simply decided, in less than five minutes, that the obvious way to solve the problem was to draw lots to see who should decide on the object and then be given it. “That’s obviously the fairest way to do it.” Paradoxically, even in the most successful hard engineering models of behavior operators use context to shape their behavior. When classical control theory is used to model the human operator in an HMS, it is found that for well-practised operators there is no single transfer function that describes the human. Instead the human possesses a set of transfer functions, from which he or she selects one that, when coupled to the machine’s transfer function, is appropriate to provide an adequate margin of stability (McRuer & Krendel, 1976; Young, 1973). Furthermore, at very low bandwiths operators seldom exercise continuous control of a continuous disturbance, but develop idiosyncratic patterns of intervention that overall satisfy their desire to keep the system stable. It is often observed in a tracking task that when given an indication of the acceptable error, operators choose work to a criterion of their own that may be stricter, if that makes their task subjectively easier: they modify the task to suit themselves in so far as that is possible. As Simon has pointed out, the most impressive characteristic of humans is that they adaptive to the context in which they find themselves.

The human mind is an adaptive system. It chooses behaviors in the light of its goals, and as appropriate to the particular context in which it is working. Moreover it can store new knowledge and skills that will help it attain its goals more effectively tomorrow than yesterday. It can learn. As a consequence of the mind’s capacities for adaptation and learning, human behavior is highly flexible and variable, altered by both circumstances and experience. Nor are the alternatives from which the actor might choose usually known in advance. Human beings spend much of their time inventing or discovering actions that fit the circumstances (Simon, 1992, quoted by Wisner, 1995, p. 1549). 36 NEVILLE MORAY

Perhaps “at the margins,” and where a task is heavily constrained by the anatomical and physiological properties of sense organs and effects, it may be true that there are universal laws. Unfortunately, there are relatively few cases in real work as distinct from laboratory tasks where these are the only constraints. For example, Clement et al. (1968) observed eye movements of pilots in the cockpit of a large transport aircraft, and while they found a rough relation between sampling and bandwidth, it was not strictly proportional. Different instruments showed different multiples of what the sampling theorem would have predicted. Similarly Moray et al. (1980, 1983) found that radar operators sampled displays almost continuously even when there was little change on the display, and speculated that this was because it was very important that operators did not lose track of aircraft movement. To avoid forgetting they greatly over-sampled compared with the prediction of the Sampling Theorem. Overall, the theme developed here is charmingly summarized by Welford:

Information processing determines some of the limits of what a man can do but this is not the same as what he is likely to do “in the service of certain ends directly and indirectly connected with family or social position (Welford, 1951, quoted by Singleton, 1997, p. 132).

The world of work is frequently more akin to the novels of Trollope than to the laboratory of Skinner. Let us now look more directly at examples of how cultural differences can constrain performance.

ANTHROPMETRICS

Factors as simple as differences in anatomical dimensions can be cultural. Chapanis (1975) provided evidence for large differences in the anthropometric measurements of military personnel of difference races., and such differences have been frequently confirmed and the range of populations widened over which they have been measured. Fernandez et al. (1989) studied Korean female factory workers, and found that while there was little difference between Korean and Western workers’ anatomical characteristics on many measures, there were some very significant differences. Table 1 reproduces some of their data. The eye height difference is enough to cause difficulties if Asian women have to look over a high control panel when operating equipment manufactured in the West. Similarly, Okunribido (2000) surveyed the dimensions of the hands of western Nigerians and compared them with American, British and Hong Kong hands. He found that there were sufficient differences to affect the ability of Nigerians to use a “power grip” when using Western tools, and that it might be difficult for them to work when wearing safety gloves sized for Western hands. Culture, Context and Performance 37

Table 1. Comparative Anthropometry of Asian and Western Women. All Heights in mm (Fernandez et al., 1989).

Population Measure American British Korean Japanese

Standing height 1625 1620 1580 1530 Standing eye height 1525 1515 1471 1425 Sitting height 860 855 833 845 Sitting eye height 750 745 724 735

Singh et al. (1995) concluded that on anthropometric criteria alone, 40% of local applicants wishing to be trained as Singapore military pilots would be rejected, since cockpit dimensions of fighter aircraft are based on the 5th and 95th percentiles of Western anthropometric data. They comment that,

...pilots in Asia are endowed with a different genetic and physical constitution. In general they are smaller and more slender than their Caucasian counterparts. These differences have serious ergonomic implications. It is extremely rare to find a mechanical product from the West that is designed specifically for use in the Far East ...(Singh et al., 1995, p. 657). Ong (1995) similarly found that if French anthropometric design standards were applied to the “developing world” countries, the equipment would be suitable only for 15% Thais and 20% Vietnamese, a result very similar to those quoted by Chapanis (1975). It has been suggested by ecological psychologists that such difficulties could be reduced if instead of using lengths in millimeters or inches to specify dimensions in design, a more ecological approach were taken. For example, it is known that the optimal dimensions of a staircase, in terms of risers and runners, scales with the length of a person’s leg, so that the design could be “tuned” to different populations of users to minimize energy consumption. For examples of this approach, see Flach et al. (1995), and Hancock et al. (1995). It is not clear, however, what advantage this would have for design, since the basic units (leg length) against which dimensions would be scaled would still differ markedly between and among the members of different populations, meaning that design would have to be specialized for particular individuals or for the mean of the population, and this could equally well be done using traditional measurements.

POPULATION STEREOTYPES

There are, then, differences between cultures at the level of anatomy and physiology. At the same time there are what seem to be characteristic patterns 38 NEVILLE MORAY

Fig. 2. Population Stereotypes for Switches. (i) North America, (ii) Europe, (iii) Japan. of behavior that are common within large populations. A natural starting point to explore these is the notion of “population stereotypes,” and their contrast with “stimulus-response stereotypes.” The latter, which are particularly strong and apparently intuitive mappings from stimulus to response, are probably determined almost at the level of physiology or anatomy: at least they indicate exceptionally tight coupling between stimulus and response. A simple example is where one of several buttons must be pressed, and which button to press is indicated by its being illuminated. The mapping from stimulus to response seems inherently “natural.”3 A population stereotype by contrast may feel equally compelling to a member of the culture in which it is localized, but may seem completely unnatural to someone from a different culture. Figure 2 shows the classical example of switches. Such stereotypes are often incredible to members of one culture who encounter those of another, but nonetheless they are all arbitrary. They are also extremely important in HMS design. As industrial globalization proceeds apace, and as military equipment is shared among multi-national forces, violation of population stereotypes will become an increasingly potent source of human error, especially where equipment incorporates components constructed by suppliers from different cultural backgrounds. Note particularly that the literature on human reliability indicates that an incorrect stereotype greatly increases the probability of moving a switch in the wrong direction, especially in an emergency where a response is almost automatic and the habitual response dominates performance. Moreover, Swain and Guttman (1983) estimated that when a series of switches is to be activated rapidly, if one is moved in the incorrect direction, the probability that all the rest will be so moved is very greatly increased. A classic investigation of the strength of population stereotypes is that by Smith (1981), who examined the stereotypes of 92 (American) engineers, 85 women (sic) who were the wives, daughters, and secretaries of the engineers, and “55 human factors specialists at the HFS annual meeting in 1977.” His introduction to the study can hardly be bettered:

In many instances, of course, spatial relations provide little guidance. Why do Americans expect that flipping a wall switch upward will turn on a ceiling light? Presumably this is a matter of custom rather than geometry, since in Europe the design convention is reversed. Culture, Context and Performance 39

Which way should one turn a bolt to loosen it? Our culturally acquired expectations are influential as we remove wheel nuts confidently on one side of a car, but are confused by the reversed logic on the other side. Some learned expectations have nothing to do with spatial relations. Why do we use the color red to command stop or signal hazardous conditions? This is presumably a cultural association that has become an accepted design standard. Once established, a design standard need not depend upon association: as a concession to international standards, the Chinese use red stoplights to control traffic, although in their culture red is associated with good fortune rather than danger. To say that a particular compatibility relation is based on a learned expectation, or custom, rather than on spatial congruity, does not necessarily mean that this relation is weak and can be ignored. Expectations based on customary usage can be strongly compelling. How many people could easily break lifelong habits to read a clock whose hands turn counter-clockwise?4 (Smith, 1981, p. 306).

Sakuma and Yaguchi presented a questionnaire based on that of Smith to groups from four nationalities.5 Like Smith they used a paper and pencil test. The questionnaire is shown in Fig. 3, and the results, together with some other data collected by Moray, in Table 2. The numbers of Japanese participants, all employees of Japan Airlines, and drawn from all levels of personnel from blue collar to boardroom, is sufficient to be confident about the stereotypes. The numbers for the other three groups provide only a small sample study (about 20 per group) of several European countries using military personnel. However, while the samples are very small by comparison with the Smith study, the U.S. group results agree well with Smith’s original findings for U.S. participants, with only one or two exceptions. It is not intended here to perform a detailed analysis of all the questions, but to draw attention to three obvious properties of the data. Firstly, there are several items where there are large national differences in the stereotypical expected relations between stimulus and response, and indeed on several items there is a highly significant difference between the national groups on a ␹2 test. Secondly, there are some items where there is no strong stereotype for any culture, suggesting that one cannot rely on any particular response from any population of users. Given the extent of globalization in product marketing these days, the export of “turn-key” technological systems from one country to another, and especially the existence of multi-national military forces sharing equipment, it seems highly probable that errors in operation will be not uncommon when HMSs that embody one population’s strong stereotype are used by a different population from or in a different culture. Many readers will be familiar with the problems of North American or Continental Europeans (who habitually drive on the right-hand side of the road) when required to drive on the left in Great Britain or Japan, and the equivalent problems for the British or Japanese when driving abroad. 40 NEVILLE MORAY

Fig. 3. Questionnaire Based on that of Smith (1981) to Investigate Stimulus – Response Stereotypes. Culture, Context and Performance 41

Fig. 3. (Continued ) 42 NEVILLE MORAY

Table 2. Proportion of Responses of Participants to the Questions in Fig. 3. The Numbers in Each Group were 372 Japanese, 20 US, 15 Netherlands, 24 Greek. The Japanese Participants were Employees of Japan Airlines from All Levels of Personnel. The Other Three Groups were Military Personnel.

Nationalities USA Japan Netherlands Greece

Question 1 Clockwise 0.00 0.09 0.00 0.08 Counter-Clockwise 1.00 0.91 1.00 0.92 Question 2 Raise 0.25 0.58 1.00 0.63 Lower 0.75 0.42 0.00 0.37 Question 3 D A 0.25 0.49 0.28 0.17 CB A B 0.45 0.22 0.43 0.28 CD A B 0.10 0.17 0.29 0.55 DC Other 0.20 0.12 0.00 0.00 Question 4 10, 9,8,7,6 – 1,2,3,4,5 0.21 0.36 0.46 0.67 1,2,3,4,5 – 6,7,8,9,10 0.79 0.49 0.54 0.33 Other 0.00 0.15 0.00 0.00 Question 5 0.10 0.09 0.00 0.08 0.40 0.65 0.85 0.75 0.30 0.22 0.01 0.13 0.20 0.04 0.14 0.04 Question 6 Clockwise 0.95 0.99 1.00 0.92 Anti-clockwise 0.05 0.01 0.00 0.08 Question 7 Left bank 0.20 0.23 0.20 0.21 Right bank 0.80 0.77 0.80 0.79 Question 8 A 0.25 0.10 0.33 0.25 B 0.75 0.90 0.67 0.75 Culture, Context and Performance 43

Table 2. (Continued )

Nationalities USA Japan Netherlands Greece

Question 9 Push 0.47 0.76 0.86 0.71 Pull 0.53 0.24 0.14 0.56 Question 10 0.40 0.58 0.07 0.30 0.07 0.23 0.58 0.13 0.07 0.13 0.21 0.52 0.46 0.06 0.14 0.05 Question 11 RPF 1.00 0.74 1.00 0.78 FPR 0.00 0.21 0.00 0.22 Other 0.00 0.05 0.00 0.00 Question 12 S,1, ..., 6,H 1.00 0.76 0.93 0.79 H,1, ..., 6,S 0.00 0.24 0.07 0.21

Eventually conscious reactions change when living in a new culture. But it seems highly likely, in view of Jost’s Law (Jost, 1897; Woodworth, 1938, p. 58) that the stereotypical habit acquired over many years in one population will from time to time interfere with performance even after much practice, and particularly in an emergency when a rapid “skill-based” reaction (Rasmussen, 1986) is required. There are now several studies of population stereotypes with adequate sample sizes. For example, Courtney (1988) tested over 700 Chinese in Yunan province, and found a tendency for “clockwise for anything” and some violations of Warwick’s Principle (Warwick, 1947) in stereotypes for the relation between direction of motion of controls and displays, using the drawing shown in Fig. 4, although many stereotypes were similar to those in the West. He also studied Hong Kong Chinese, and found a number of stereotypes that were not found with Western subjects (Courtney, 1994). Courtney also investigated Chinese stereotypes for the meaning of colored signals, and found considerable differences from what is common in the West. Among his conclusions were,

In general, color associations for the colors used here did not seem as well established for the Chinese as for the American subjects. The most striking comparison was between the near 100% American responses for the red for stop and green for go associations, and the less than 50% 44 NEVILLE MORAY

Fig. 4. The Drawing Used by Courtney (1988, 1994). The Suggested Motions of Lever, Knob, and Push-Rod were Tested in All 6 Combinations of their Relative Positions.

Chinese responses for these two very important associations ...The results here indicate that the Chinese have broadly similar but not such clearcut color associations as U.S. subjects ...The red for danger in the Western world may not be so clear for the Chinese because red is the color of communism in the People’s Republic of China and for happiness and prosperity for all Chinese people (Courtney, 1986, p. 99).

Hsu and Peng (1993) gave the classical stove-burner problem (Fig. 5) to both Taiwanese Chinese and to Americans, and found a clear difference in the way the two populations allocated controls to the burners. The Americans preferred an anticlockwise labelling,

AD BC

The Chinese an “inverse N” AC BD

Fig. 5. The Task is to Allocate Each Control, A, B, C and D to one of the Burners. Culture, Context and Performance 45

Perhaps the most extreme finding to date is that of Wong and Lyman (1988) who in a study for NASA for the design of the International Space Station found only one configuration that elicited a response stereotype that was common to both American and Japanese culture, even though, as they point out, both cultures are familiar with modern advanced comnplex technological systems! Hoffmann (1997) in another study of direction of motion stereotypes concluded that the findings depended strongly on the group of participants used, even within a single nationality. For example, engineers seemed to have a mental model of direct mechanical coupling between the control and the display, which psychologists did not. Hoffmann warned that pencil and paper tests of the kind used in most of the studies quoted above can give results that are considerably different from those found when physical devices are handled, but nonetheless a conclusion seems clear. There are major differences in population stereotypes across cultures. There may even be significant differences within a culture. For example, to seamen the strongest association of the color “red” is almost certainly not danger, but “port,” (i.e. the left side of a boat when facing the bows), since this is an extremely important association for navigation and pilotage, and one that is encountered far more frequently than red-for-danger. In studies of industrial “safety cultures” it has been suggested that within a single industry or even within a single organization there are cultural differences in the meaning of “safety” and the behavioral consequences of such differences. See, for example, several papers in a special edition of the journal “Work and Stress” (1998) and the work of Rochlin and von Meier (1994) and Bourrier (1996, 1999) who have found great variations in the management and practice of maintenance outages in nuclear power plants even within a country, let alone between countries, in what are very similar installations.

LANGUAGE

One of the most obvious differences between cultures is the difference of language. There are well-analyzed cases where linguistic differences have caused performance problems, for example in the Teneriffe air disaster. But most of these cases have involved situations where one or more of the people involved have had to use a language which is not their native language. (By international agreement, all air traffic control communication must be carried out in English.) It is a more interesting question whether there are characteristics of some languages that may make certain behavior inherently easier or inherently more difficult than is the case in other languages. There is a long-running dispute as to whether different languages embody semantics in sufficiently different ways to prevent things that can be said in one 46 NEVILLE MORAY language from being said, or even thought, in another. As Wittgenstein (1953) once remarked, “To imagine a language is to imagine a way of life.” It is notoriously difficult to translate some words and concepts between languages. An interesting analysis of the way meaning is carried by language is that of Ezra Pound (1935). When discussing how meaning is carried by language in poetry he suggests that in some cases translation may literally be impossible, and it is interesting to consider how his ideas might apply to the design of communications, warnings, etc. in HMSs. Pound distinguished between phanopoeia, melopoeia, and logopoeia. In the first, meaning is carried by images conjured up in the reader’s mind by the words read. In that case, translation is possible, although a literal (word for word) translation may not be the best way. Translation from Western to Chinese is probably usually based on phanopeia. For example, to translate “red” one would use words, or characters, that conjure up equivalent images in the mind’s eye in the two languages. In melopeia meaning is carried by the sounds of the words used, and this can make translation particularly difficult if the phonetic systems of two languages are very different. A French dog “bouya” whereas English dogs “bark.” At least to an English ear the associations of the sounds are different: for example, there is no secondary association with the rind of a tree in the case of French, and the French word has a less urgent sound than that carried by the hard “k” at the end of the English word. In logopeia, a string of words carries a meaning expressed by the content of the text (what we normally think of as the meaning of a sentence). In this case translation will be possible, although again, there may be difficulties that require paraphrasing. (It is notoriously difficult to translate the French “sympatique” or the Italian “simpatico” into English, even with an extensive paraphrase.) See also Osgood et al. (1975). Well known misun- derstandings between English and French include the almost opposite meanings of “experimente”´ (= “experienced”) and “experience” (= an experiment), and the notorious phanopoeic mistranslation of the word “preservative.”6 It follows from such considerations that when technology is transferred from one linguistic culture to another, either operating instructions must be translated, or the receiving culture must learn the language of the transmitting culture. As was mentioned above, we already know that imperfect learning of a language can cause operational difficulties, and that these have led to catastrophic accidents. On the other hand, imperfect translation can equally cause dangerous misunderstandings. English has a particularly rich vocabulary, because its origins lie in a mixture of several languages (Latin, Greek, French, Anglo-Saxon, Norse, etc.). Experiments and surveys have shown that in English the connotations of “start,” “begin,” “initiate,” and “commence,” are not identical, at least in some technical sub-cultures. It may not be easy to capture the nuances of these semantic differences in languages which do not possess so may quasi-synonyms. Indeed Culture, Context and Performance 47 when procedures are written by professional writers who lack technical domain knowledge, there may be problems even within English, for example if there has to be “translation” between, say, American and English English. Perhaps the best known of the latter is the classical example of an Englishman seeing a sign saying, “Vehicles must park on the pavement.” In an English English culture this means what an American would state as, “Vehicles must be parked on the sidewalk.” To show the extent of differences that can arise on the basis of linguistics, consider Question 2 in the Questionnaire for population stereotypes:

Working with a fire crew the hoseman calls “Pressure High!” What should be done? Raise the Pressure or lower the pressure? The responses to the questionnaire for the different cultural groups was as follows:

U.S. Japanese Dutch Greek

Raise 0.25 0.58 1.00 0.63 Lower 0.75 0.42 0.00 0.37

Consider the difference between the American and Japanese group. In the first place, the questionnaire was administered in English to all participants, in the form shown in Fig. 2. It may be thought that a more valid test of the comparability of stereotypes would require the translation of the questionnaire into Japanese for the Japanese group. But that could not be done. The following is a commentary on the question from a Japanese colleague. (If) I heard a man shouting “Pressure High!”, I would imagine he wants the pressure made lower. Many Japanese would adopt this interpretation. Suppose a firefighter shouts “Pressure High!” in Japanese. Literal translation would be “Atsuryoku Takai,” where “Atsuryoku” is a Japanese noun for “pressure,” and “takai” is a Japanese adjective for “high. “Atsuryoku Takai” is a shorter expression of “Atsuryoku ga Takai,” where “ga” is a postpositional particle expressing that Atsuryoku is (the) subject (of the phrase). “Atsuryoku Takai” or “Atsuryoku ga Takai” describes the current status that the pressure is high. If a man shouts like that, we imagine he is not satisfied with the pressure, and thus the pressure should be made lower. If the man needs higher pressure, ...he would shout in Japanese, “Atsuryoku Takaku!” Japanese adjectives change their ending depending on words which follow or words that must be modified. “Atsuryoku Takaku!” is a shorter form of the expression “Atsuryoku wo Takaku shiro!”, where “wo” is a postpositional particle to express that Atsuryoku is (the) object (of the verb), and “shiro” corresponds to “let” or “make”. If we hear a man say “Atsuryoku Takaku!”, we interpret it without any ambiguity (to mean) “Pressure Higher!” if we are allowed to use only two words (Inagaki, 2003). 48 NEVILLE MORAY

So Question 2 of the questionnaire could not be asked in a Japanese translation, since the ambiguity in English would not arise. It seems likely that the lack of a strong stereotype in the Japanese response group arose because of the inability to understand the problem, and that people were simply guessing, giving a 50% split. It is interesting also to ask why the American stereotype is almost exactly the opposite of the Dutch, since the latter speak extremely fluent English. On the other hand, written Japanese can be very ambiguous, and strongly dependent on context to resolve ambiguities. A single Kanji character may have several pronunciations and several meanings. In some cases a character or series of Kanji characters will be superscripted with Hiragana characters to indicate the pronunciation to be used on a particular occurrence of the character. It is not unknown for highly literate Japanese readers to have difficulty in reading formal texts in their own language. Other differences include the fact that difference between singular and plural is not clearly marked graphemically, that there are different words for cardinal numbers depending on the class of thing being counted, etc. There is no reason to think that such differences in linguistic structure need cause particular problems for those native to the culture, whether in everyday intercourse or in interaction with interfaces in HMSs, although a comparative study across cultures would be of great interest. Certainly, the fact that Japanese requires more reliance on context does not imply that there will be more errors made in Japanese HMSs: after all, there have been plenty of errors due to ambigu- ous labeling of written instructions in English. But to the extent that the coupling between spoken and written words and their semantics differs among cultures, we may expect there to be an increase in mistakes (and perhaps slips, Reason, 1990) when people communicate in a non-native language. The errors will be not just in direct mistakes, such as not knowing a particular word, but in a more general sense of not being able to detect nuances of semantics or emotional loading. The extent of such potential problems in HMSs has not been extensively studied, even for bilingual let alone multilingual populations. However there is, for example, no evidence of incidents or accidents caused by switching languages in U.K. nuclear power plants in Wales, where it is common for workers to use Welsh for informal conversation, but English for technical communication, and where management tends to be monoglot English-speaking while many of the work force are bilingual. An interaction between technical design and linguistics is indicated by the work of Wang (1990) and of Fukuda (1990). The former found that to represent Chinese characters on a dot-matrix printer requires at least a 12 × 12 matrix, and preferably a 15 × 15 dot matrix; and that increasing the space between adjacent characters from 2/16 to 5/16 of the width of a character progressivley improved Culture, Context and Performance 49 their legibility. Fukuda found that to handle all three Japanese “alphabets” (Kanji, Hiragana, and Katakana) together with Roman characters required a matrix of 15 × 18 dots. Both of these studies point to a much larger matrix than is required for western (Roman) characters, which can be handled by a 7 × 9 matrix (Cakir et al., 1980). While many Chinese and Japanese characters would be legible in a standard Western matrix, there would also be many that were not. A specific problem in linguistic and symbolic communication has arisen very recently. The International Atomic Eneergy Authority is trying to find a set of symbols indicating hazards due to nuclear energy which will be immediately intelligible to people of all cultures and all ages, at all times, and without instruction, even when reduced to an icon of less than 2 cm by 2 cm. The aim is to protect people who may pick up materials contaminated by radiation that are found lying about the environment (for example, used equipment or materials from nuclear medicine facilities that are accidentally lost). The symbol must not be confusable with existing synmbols for other hazards such as explosives, poison, etc. The meaning of the existing symbol, three yellow “petals” or “blades” or “rays” on a black ground, which has been used for decades, is well-known to be less than obvious. (The most intuitive interpretation is not that there is a radiation hazard, but that the object is, or contains, a fan or propeller.) From what has been said above, even though we have not considered the semantics of icons as distinct from linguistic representations, it is safe to say that the search is doomed to failure. Since alpha, beta and gamma radiation and their emission is invisible, and since few people have seen even evidence of their existence, it is certain that there is no universal symbol that is intuitively meaningful to all people without instruction.

TECHNOLOGY TRANSFER

So far we have looked at reasons to think that there will be differences in human performance when people from one culture are faced with equipment (displays, controls, instructions) designed for use by those from a different culture. Let us now consider the larger picture. Such confrontations often arise because of technology transfer, and such transfer is almost always from advanced industrialized countries (usually in the West) to countries undergoing industrial development in other parts of the world (although some countries, such as Japan, South Korea, and increasingly China) have already reached the point where they in their turn are exporting technologies. We may expect there to be difficulties in using such technologies, even when they are supplied as “turn key” systems. These difficulties will arise because of 50 NEVILLE MORAY mis-matches between users due to anthropometric dimensions, stimulus-response stereotypes, and misinterpretation of linguistic (or even iconic) operating instructions. But there are also more complex factors, to do with the expectations that different cultures have about how life should be lived, the relations between people of different rank in a social or official hierarchy, whether motivation is based on money, competition or co-operation, and so on. For an example of how such factors operate even in industrialized countries, see Zuboff (1988); for a superb fictional account, see Ireland (1970). High technology is almost always capital intensive and labor sparse, but as Ong (1991) has pointed out, and as is obvious to anyone who travels in industrially developing countries (IDCs), such technology is often ill matched to an IDC. In technology transfer to IDCs it is often necessary to maximize the use of labour rather than design for low manpower high technology, in order to avoid unemployment, and resulting social instability. Ong emphasises the importance of socio-cultural attitudes to work:

...for example, the relationship that exists between control and response and display and control compatibility, etc., affect the efficiency and safety of human-machine systems, and are population-dependent variables (Ong, 1991, p. 804).

Fukuda (1990) reports that in China half of industrial accidents occur because of violations of procedures, that is, the unwillingness of operators to follow the procedures provided (Reason, 1990, 1997). It is known there are cases where highly skilled and experienced operators find that it is both more efficient and safer sometimes to violate procedures (Mearns et al., 1998). But it is highly unlikely that a violation rate of 50% is due to half the operating procedures being impossible to follow. More likely are other factors arising in the social psychology of the workplace. Shanavaz (1996) notes that the rate of accidents and injuries at work in IDCs is more than 10 times that of the fully industrialized countries who export major technology:

It is generally agreed that the characteristics of the technology-recipient country are decisive for a successful technology transfer. These are: government policy, technical absorptive capacity, and cultural distance ...On the question of cultural distance, the national culture of the technology recipient country and its similarities with the technology-producing country can greatly influence the success of the transfer of technology. Cultural values, social organizations and behavioral patterns have a direct bearing upon a society’s willingness and ability to adapt and absorb technology ...A successful transfer of technology therefore requires that the cultural issues are recognized and overcome, and that cultural issues are considered in the process of technology transfer (Shanavaz, 1996, p. 1399).

Similarly, Simard and Marchand (1997) suggest that micro-organizational factors are the primary determinants of safety compliance behavior, with social Culture, Context and Performance 51 relationships at the shop floor level being the best predictors. Their work also suggests that the fabric of these social relationships can be substantially influenced by managerial actions in developing a participative approach to the management of safety, but as we shall see, this may be very difficult in some cultures. Wisner (1995) points out that the mental models and expectations of designers concerning the characteristics, education, psychology, etc., of the target population may be wildly incorrect, often amounting to an attempt to force the adoption of a Western viewpoint on the society and workforce, “in order to avoid consideration of socio-cultural characteristics.” The importance of social and organizational factors, and hence of culture, is emphasised by the work of Rochlin and his colleagues on Highly Reliable Organizations (HROs). They have studied hazardous HMSs that seem to have a far lower frequency of accidents than one would expect, and have identified the following characteristics in such systems.

HROs consciously aim for reliability and acceptance by society, emphasising a “safety culture.”
Their personnel have strong technological and social relationships, and have well trained experts who communicate freely and effectively with one another, and who at all times understand system state.
The organization promotes activity to enhance reliability, with easy access by lower grades of personnel to senior levels of management to report problems.
HROs have a great capacity to learn from records of past accidents and errors, constantly seek to improve their performance, and have a flexible organisational style. Many of their personnel are able to perform more than one task, and the teams show adaptive self-organization in the face of unforeseen demands during hazardous incidents.
There is a sense of collegiality and shared responsibility, and this allows decisions to be made without a need to refer to a central authority. That is, in the face of hazardous events the organization is not strongly hierarchical, but informally devolves authority to those on the spot who are temporarily the best placed to deal with the problem.
HROs have people whose job is to seek out and report problems and signs of weakness. If there are long periods where there are no incidents or near accidents personnel begin to worry that their standards of safety have been lowered, rather than congratulating themselves that the problems have been overcome.
A very important characteristic is that when people have a “near miss,” make an error, or cause an incident and report it themselves, they are rewarded and the 52 NEVILLE MORAY

information is used for “lessons learned” which are transmitted to upper levels of management. Such people are praised, not blamed (LaPorte, 1996; Rochlin, 1999; Rochlin et al., 1987). These studies were all of American organizations. Do they reveal characteristics of HROs worldwide? Or to put the question another way, can other cultures be expected to show the same characteristics in the search for high reliability and safety? Notice that these studies make very little explicit reference to the engineering designs of the system. What is emphasised is the character of social and organiza- tional interaction among the personnel, and the effectiveness of communication within and between levels of groups of workers and management. We saw earlier that several researchers on technology transfer emphasised the cultural patterns of behavior found in different countries. Are cultural patterns sufficiently different to cause inherent difficulties in the human management and operation of socio-technical systems? Consider the lists of “national characteristics” in Table 3. They are taken from commercial briefings for business people planning to do business in Russia and Japan. To some extent all such “national stereotypes” are of course a parody of what people of those countries are “really” like, as would be a similar list describing the “typical” American. But is the description of an HRO perhaps an approximation to (at least certain type of) the latter? If so, it seems very unlikely that we will

Table 3. National Characteristics as Portrayed in Commercial Briefing Documents.

“Russian” characteristics 1. Risk averse 2. Subordinates will not talk to boss about a problem 3. Avoid initiative because it will be punished by increase of work 4. Avoid sharing information because knowledge is power 5. Do not admit not knowing the answer to a problem 6. Work around rules when these are too demanding, using personal contacts and influence. “Japanese” characteristics 1. Avoid direct confrontation during discussions 2. When threats occur withdraw and consider the meaning of the threat 3. Everyone must have a chance to voice their opinion 4. Decision making is by consensus 5. Much forethought and discussion precedes decisions 6. Decision is “middle-up-and-down” not “top-down” 7. Conclusions are reached with reluctance. Culture, Context and Performance 53

find similar HROs in either Japanese or Russian cultures, for the social dynamics described by Rochlin et al. seem unlikely to occur among a group having the characteristics listed in Table 3. It may be that each culture will need to develop a style of HRO suited to its national characteristics, and even that some will be unable to do so.

THE SOCIAL SCIENCE OF NATIONAL CHARACTERISTICS

To develop a particular organizational policy in pursuit of safety, for example, requires incentives to change behavior. Looking at the literature on ergonomics and human factors, there is almost a complete absence of discussion of this factor. Paradoxically human factors practitioners seem to discuss almost exclusively technical solutions (albeit behavioral) to problems of risk, fault and error. While castigating engineers for too hardware-orientated an approach to reliability, ergonomists themselves are perhaps also guilty, in so far as they concentrate on human-centered design, but interpret that all too often merely in terms of better interfaces, or better training regimes. The social sciences, economics, sociology, anthropology and so on are largely neglected, at least in British and North American human factors, except for some of the work in macroergonomics and the relatively few uses of ethnomethodology. For example, I can recall no published papers where a solution based on economics, rather than ergonomics, has even been suggested, let alone discussed in detail. Certainly one sees papers in which it is pointed out that changing pay-offs will shape behavior. But where, in human factors literature, are there detailed discussions of how to shape behavior by incentives, based on detailed cost-benefit analyses, and related to the psychology of wage structures, work patterns, manning levels, etc.? While traditional ergonomic solutions are perhaps relevant to the inner levels of Fig. 1, at outer levels economic policy is almost certainly a far more potent method of shaping policy and behavior. Ergonomics must take economics seriously. There are reasons to believe that few results of ergonomics research are universally applicable. The great importance of social and cultural characteristics is emphasised by the work of Hofstede (1984, 1994) who studied the attitudes and values of employees in a single organization, IBM, that has offices world-wide. The advantage was that to a considerable extent the people employed by IBM had common levels of education, and worked in an organization that has common standards, incentive schemes, career paths, managerial policies, etc., throughout the world. The study comes close to leaving national characteristics as the major factor accounting for differences between the groups studied. Hofstede reported 54 NEVILLE MORAY several major dimensions on which people performing the same kind of work but coming from different cultures appeared to differ. The three most important dimensions are Power Distance, individual- ism/collectivism, and Uncertainty Avoidance. He defined the Power Distance Index (PDI) by the fact that people who show a high PDI and were located low in an organization feel remote from those higher in authority. Those with high uncertainty avoidance index (UAI) are intolerant of uncertainty, and show characteristics reminiscent of Adorno’s classical “Authoritarian Personality,” both desiring to exercise power and desiring to have it exercised over them rather than to accept variability in performance and freedom to make decisions outwith rules. Moray summarized Hofstede’s results as follows:

PDI scores are higher in jobs requiring lower levels of skill, and the national differences were less among lower skilled workers. Locations with PDI scores in the top third of the ranked order included Malaysia, Philippines, India, Singapore, France and Hong Kong. Those in the middle third included South Korea, Taiwan, and Japan. Those with the lowest scores included the USA, Canada, Netherlands, Scandinavian countries and the UK. The USA has a moderate to low PDI score (40 on a scale of 0 – 100), and typically has a management style in which there is considerable collaboration of employees with management, but largely at the discretion of management. In Scandinavian cultures equality and consensual decision making is largely enshrined in legislation as well as in psychological attitudes. In Greece, with a PDI of 60 on a 100 point scale, Hofstede reports typical attitudes as being, “If he is the manager, why should he want to consult my opinion – he should know what to do” (Moray, 2000, p. 864).

If we compare such attitudes with those found in HROs by Rochlin, LaPorte, and Roberts (vid. sup.) there is a poor match. For example, a willingness to report one’s own errors, and the presence of strong communication links between the workforce and management would be very hard to maintain in cultures with high PDIs, where the workforce perceives itself as very distant (alienated?) from management. Moray’s summary of Hofstede continues,

Countries with high Individualist scores on the ICI include the USA, UK, Canada, Netherlands, the Scandinavian countries, France and Germany. Those with intermediate scores include Japan, India, Brazil and Greece. Those with low scores include Singapore, Hong Kong, Malaysia, South Korea, Taiwan and Pakistan. ...In collectivist cultures much can be taken for granted – in individualist cultures communication must be explicit: For example American commercial contracts are much more detailed than Japanese ones (Moray, 2000, p. 864).

In Hofstede’s terms the HROs studied by Rochlin and his colleagues had a strongly “collectivist” attitude. The willingness to take on several jobs as the situation called for it, the effort to present a unified highly efficient image to people outside the organization, the willingness to learn from one another, and Culture, Context and Performance 55 the characteristic ability to be self-organizing in response to disturbances – all these are collectivist characteristics. (The strange thing is that these HROs were, of course, American, and hence created in a culture that Hofstede found to be highly individualistic – but then of course the point is that these HROs were exceptional.) A particularly interesting study quoted by Hofstede compared American workers and Chinese workers performing the same task, namely to manufacture baskets, to the same productivity target, 200 baskets, and with the same number of workers, 20. It was found that Chinese workers worked best when the goal was a collective one (200 baskets to be made by the team of 20 people as a whole). American workers worked best when the 20 workers each had to make 10 baskets.

The uncertainty avoidance index (UAI) measures the extent to which members of a culture feel threatened by uncertain or unknown situations. High scores were found for Japan, South Korea, and France; intermediate scores for Norway, Canada, USA, Finland and Germany; and low scores for UK, Hong Kong, Sweden and Denmark. ...Countries with a high UAI tend towards a proliferation of rules and regulations. There is a relation with PDI, since if PDI is large, the power of superiors can substitute for explicit rules. In high UAI countries even ineffective rules can satisfy the psychological need – e.g. in France under the ancien r´egime there was a saying, ‘un regle rigide, une pratique molle’ (Moray, 2000, p. 864).

Again, it seems that the personnel in the HROs would not feel threatened by uncertainty. They have a firm belief in their expertise, and are willing to use self-organization in the face of official rules to ensure that their expertise is put to best use. Self-organization implies a tolerance of uncertainty, since the formal hierarchy of authority may be violated to a very considerable extent so that individual expertise may come to the fore for brief periods. In the light of Hofstede’s work consider the so-called national stereotypes listed in Table 3. Clearly a culture which inhibits the transfer of power, which discourages initiative, or inhibits communication between different ranks in a hierarchy will not be able to develop HROs which depend on the characteristics which were observed by Rochlin, Roberts, La Porte, Bourrier and others. Note that in taking cultural differences into account we are looking at the outermost layers of Fig. 1, and that we find that the factors located there place just as strong constraints on system performance as those due to the innermost levels that have traditionally been seen as the proper concern of ergonomics. This is not just an academic issue. For example, both the sinking of the “Herald of Free Enterprise” and the disaster at Chernobyl were largely “outer level” accidents. It seems that in a culture where it was more acceptable to challenge the commands of management Chernobyl at least would not have happened (although it would be rash to claim that it could not happen in a culture which accepted more challenge to authority). 56 NEVILLE MORAY

It is not known to what extent these national characteristic discovered by Hofstede are stable over time, nor what are their social dynamics,. For example, Wisner remarks: Social distances that are increased by substantial wage differences and sometimes class and political differences may create conflict situations in the company that further complicate the comprehension of and the solution to technical difficulties (Wisner, 1995).

CONCLUSION

Figure 1 shows that many different kinds of constraints operate to determine the quality of performance in HMIs. The data reviewed in this paper show that those factors in their turn are all affected to some extent by social and cultural forces. Furthermore, such cultural factors are not restricted to large complex systems, but can have important impacts on the efficiency even of simple things such as the use of individual hand tools. The supposedly universal psychological constants and laws, reaction times, transfer functions, mathematically optimal sampling rates, etc., are best regarded as boundaries, not as constants, and the effects of culture can make it hard for people to reach, or even to come close to those boundaries. We may indeed ask whether in the pursuit of efficiency we should demand that people always work “on the margin.” There is something to be said for maximizing profit and efficiency. There is also something to be said for the 2-hour lunch break with a bottle of wine that characterizes many French workplaces. The aim of ergonomics, it is said, is to improve the lot of the worker and the user of equipment and products. The solutions may be found, in cross-cultural studies, to lie in regions unknown to the (scientific and technological) philosophy of traditional ergonomics and human factors.

NOTES

1. Throughout this paper the words “ergonomics” and “human factors” will be held to have the same meaning and will be used interchangeably. 2. “Shared values (what is important) and beliefs ...that interact with an organisation’s structures and control systems to produce behavioural norms the way we do things around here.” Reason, (1998, p. 298). Substitute “Society” for “organisation” and change the last phrase to “the way we live around here.” 3. Note that there will, even here, be non-industrialized cultures where the notion of turning something on or off, and the meaning of an illuminated button, will be unknown. 4. The present writer owns such an “anti-clock” which resides permanently on display in his kitchen, and the response of visitors is interesting. Almost all feel that there is something wrong about it, but frequently it takes a considerable time for them to realize what is the matter, even though, geometrically, the clock would appear to be up Culture, Context and Performance 57 to 6 hours out if you did not notice the digits printed on the face. The writer can now read the time – after several years practice – in about twice the time required to read a conventional clock. 5. The questionnaire was modified from Smith’s version by Prof. P. M. Sanderson of University of Queensland, and provided to Yaguchi and Sakuma by the present writer. 6. Which in English means “a substance which prevents decomposition of food or wood,” and in French means “condom.”

ACKNOWLEDGMENTS

My thanks are due to Prof. T. Inagaki of Tsukuba University, Japan for his discussion of the linguistics of the “hoseman’s problem”; and to Sakuma-san and Yaguchi-san of Japan Airlines for their invaluable assistance in administering the stereotype questionnaire to the personnel of Japan Airlines.

REFERENCES

Cakir, A., Hart, D. J., & Stewart, T. F. M. (1980). Visual display terminals. New York: Wiley. Chapanis, A. (1974). National and cultural variables in ergonomics. Ergonomics, 17, 153–175. Chapanis, A. (1975). Ethnic variables in human factors engineering. Baltimore: Johns Hopkins University Press. Clement, W., Jex, H., & Graham, D. (1968). A manual control theory applied to instrument landings of a jet transport. IEEE Transactions on Man-machine Systems, MMS-9, 93–110. Courtney, A. (1994). Hong Kong Chinese direction-of-motion stereotypes. Ergonomics, 37(3), 417–426. Courtney, A. J. (1986). Chinese population stereotypes: Color associations. Human Factors, 28(1), 97–99. Courtney, A. K. (1988). Chinese response preferences for display-control relationships. Human Factors, 30(3), 367–372. Fernandez, J. E., Malzahn, D. E., Eyada, O. K., & Kim, C. H. (1989). Anthropmetry of Korean female industrial workers. Ergonomics, 32(5), 491–495. Flach, J., Hancock, P., Caird, J., & Vicente, K. (1995). Global perspectives on the ecology of human-machine systems (Vol. 1). Hillsdale, NJ: Lawrence Erlbaum. Fukuda, T. (1990). New electronic media and the human interface. Ergonomics, 33(6), 687–706. Hancock, P., Flach, J., Caird, J., & Vicente, K. (1995). Global perspectives on the ecology of human-machine systems (Vol. 2). Hillsdale, NJ: Lawrence Erlbaum. Hoffmann, E. R. (1997). Strength of component principles determining direction of turn stereotypes – linear displays with rotary controls. Ergonomics, 40(2), 199–222. Hofstede, G. (1984). Culture’s consequences. Newbury Park, CA: Sage. Hofstede, G. (1994). Cultures and organizations. London: HarperCollins. Hollnagel, E. (1998). Cognitive reliability and error analysis method. Amsterdam: Elsevier. Hsu, S.-H., & Peng, Y. (1993). Control/display relationship of the four-burner stove: A reexamination. Human Factors, 35(4), 745–749. 58 NEVILLE MORAY

Ireland, D. (1970). The unknown industrial prisoner. Melbourne: Angus & Robertson. Jones, R. E., Milton, J. L., & Fitts, P. M. (1949). Eye fixations of aircraft pilots, I. a review of prior eye-movement studies and a description of a technique for recording the frequency, duration and sequences of eye-fixations during instrument flight. USAF Technical Report No. 5837. Wright Patterson Air Force Base, Dayton Ohio. Jost, A. (Title not available). Zeitschrift f¨urpsychologie, 14, 436–472. Kleinman, D. L., Baron, S., & Levison, W. H. (1970). An optimal control model of human response. Part 1: Theory and validation. Automatica, 6, 357–369. LaPorte, T. R. (1996). High reliability organizations: Unlikely, demanding, and at risk. Journal of Contingencies and Crisis Management, 4(2), 60–71. McRuer, D. T., & Krendel, E. (1976). Mathematical models of human pilot behavior.NATO AGARDograph No. 188. Brussells. Mearns, K., Flin, R., Gordon, R., & Fleming, M. (1998). Measuring safety climate on offshore installations. Work and Stress, 12(3), 238–254. Moray, N. (2000). Culture, politics and ergonomics. Ergonomics, 43(7), 858–868. Moray, N., & Inagaki, T. (2000, 2001). Attention and complacency. Theoretical Issues in Ergonomic Science, 1(4), 354–365. Moray, N., Neil G., & Brophy, C. (1983). Selection and behavior of fighter controllers. Contract Report, Ministry of Defence, London. Moray, N., Richards, M., & Low, J. (1980). The behavior of fighter controllers. Contract Report, Ministry of Defence, London. Moray, N., Synnock, G., & Sims, A. (1973). Tracking a static display. IEEE Transactions on Systems, Man and Cybernetics, SMC-2, 518–521. Okunribido, O. O. (2000). A survey of hand anthropometry of female rural farm workers in Ibadan, Western Nigeria. Ergonomics, 43(2), 282–292. Ong, C.-N. (1991). Ergonomics, technology transfer, and developing countries. Ergonomics, 34(6), 799–814. Osgood, C. E., May, W. H., & Miron, M. S. (1975). Cross-cultural universals of affective meaning. Chicago: University of Illinois Press. Pound, E. (1935). The ABC of reading. London: Routledge & Sons. Rasmussen, J. (1986). Information processing and human-machine interaction: An approach to cognitive engineering. Amsterdam: North-Holland. Reason, J. (1990). Human error. Cambridge: Cambridge University Press. Reason, J. (1997). Managing risks of organizational accidents. Ashgate Gower. Reason, J. (1998). Achieving a safe culture: Theory and practice. Work and Stress, 12(3), 293–305. Rochlin, E., LaPorte, T., & Roberts, K. (1987). The self-designing high reliability organisation: Aircraft flight operation at sea. Naval War College Review (Autumn), 76–91. Rochlin, G. I. (1999). The social construction of safety. In: J. Misumi, B. Wilpert & R. Miller (Eds), Nuclear Safety: A Human Factors Perspective (pp. 5–23). Senders, J. W., Elkind, J. I., Grignetti, M. C., & Smallwood, R. (1965). An investigation of the visual sampling behavior of human observers. Technical Report NASA-3860. Cambridge, MA: Bolt, Beranek & Newman. Senders, J. W. (1983). Visual sampling processes. Tilburg, The Netherlands: Katholieke Hogeschool Press. Shanavaz, H. (1996). Making ergonomics a world-wide concept. Ergonomics, 39(12), 1391–1402. Shannon, C., & Weaver, W. (1947). The mathematical theory of communication. Urbana: University of Illinois Press. Culture, Context and Performance 59

Simard, M., & Marchand, A. (1997). Workgroup’s propensity to comply with safety rules: The influence of micro-macro organizational factors. Ergonomics, 40(2), 172–188. Simon, H. A. (1992). What is an ‘explanation’ of behavior? Psychological Science, 3, 150–161. Singh, J., Peng, C. M., Lim, M. K., & Ong, C. N. (1995). An anthropometric study of Singapore candidate aviators. Ergonomics, 38(4), 651–658. Singleton, W. T. (1997). A. T. Welford – A commemorative review. Ergonomics, 40(2), 125–140. Smith, S. L. (1981). Exploring compatibility with words and pictures. Human Factors, 23(3), 305–315. Swain, A. D., & Guttmann, H. E. (1983). Handbook of human reliability analysis with emphasis on nuclear power plant applications. NUREG/CR-1278. Washington, DC: U.S. Nuclear Regulatory Commission. Swets, J. A. (1996). Signal detection theory and ROC analysis in psychology and diagnostics. Mahwah, NJ: Lawrence Erlbaum. Wang, Z. M. (1990). Recent developments in ergonomics in China. Ergonomics, 33(7), 853–865. Warwick, M. J. (1947). Direction of movement in the use of control knobs to position visual indicators. In: P. M. Fitts (Ed.), Psychological Research on Equipment Design (pp. 137–146). Army Air Force Aviation Psychology Program Research Department. Welford, A. (1951). Skill and age: An experimental approach. London: Oxford University Press. Wisner, A. (1995). Situated cognition and action: Implications for work analysis and anthropotech- nology. Ergonomics, 38(8), 1542–1557. Wittgenstein, L. (1953). Philosophical investigations. Oxford: Blackwell. Woods, D. D., Johannsen, L. J., Cook, R. I., & Sarter, N. B. (1994). Behind human error: Cognitive systems, computers and hindsight. Columbus, OH: CSERIAC. Woodworth, R. (1938). Experimental psychology. New York: Henry Holt. Work and Stress (1998). Special edition on safety culture (Vol. 12). Young, L. R. (1973). Human control capabilities. In: J. F. Parker & V. R. West (Eds), Bioastonautics Data Book (pp. 751–806). Washington, DC: National Aeronautics and Space Administration. Zuboff, S. (1988). In the age of the smart machine. New York: Basic Books. This Page Intentionally Left Blank 3. THE DEVELOPMENT OF CULTURE-ORIENTED HUMAN MACHINE SYSTEMS: SPECIFICATION, ANALYSIS AND INTEGRATION OF RELEVANT INTERCULTURAL VARIABLES

Kerstin Rose¨

ABSTRACT

This chapter is focused on the specification and integration of intercultural variables for human machine systems and the description of content analysis for these variables. Starting with basics of culture-oriented design, these are followed by an approach to machine localization issues and a cost model, then basics of the intercultural design and human machine system engineering process, a definition and specification of intercultural variables, a systematic treatment for their integration into the process, and a description of how to use these variables in the process. Finally, an example of an intercultural variables approach to “information coding” in a human-machine system is presented for China and Germany.

Cultural Ergonomics Advances in Human Performance and Cognitive Engineering Research, Volume 4, 61–103 Copyright © 2004 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1479-3601/doi:10.1016/S1479-3601(03)04003-7 61 62 KERSTIN ROSE¨

CULTURE AND GLOBALISATION

With the globalization process, there is a need for communication that goes beyond the borders of countries and cultures. The global interaction between different cultures involves sharing the values of all interaction partners and sometimes, they define their own communication culture to interact (Honold, 2000). Bourges-Waldegg (2000) says: “...Design changes culture and at the same time is shaped by it. In the same way, globalization is a social phenomenon both influencing and influenced by design, and therefore, by culture ..., both globalization and technology have an effect on culture, and play a role in shaping them.” It is therefore important to know more about the similarities and differences in culture-influenced user philosophies, e.g. of European and Asian culture. Some well-known and established theories in the field of cultural differences have been introduced by Hall, Hofsteede, Victor and Trompenaars (see Hoft, 1995). Based on these theories, many studies have been conducted on various special aspects of cultural differences, e.g. icon coding, recognition and usage, and colour coding and usage (studies by Choong & Salvendy, 1998; Piamonte & Ohlsson, 1999; Prabhu & Harel, 1999; Rose&Z¨ uhlke,¨ 1999; Zuhlke¨ et al., 1998). The results of the studies show that cultural influences need to be considered above all in the area of information coding. The use of colors and pictorial symbols are only examples of this. The study results show that there are the following challenges for the future:
An increasing globalization requires a concrete consideration of other cultures.
The exploration of new markets is required in the area of ergonomic design of human-machine-systems, also the consideration of culturally dependent user habits.
Global markets are not automatically associated with global design. Often a local design is necessary to obtain user-friendliness and to avoid operating mistakes at an early stage. One could say that in times of globalization, user-oriented design also means culture-oriented design (see also Fig. 2). Are we fit for the challenges of the future? What do we know, and what do we not know? How can we support culturally-oriented design in practice?

CULTURALLY-ORIENTED DESIGN

Culture as a User Variable

A useful way of thinking about product design is that it should be a message from the designer to the user. The designer has to be able to convey to the user the functionality of the system in a self-explanatory way. The Development of Culture-Oriented Human Machine Systems 63

If we talk nowadays about user-oriented design, we must realize that in times of globalization, culture-orientation is one essential component for successful user friendliness and self-explanatory functions and design. Culture is both an influence factor on user-interface design and also an element of user experiences. Culture exercises an influencing factor on the user-interface design process. It shows an effect on the designer and his theories, his thinking and working style, and the design process itself; because each culture has its own style for product development, and the design result is influenced from the culture-specific aim of a development process (de Souza & Dejean, 2000). In an age of globalization, culture orientation is one essential component for successful user-centered design. Therefore, culture has the same importance as other factors such as the user’s profession, choice of operating systems, learning style and other elements. The following model of mono- vs. multi-cultural user, based on Bourges-Waldegg and Scrivener (1998), is a good illustration (see Fig. 1). Figure 1 shows the example of a mono-cultural user. A mono-cultural user stays in his primary culture and has no experiences of other cultures. He is familiar with a single culture and its special representational meanings. His interpretations are based on representational forms known to him. If such a mono-cultural user substitutes a representation form from his culture for a certain meaning, he may not be able to find the correct interpretation. He is only familiar with one set of meaning-representation-interpretations, where as a multi-cultural user knows his primary culture experiences and other secondary cultures. He knows more than one culture and their relevant representational meanings. He is able to find an interpretation for the meaning which is based on more than one well-known representation form. His interpretations are based on a number of different sets of meaning-representation-interpretations.

Approach for Culture-Oriented Design

With the globalization process, there is need for communication that goes beyond the borders of countries and cultures. The global interaction between different cultures involves sharing the values of both interaction partners. The key problem of intercultural design is how the designer can get his message across to the user of another culture. This is not simply a question of language. The most important fact is that the designer and the users of different cultures agree on the information meaning and its interpretation. There must be a significant element of shared meaning between the user and the developer. The importance of cultural differences as opposed to just language factors is now widely recognized. The area of culture-oriented design deals with the challenge of intercultural aspects of product design. There are two established approaches for intercultural 64 KERSTIN ROSE¨ Mono vs. Multicultural User. Fig. 1. The Development of Culture-Oriented Human Machine Systems 65 design: internationalization and localization. Internationalization describes a basic structure with the consideration of future integration of culture-specific requirements. This design concept takes into account some general culture- specifics (format, language etc.) and is often designed for flexible switching between user cultures. Localization focuses on one specific user culture. The aim of this design concept is the best fitting product for a specific user culture. In application areas like the aerospace industry and sometimes the car industry, a third approach to intercultural design, global design (one for all), should be required. But in areas of high individual user orientation, an international or local design is necessary. Another important aspect of culture-oriented design is cultural models. These models are descriptive artifacts, which enable the separation of cultures, with regard to specific cultural dimensions.

Culture Models and Their Application for User-Interface Design

To enable communication between individuals, certain rules have to be followed. Interpersonal communication is based on cultural and social influences, and has developed over millennia into many regional communication forms. In the same way, human-computer(machine)-communication needs defined rules so that the information displayed by the machine expresses the current state of the process in an understandable way. In previous research, we have classified user factors into the following types:
objective factors e.g. gender, age, ethnic background, mother-tongue, these can all be objectively identified for each user subject and used to categorize different user groups,
subjective factors which cannot be directly measured or identified e.g. values, beliefs and rituals that lead to patterns of behavior. Some well-known and established theories in the field of cultural differences have been introduced by Hofstede (1997), Trompenaars (1998) and Hall (1989). All models organize human behavior into cultural dimensions. The dimensions are called cultural factors and illustrate the dichotomies of user behavior. The investigated human behavior (human-human-interaction) of each study decided the cultural factor. Geert Hofstede: determining patterns that form a culture’s mental program- ming. Hofstede’s cultural factors: Individualism/Collectivism, Power Distance, Masculinity/Feminity, Uncertainty Avoidance, Long-Term/Short-Term. 66 KERSTIN ROSE¨

Fons Trompenaars: determining the way in which a group of people solves problems. Trompenaars’ cultural factors: Universalism/Particularism, Individual- ism/Collectivism, Neutral/Emotional, Specific/Diffuse, Achievement/Ascription. Edward T. Hall: determining the right response rather than sending the right message. Hall’s cultural factors: Context – High/Low, Message – Fast/Slow, Time – Polychronic/Monochronic, Information flow. All factors describe behaviors of members of a specific culture. The cultures are assignable between the poles of each dichotomy, the intensity that is shown depending on behavior in this culture. The study of Hofstede is unique in its dimensions. He has evaluated 116.000 questionnaires of IBM-employees in 50 countries and more than 20 languages. The results of a factor analysis for 32 presented statements show five dimensions: individuality, power distance, gender, uncertainty avoidance and time orientation. Results of the evaluation can be found in Hofstede (1997) “Culture and Orga- nization.” His study is used very frequently to interpret cultural differences of interpersonal and personal-technical communication. The application of Hofstede’s cultural dimensions to user interfaces is not simple. At first, a description of user interface features is needed as a descrip- tive artifact to separate between the features. These interface factors, called “characteristic user interface factors” are shown in Table 1. These five factors are the essential descriptive features for a user interface. The combination of Marcus’ characteristic factors and Hofstede’s culture dimensions makes a new kind of cultural description for user interfaces possible (comp. Marcus, 2001a). The result of the combination is called: cultural user interface factors. These factors are the focus of the following section. Based on Hofstede’s cultural dimensions the following cultural user interface (UI) factors are known: cultural UI-factor “Power distance,” cultural UI-factor “Individuality,” cultural UI-factor “Gender,” cultural UI-factor “Uncertainty avoidance” and the cultural UI-factor “Time orientation.” An example of the UI-factor “Individuality” is shown in Table 2. The cultural user interface factor is oriented on Hofstede’s dimension with the two poles “individualism” and “collectivism.” In an individualistic culture, the

Table 1. Characteristic UI Factors (Marcus, 2001).

Metaphor Basic concepts of words, pictures, sounds, etc. Mental model Organization of data, functions, tasks and rules Navigation Movements trough the system, dialogs, buttons and the linking Interaction Feedback, input-/output-technique Presentation Visual, verbal, acoustic and tactile information presentation The Development of Culture-Oriented Human Machine Systems 67

Table 2. Cultural UI-Factor “Individuality.”

Individualism Collectivism

Metaphor Action-oriented Relationship-oriented Mental Model Task- and product-oriented Rule-oriented Navigation Rule-independent path Rule-dependent path Individualistic path Group-specific path Arbitrary choices Limited choices Interaction Multiple devices Official devices Keyword search Restricted choices User-specific adaptability Role-specific adaptability Presentation Figures of products and people Figures of groups and organizations Low context-orientation High context-orientation Countries France, Germany, USA, Italy, Costa Rica, Korea, Mexico, Malaysia, Israel, Finland Brazil, Japan individuality of a person is the most important and promotable skill. The society is promoting the definition of a person as individual. In a collectivistic culture, the status of a person, based on his group membership, is very important. The rule definition within a group is the basis for the incorporation of a person into the society. Most activities of culture-oriented design are recorded in the area of web design. Therefore Web Pages as seen in Fig. 2 are shown as an example. Both are from the company Siemens. On the left side is an example for an individualistic culture: task-oriented, Keyword-search, multiple choices for navigation and action-oriented design. This page supports the individualized usage. The right side of Fig. 2 shows the other way of design, an example for a collectivistic culture like Malaysia. It shows a high context-orientation, rule-orientation and restricted choices. This design offers the same manner of usage for each visitor to this page. Both examples for USA and Malaysia show only some features of the cultural user interface factor “Individuality.” There is a wide range of possibilities between the poles “individualism” and “collectivism.” Each culture fits between these poles, but it also means that these cultures are not 100% “individualistic” or “collectivistic.” The assignment of a culture to one pole is only an orientation for the designer. It does not replace the empirical analysis of the specific culture- dependent user requirement in the focused market for the designer. The cultural user interface factors are a good help to describe cultural differences for user interfaces. With that, an artifact for the description of culturally-based differences in the area of user-interface is found. Therefore, 68 KERSTIN ROSE¨ Individualistic Culture, USA (Left); Collectivistic Culture, Malaysia (Right). Fig. 2. The Development of Culture-Oriented Human Machine Systems 69 these factors are a good element for description, explanation and interpretation of cultural differences in user-interface design. But these factors do not contain a design guideline on “how to design for a specific culture.” Developers of culture-oriented products need more information and better support. At the moment, each approach of user-oriented design shows that “user requirements” are the key to successful and ergonomic design. That is also valid for culture-oriented design. User requirements include the analysis of user preferences for specific tasks, products and cultures. Resulting from this, a culture-oriented design is not possible without the empirical analysis of the user requirements in each culture, and the product to be developed for the respective markets. Culture models and factors are only an instrument for interpretation of empirical results. It does not seem possible to create a model with a data bank of all culture specifics for all tasks. With this in mind, it is useful to develop cultural models and factors to promote the understanding of cultural differences, but they will never be a complete substitute for empirical user analysis.

DESIGN ISSUES AND AN APPROACH FOR CULTURE-SPECIFIC DESIGN

User-centered design issues for product localization have by now been directed to the design issues within user interfaces. In the computer software industry, the need for localization of these issues for the target market has been widely rec- ognized. Many recent software products have already incorporated the localized design features for their user interfaces. The much-concerned culture-specific design issues within user interfaces have been listed by del Galdo (1990), and Fernandes (1995), which include color, icon, symbol, date format, time format, number format, language translation, etc. Other design issues such as menu direction, interface structure, information flow, etc. are addressed by some other recent researches (Choong, 1996; Dong & Salvendy, 1999). In the area of the production automation industry, the localization of machine user interfaces for the target market is much slower than in the computer software industry. In the case of machine exporting to Asian countries, many Western com- panies have only dealt with basic considerations of user interfaces (e.g. translation for the target market). Other user-interface design issues such as the information- presentation options and physical design of interface components, etc. have been usually neglected. Often, the user-interface design standards for Western countries are simply applied to the machines which are actually going to be used by other user groups with quite different cultural characteristics. This will probably result in many problems in human-machine interactions or even cause serious 70 KERSTIN ROSE¨

Table 3. Design Issues for Machine Localization.

Design Issues Within User Interfaces Design Issues Beyond User Interfaces

Interaction concept Machine functionality Menu structure Appropriate technology Navigation, dialogue form Service model Information coding Technical documentation Help and diagnosis General machine design Language industrial accidents. However, as the competition in the target market increases, many Western companies have now recognized the importance of localized machine user-interface design to their business success in the market, and more and more culture-specific user-interface design issues are being addressed in the design process. Furthermore, considering the actual requirements of machine localization in the foreign market, many design issues which are “beyond” user interfaces but are actually closely related to the user’s interaction with machines should be well addressed in the design process. These issues include: machine functionality, appropriate technology, service models, and some general machine design issues, etc. In some Asian countries which have no high industrial development, these functional design issues are often very important for the user. The main culture-specific design issues for machine localization in the foreign markets are summarized in Table 3. Obviously, cultural requirements should be well addressed in machine local- ization for the target market. There are still questions in practice about how the designers could find out these requirements based on the analysis of the target culture and then create a basis for the localized design. A good understanding of culture could provide the designers with clues to answer their questions. Culture can be defined as “learned behavior consisting of thoughts, feelings, and actions” (Hoft, 1996). For the cultural studies, culture models have been proposed to organize the cultural data with different international variables. Hoft (1996) has reviewed some culture models which could help the designers understand the features of culture and cultural variables. The meta-model from Stewart and Bennett (1991) classifies culture into two layers: the objective culture and the subjective culture, which correspond respectively to the visible portion of a culture, such as its economic system, political structures, social customs, arts, etc., and the psychological features of a culture, including assumptions, values, and patterns of thinking. Other meta-models of culture also suggest that a large portion of culture is “below the surface.” The Development of Culture-Oriented Human Machine Systems 71

For effective communications and interactions, the invisible portion of culture is very important. Some tested culture models identify this unconscious portion of culture and provide cultural information about the attitudes, behaviors, problem-solving strategies, thinking patterns, etc. of the cultural groups with some cultural variables (Hoft, 1996). (See the proceeding section.) It is necessary to combine cultural variables and meta-models of culture to form an approach of culture for application in culture-specific design practice. Consider- ing the unique importance of the user’s requirements in the whole design process, it is necessary to organize the existing cultural data into two distinct categories: cultural mentalities and cultural environments. The cultural mentalities refer to those cultural facets that are “within” a cultural group’s thought and behavior, and the cultural environments refer to those cultural facets that are “around” them (Fig. 3). These two categories correspond well to subjective culture and objective culture but are clearer for designers seeking to understand their contribution to designs. The cultural mentalities that are especially important for user-oriented design issues are shown in Fig. 3. They are determined by the cultural envi- ronments and, in turn, determine the designs. Some of these mentalities are also addressed by Plocher et al. (1999) as “the culturally determined user characteristics.”

Fig. 3. An Approach to Culture-Specific Machine Design. 72 KERSTIN ROSE¨

MODEL OF USABILITY ENGINEERING COSTS FOR LOCALIZATION

Machine Localization Issues

For the purpose of successful machine localization, those user-oriented design issues that are affected by cultural conventions should be well-addressed. The localization issues are distinguished into two groups: issues within user interfaces and issues beyond user interfaces, which are shown in Table 3.

Machine Localization Costs

Machine localization will involve additional costs to analyze, design, and realize localization-design features for the target market. These costs will then cause an increase in the total machine-development costs. However, for different machine localization issues and different kinds of machine developments, the localization costs could be quite different. If the localization is well organized from the beginning of the machine-development phase, the costs could be significantly reduced. The developers should also recognize the benefit (high user satisfaction, lower error rate) of machine localization when they design their product. One aspect of machine-localization cases will largely influence the localiza- tion costs, i.e. the invariability of the localization cases for different machine application cases, e.g. different machine types releasing the same product, etc. (Rose¨ et al., 2001). The localization case with high invariability is “fixed,” which means that the design features for the case could stay unchanged for different machine-application cases. The typical fixed case is fixed format such as time, date, etc. They only need to be changed once and then could stay the same for the design process of other machine application cases that follow. The localization case with low invariability is “variable,” which means that the design features for the case might be quite different for different application cases (not fixed item, all releases need adaptations). The developers should apply different features for different application cases. One typical variable issue is language translation. This factor has largely influenced the localization costs for different cases. Fixed localization cases usually require lower total localization costs than those of variable localization cases. (In accordance with this separation, LCfix is used for fix localization costs, and LCvariable is used for variable localization costs.) The design features for them are easily transferable and the requirements of the target users regarding these cases are normally not difficult to elicit (typically as standards and norms). The developers only need to spend once for the needed The Development of Culture-Oriented Human Machine Systems 73 localization results of these cases and then can apply them in different application cases. In comparison, the variable localization cases always need extra expense for the localization features in the new application case and their total localization costs are higher. The localization costs (LC) for one machine application case are composed of two localization cost parts:

LC = LCfix + LCvariable (1) The fixed localization cost will not occur for the subsequent machine application cases, so the total localization costs for the developers for their whole range of machine application cases can be expressed as: n LCtotal = LCfix +
LCvariable (i) (2) i=1

Localization Efficiency for Different Localization Levels

According to the localization difficulty and cost, the implementation of different machine localization cases may be distinguished as three localization levels. These levels are: surface level, interaction level and function level, as shown in Fig. 4. The surface and function level are described as levels with an uncovered cultural context and the interaction level with a covered cultural context. The covered cultural context is difficult to measure. Figure 5 shows an approach to the measurability of cultural behavior. Design issues on an uncovered level are easier to measure and to identify with user analyses. Therefore, they have lower localization costs.

Fig. 4. Machine Localization Levels. 74 KERSTIN ROSE¨

Fig. 5. Measurability of Cultural Issues.

In practical operation, when the developers decide on the most suitable localiza- tion range for their actual machine-application cases, besides the above mentioned factors for localization costs, the localization efficiency of different issues should also be taken into consideration. The localization efficiency (relationship of localization cost and localization benefits) could then be analyzed in comparison with localization cost (see formulas) and localization benefits for each level. The estimation of the localization efficiency for different localization levels is shown in Fig. 6. The issues on the surface level are mainly related to the

Fig. 6. Estimation of Localization Efficiency. The Development of Culture-Oriented Human Machine Systems 75 superficial cultural conventions, so the localization costs are not very high. They are closely related to the user’s correct understanding of the machines, so they are very important. These issues have the highest localization efficiency. The issues on the function level are the prerequisites for machine application, so they are also very important. However, the adaptation of some functional issues to user’s requirements may involve major changes of machine functionality and will increase the localization cost substantially. Therefore, machine localization in this level is usually implemented only when the functional features of the machine are not applicable without the adaptations. The issues on the interaction level are related mainly to the covered cultural conventions. The localization costs are very high, so these issues can not yet ensure high localization efficiency at present.

Costs for Machine Internationalization

For machine producers who must face up to different international markets at the same time, another aspect of machine localization issues should be taken into consideration to effectively control all machine internationalization costs. This aspect is the sameness of the issues for different markets. The same issue has identical design features for a machine for different markets, but dissimilar issues have different design features. The careful organization of different machine

Fig. 7. Characteristics of Localization Issues. 76 KERSTIN ROSE¨ Hierarchical Cost Structure. Fig. 8. The Development of Culture-Oriented Human Machine Systems 77 localization cases according to their characteristics of “fixed” or “variable” and “common” or “uncommon” could surely help the machine developers to effectively control their total international user-interface design costs according to their own product and market spectrum. The localization cases with their different characteristics are illustrated in Fig. 7. These postulated coherences have a strong impact on machine localization and cost efficiency. The following derivation of a hierarchical cost-structure is based on the previous descriptions. The cost-hierarchy is built up in three levels. The first is the level of user-interface issues, the second is the level of culture issues, and the third is the level of product issues. All together make up a hierarchical structure for cost estimation for one company. Figure 8 shows the level. The localization costs for a company are relatively high at the first level and decrease by the third level. These structure models take into account one company, with several user-cultures and several products. If you look at this model, it is easy to understand why localization costs are not as high as expected. If the developer of a product localizes the fix and flexible variables of a user interface for one user culture, than he has done 70–80% of the localization work for the next years. Why? If a user interface is localized for a culture (A), than the developer must look for the cultural difference between the culture (A) he designed for and the culture (B) he expects. Localization for the user interface for culture B will follow. The local- ization costs depend on the cultural difference between culture A and culture B. An increasing number of cultures do not automatically mean an increase of costs. If the cultures have only few differences (e.g. color coding), the effect of culture- oriented design is really satisfying for the user with low costs for the company. Usually, it is possible in one company to transfer the user philosophy from one product to another. The products have different functions, but the culture-specific user requirements are often the same. There are only differences in a special area or with special functionalities. Therefore, it is very important to start really early with localization, because localization is a long-term investment with high cost for user interface design in the beginning and good effects and low costs over several user cultures, products and years. The effect will be 100% user-satisfaction.

CULTURE AND THE USER INTERFACE (UI) ENGINEERING PROCESS

To understand how the cultural factors of interaction influence the development of user interfaces, it is necessary to take a look at the entire user interface and the system it is based on. The complete user interface development process must be examined and a look taken at the user interface as a whole. 78 KERSTIN ROSE¨

Fig. 9. Management of UI Engineering.

User Interfaces: An Integral View

Based on the model of the management for complex socio-technical systems from Rasmussen (2000), a model for the management of UI-design was created, see Fig. 9. The model for UI design in Fig. 9 shows on the left side the basic model components of a human-machine-interaction. Centered are the system levels: Platform, Data base, Structure, Presentation and Interaction. This is the order for the chronology of the developing process, which started with the decision for a system platform and ended with the interaction design. The developing process is a heterarchical process, which is based on a hierarchi- cal structure (see Hacker, 2000). That is the reason for defining it as a chronology for UI development. In this way, the results from the platform level are technical and system restrictions. The result of the data base level is the knowledge base of the system. Based on this, next is the structure level. Results from the structure level are the plan for the UI (structure for navigation, general layout). The results from the presentation level are design features (e.g. information classes, color coding, visualization technique etc.). Finally, the result of the interaction level is the action design (e.g. dialog-basics-elements). All levels are strongly influenced by and dependent on each other. There is interaction between them, forward The Development of Culture-Oriented Human Machine Systems 79

Table 4. Culture-Environmental Stressors.

System Level Environmental Stressors Culture-Environmental Stressors

Platform Changing platform Differences of preference and prerequisite Software update Culture-specific Software Version New processor/technology Culture-specific environmental conditions (climate, reliability of power supply) Data base Different knowledge High variances of user knowledge in different cultures New data (format) Culture-dependent data format (if new implication, than for each cultural software-version) User group User group has cultural-dependent variances Changing user-/ Cultural-dependent differences of the system structure user-/system structure must be taken into account Structure Different user tasks Cultural influences variances of user knowledge Changing functionalities Integration of translation and explanation in all culture-specific software-versions Changing structure Culture-specific differences of the developer and user model must be coordinated, culture-specific structure must be evaluated New user group/ Evaluation of the culture-specifics for user application area group/application area Presentation New visualisation Culture-specific preferences, prerequisites and technologies taboos Different in-/output devices Anthropometrical specifics, production and working conditions New environment Culture-specific working conditions/structure and environment Changing surface Culture-specifics for color and icon usage and taboos Interaction Changing user group Cultural-specifics of working structure (training system) and user-knowledge Changing interaction mode Cultural taboos (esp. Assistants & Avatare) and preferences Changing dialog elements Evaluation of publicity and usability for the specific culture and backward (shown in Fig. 9). The UI engineering process is really complex, because there are a lot of environmental stressors which one must manage (in Fig. 9 shown on the right side and Table 4). Some environmental stressors are influencing only one level, others more than one. This model for the management 80 KERSTIN ROSE¨ of the UI engineering process takes into account “environmental stressors” (shown on the right side), which influence the UI management. These stressors are described for each system level of the process. The UI Engineering Process is influenced from the cultural context of the current developer and the former user. Therefore, it is really interesting to focus on them. Culture influences the whole UI engineering process. As a result, management of intercultural usability engineering is necessary. Culture is a context variable and also an “environmental stressor.” Based on these points, Table 4 lists the “environmental stressors” for an Intercultural UI management process. Culture-environmental stressors are shown as an application of the UI man- agement model in Table 4. This section focuses especially on the intercultural UI Engineering process and its management, because most studies and papers in this field only concentrate on one specific variable of cultural influences, e.g. color or icon usage. Culture is a context variable of the user. Intercultural UI design implements the integration and coordination of the user and his cultural context. This is the only way to design user-oriented UI in a time of globalization. To be successful, the influence of culture on the complete management of UI engineering must be realized and converted into an Intercultural UI management process. Therefore, models – like those shown in this paper – must be created as a platform for guidelines which one can develop through integration of the current study results. The following model for the Method of Culture-oriented Design (MCD) was developed first for the area of production technology and considers, in particular, the special problems of developers in this area. However, an expansion for general applicability is the real aim. In the following, the reader will find some aspects of this new approach.

MODEL FOR INTERCULTURAL DESIGN OF HUMAN-MACHINE SYSTEMS

The Idea of MCD

To support developers of internationally usable human-machine systems, the idea for the Method of Culture-oriented Design (MCD) was developed. The MCD integrates the aspect of culture-oriented design (see Coronado, Day & Hall, 2002; Day, del Galdo & Prabhu, 2000; Day & Dunckley, 2001; del Galdo & Nielsen, 1996; Hoft, 1995) and established user-oriented design theories (EN ISO 13407; Johannsen, 1993; Preim, 1999; Shneiderman, 1998) into existing concepts for UI design. This method should help developers with the draft of the human-machine interface. To realize this, the MCD has to consider the existing developer methods The Development of Culture-Oriented Human Machine Systems 81 for drafts of human-machine systems. The MCD must be based on an integral approach, i.e. the knowledge about cultural differences must be integrated into the developer’s existing methods of human-machine system design. As a basis for this procedure of MCD, the existing concepts by Schneider (1997) and Wahl (1999) for ergonomic user interfaces in production technology are used. Combined with an existing concept (module-based combination) for user-interface design, it will inform the developer about the relevant information for culture-oriented design at the very moment when he needs this information. Culture orientation is only one component of ergonomic UI design. Therefore, a complete new method is not useful. In order to be able to carry out this procedure, those modules of the existing concepts which are relevant for a culture-oriented design have to be determined. The culture-oriented factors must be integrated into these modules, after redesign or adoption with models of UI management (see Fig. 10). A MCD matrix will give the overview for integration, which contains all relevant influences for the design modules derived from culture-oriented factors. Only the integral approach is able to support the developer’s work, by enabling him to work with well-known methods. He does not have to learn any new method, making it possible for the developer to acquire new knowledge (about culture-oriented design) in an application-oriented way.

Fig. 10. Integrative Approach of the Culture-Oriented Design. 82 KERSTIN ROSE¨

The integration of intercultural variables into a standard method of user-oriented design should make it possible for the developer to create user- and culture- oriented products. In this case, the “intercultural design of man-machine systems” is possible and should be named intercultural user-interface design (integration of user and cultural specifics). The following method of culture-oriented design (abbreviated: MCD) is based on this integrative approach. To develop the MCD, the following steps were needed:
Analyzing and evaluating the developers’ requirements.
Definition of intercultural variables and factors.
Definition of docking places to the “normal” developing process of Human- Machine Systems.
Making a systematic to identify and integrate the qualities of intercultural variables.
Definition of application steps for the integrative design. The developers’ requirements in the application area of production automation were analyzed with a questionnaire and documented. Some selected results:
89% think globalization is a challenge (less than 42% think the globalization problem is solved);
44% would like to adapt an existing user system to foreign countries (internation- alization), 11% think they must develop specific user interface for the specific user culture (localization);
more than 52% think support for intercultural design by developing guidelines would be helpful;
the information sources with a growing meaning in the future would be: internet, information on CD and support systems. Detailed information on the developers’ evaluations are published in Rose¨ and Zuhlke¨ (2001) and Rose,¨ Zuhlke¨ and Liu (2003). The next step is the definition of intercultural factors and variables. Intercultural factors from Marcus (2001) (see Table 2) could also be used for the MCD. Intercultural variables have as yet to be defined. The definition of intercultural variables is described in the following.

THE MAKING OF INTERCULTURAL VARIABLES

Intercultural Variables

Relevant variables for intercultural aspects are extracted from the standard vari- ables of a human-machine-design approach (see Wahl, 1999). These procedures The Development of Culture-Oriented Human Machine Systems 83

Fig. 11. Intercultural Variables for Human-Machine-Systems. were carried out in the context of human-machine systems. As a result, there are particular variables which must be taken into account for an intercultural design: the intercultural variables. These intercultural variables are divided into: direct and indirect design variables. According to the idea of an integral approach, the frame variables should be also taken into account. Figure 11 shows all intercultural main variables. These are three main groups of intercultural variables. First are the direct variables, which have an impact on usability and acceptability of an interface design. Second are the indirect variables, which have an impact on applicability and feasibility of a human machine system. Third are the frame variables, which have an impact on utilizability of a system. Each system is embedded into an environment. Based on a context-oriented design approach, this environment must be taken into account in product engineering, and these contextual frame considerations of usage and production are drawn in the frame variables. The direct intercultural variables have an immediate influence on the design of human-machine interaction. They determine quite fundamentally the usability and acceptance of the final product (in accordance with the DIN ISO 66234, part 8). These direct intercultural variables are: Information presentation: time format, date format, standard unities, standard format of tables and diagrams Information coding: colors, symbols, labeling (incl. abbreviations), (acoustic) signals Dialog design: menu structure, dialog form (basic elements of interaction), approximate layout 84 KERSTIN ROSE¨

Interaction design: system structure, navigation concept, anthropometric specifics, specific interface qualities Language: reading-/writing direction, text sorting and multilingual labelling, report, indication, searching algorithm The indirect intercultural variables have a mediating influence on the design of human-machine systems. They determine essentially the applicability of the system and the feasibility of the work task, especially user orientation and functional availability. Other influential aspects are customer orientation and customer satisfaction. Sole consideration of the direct intercultural variables does not guarantee these aspects; the technical components are also important. Only the consideration of all intercultural design variables allows successful user orientation and customer satisfaction. The indirect intercultural variables are: General machine design: machine construction, hardware design (including color composition), packing, product and company name Machine functionalities: size of the functions, offer or implementation of added features (specific functionalities), configurability Technical documentation: kinds of documentation and their content, structure and layout (also consists of graphical descriptions) and the translation quality Service: maintenance offers and intervals, offers of tele- service, availability of spare parts, training offers Described above was a division of intercultural variables into three main parts: direct, indirect and frame. The third part, the frame variables, is important for the utilizability of a system. Physical components (like climate conditions) and structural differences (e.g. working structure, working system) are described with these variables. They envelop the usage situation like a frame. Indirect and frame variables are a construct of context-oriented design and take into account the usage context of a product. The frame variables are: Environment factors: Climate, production conditions, production plants Technical standard: historical development (and level) of technique, energy fluctuations, technical operating conditions (e.g. networking conditions for teleservice) The Development of Culture-Oriented Human Machine Systems 85

Work structures: hierarchical structure, number of user groups, access rights, shared workspaces Education system: user’s qualification level, technical pre-knowledge, exploration behavior (problem solving strategies) Political system: laws and regulations, cooperation policy, support on the spot Frame variables are also important for the design of human-machine systems. In accordance with location conditions, they should be clarified before the development begins. Marketing department or representations on the spot are a good source for much of this information. However, for a good intercultural human-machine system, all intercultural main variables should be considered.

Intercultural Variables and UI Engineering

All intercultural variables can be assigned to the levels of the approach for human-machine system engineering. Such an assignment is shown in Fig. 12. For the assignment, the model of culture-environmental stressors is used, and the in- tercultural variables are assigned in relation to Table 4. Some variables could have an impact on more than one level, e.g. machine functionality and work structure. All previous system levels and subdivisions are useful for the specification of the intercultural variables. However, Table 5 shows an integration of the direct, indirect and frame variables into the system levels of human-machine system engineering. Establishing the “docking places” for the intercultural variables is one of the steps in developing the MCD. Culture-oriented design and intercultural variables are one integrative part of user-oriented design. The docking places describe the integration of the intercultural variables into the system levels of human-machine system engineering. For that reason, Table 5 shows the integration positions for the intercultural variables and enables a definition of the intercultural docking stations for the process of developing human-machine systems.

Description Level for Intercultural Variables and Issues

Summarizing the previous descriptions, Table 6 shows the useful specifications of intercultural variables. Three description levels are shown:
Impact of the design of a HMS (compare Table 5).
Context dependency and measurability (compare Figs 4 and 5).
Changeability during the product life cycle (compare Figs 7 and 8). 86 KERSTIN ROSE¨

Fig. 12. Assignment of Intercultural Variables to Design Levels of the HMS Engineering.

From the description of intercultural variables in Table 6 and the definition of dock- ing stations in Table 5, a systematic can be derived. The systematic of intercultural variables will be described in the following section.

Systematic to Identify Intercultural Variables and to Integrate into HMS

For the construction of a systematic for intercultural variables a catalogue of all intercultural variables is needed. The information about intercultural variables from the previous section can be used for this catalogue. The described variables could be different with regard to:
Impact of the design on human-machine systems (Fig. 11).
Docking places in the human-machine systems engineering process (Table 5).
Cultural context dependency and resulting measurability (Fig. 4), and methods to identify them. The Development of Culture-Oriented Human Machine Systems 87

Table 5. Docking Places for Intercultural Variables.

System Level Intercultural Variables Categories of Intercultural Variables Direct Var. Indir. Var. Frame Var.

Human-machine system (intern) Technology Machine functionalities ∗ Technical standard
Data/Model Machine functionalities ∗ Work structure
Structure Education system
Work structures
Service ∗ Interaction Interaction design ∗ Dialog design ∗ Technical documentation ∗ Presentation Information presentation ∗ Language ∗ Human-machine system (extern); working/environmental conditions General machine design ∗ Technical documentation ∗ Service/maintenance ∗ Education system ∗ Environmental factors ∗ Political system ∗

Stability factors for the variables and resulting care effort, and expected cost factors (Figs 5 and 6–8).

The previous section described this information. The question is: how to catalogue the intercultural variables? Intercultural variable “x” can be assigned into the catalogue with the degree of their influence on the human-machine system. Therefore, they could be system- atized as direct, indirect or frame variable (see Fig. 11). A variable status can be assigned to an intercultural variable. Furthermore, each intercultural variable has its own docking place in the human-machine system engineering process (see Table5). Both parameters have a strong coherence with the human-machine system. As described in the previous sections, variables can be different with regard to the determination of cultural context. The possible distinctive features are: covered and uncovered variables (see Fig. 4). A further parameter is derivable from this, context dependency. The analysis of the intercultural variables during the lifetime of a human- machine system and in coherence with the product releases has been also 88 KERSTIN ROSE¨ of the HMS. the HMS design. – Impact of the utilizability measurability. product life cycle. several times (depend on product specifics). less deep (shallow) cultural determination. the HMS. – Less stability of appearance during the – To define and integrate in to the HMS design – Good up to not easily accessibility and – Low up to middle context dependency and – Impact of the applicability and feasibility on – Mediate influence on the HMS design. – Contextual influence on Description Level for Intercultural Variables. Table 6. during the product life cycle. HMS design once. measurability. cultural determination. of the HMS. design. Fix variable– High stability of appearance ability Flexible variable – To define and integrate into the – Bad accessibility and Covered variable– High context dependency and deep Uncovered variable – Impact of usability and acceptance Direct variable– Immediate influence on the HMS Indirect variable Frame variable product life cycle and measurability on HMS Changeability in the Context dependency Impact of the design Description Level Specifications The Development of Culture-Oriented Human Machine Systems 89

Fig. 13. Systematic for the Description of Intercultural Variables. presented in the previous section. Thus, the intercultural variables are describable with regard to the complete lifetime of a human-machine system, and differen- tiable in fixed and flexible variables. Stability factor is the fourth parameter for intercultural variables. For the description of intercultural variables in a catalogue, the following parameters have to be used: variable status, docking place, context dependency and stability factor. Figure 13 shows the catalogue systematic. The complete catalogue with the description of all intercultural variables can be found in Rose¨ (2002). An application example will be shown in Fig. 16. Figure 13 shows the systematic for the description of intercultural variables of human-machine-systems. This grid allows the description of all variables from Fig. 11. It is “cultureless” and only a structuring aid. With this systematic, no design of an intercultural human-machine-system is possible as yet. Information about the application culture and the culture-specifics of the user are needed. The next step is the combination of the intercultural variable systematics (Fig. 13) and the intercultural user interface factors, shown in Table 2. The combination of both is presented in Fig. 14. After selecting a specific user culture, the description of cultural user-interface factors can start. Based on the cultural user interface factors (see Tables 1 and 2; also Hofstede, 1997; Marcus, 2001) the definition of the user culture is carried out. The description of a user culture starts with the assignment of cultural dimensions: power distance, individuality, gender, uncertainty working and time orientation, and ends with the definition of resulting effects for: metaphor, mental model, navigation, interaction and presentation with regard to the intercultural user-interface factors (see Marcus, 2001). 90 KERSTIN ROSE¨

Fig. 14. Combination of Cultural Factors and the Intercultural Variables Systematics.

To determine the cultural difference between the target user culture and the existing developer culture, this procedure must also be carried out for the developer’s culture. This cultural difference has a direct impact on the interna- tionalization effort for a human-machine system. Therefore, a “reference culture” must always be defined. The advantage of this approach is global usability. There is no specific reference culture, and therefore, this approach is usable for a developer in India, U.K., U.S. and so on; the first step is always the definition of the reference culture. After the definition of the cultural difference between reference and user culture, the changes for the intercultural variables can be determined. However, the cultural user interface factor must be taken into account for the definition of the intercultural variables. The change effort for the internationalization of a specific human-machine system in a specific culture is determinable as result from the cultural difference and with regard to the stability factor of the changing for these intercultural variables. The Development of Culture-Oriented Human Machine Systems 91

The combination of cultural factors and the intercultural variable systematic is an important element in the method of culture-oriented design (MCD). The following section describes the method of culture-oriented design and its application.

MCD: APPLICATION DESCRIPTION

In this section, the step by step procedure of the MCD is explained. For the explanations in this section the following prerequisites apply:

The catalogue of the intercultural variables is available (see Rose,¨ 2002).
Complete definition of intercultural user interface factors is available (see Marcus, 2001; Rose,¨ 2002).
The target user culture is defined.
The explanations are based on the systematic for human-machine-system engineering of Wahl (1999), see also Fig. 9. Adoption of other procedures and systematics is possible for 99% of all other “normal” engineering approaches.
The general concept (use, purpose, and general product specifics are known) of the human-machine system exists.

It is really important to remember that MCD is only a systematic for data analysis, data preparation and data integration. MCD does not substitute for the empirical analysis of user requirements in the corresponding user culture, especially in the case of product localization. Figure 15 shows an overview of the method of culture-oriented design of human-machine systems. This figure also serves as the application overview for the MCD. What follows is the 10-step procedure of the MCD:

(1) Definition and description of the user culture. (2) Determination of the culture UI-factors for the reference culture (developer’s culture) and the user culture (product usage culture). (3) Comparison of the cultural UI-factor profiles, concerning similarities and differences of reference culture and user culture. (4) Select the product relevant intercultural variables. (5) Derivation of variable status from intercultural variables catalogue (for each of the selected variables). (6) Determination of the culture-context dependency, using the intercultural variables catalogue (for each of the selected variables). 92 KERSTIN ROSE¨

Fig. 15. Application Overview for the Method of Culture-Oriented Design of HMS (MCD).

(7) Specification of the information to be analyzed in the user culture. Step 5 includes information about analysis content, and step 6 includes information about the analysis methods. (8) Analysis of the needed culture-specific information and user requirements in the target user culture. (9) Derivation of the docking place for the results integration; use of the intercultural variables catalogue (for each selected variable). (10) Integration of the user culture-specific information into the human- machine system engineering process for the design of a culture-oriented product/system. (11) OPTIONAL: Cost calculation and cost estimation (see Figs 13 and 15). The change effort resulting from the culture factor profile comparison (see step 2) and the stability factor of the selected intercultural variable. Starting with step 4, this procedure must be carried out for each selected intercultural variable. What is the standard usage case, how does it look? As a standard case, the developer inquires into the cultural factor profiles for the reference culture and The Development of Culture-Oriented Human Machine Systems 93 the user culture (steps 1–3). Following this, needed intercultural variables are selected, and the analysis content and evaluation methods are then derived (steps 4–7). The result is a complete analysis and evaluation concept (variable-oriented approach). Based on it, the culture-specific user requirements are analyzed (step 8). After execution of the analysis, all needed culture-specific information – from the user for the specific product – is available. The analysis result will be prepared for integration into the “standard procedure” in a variable-oriented way (step 9). Finally, the intercultural variable information will be added to the appropriate docking places into the HMS engineering process and will than be integrated (step 10). A developer now gets the culture-specific information exactly at that point in the engineering process when he needs it. Hence the MCD is a simultaneous procedure (or added procedure) of the standard HMS engineering procedure. With such handling, it is possible to find out critical application differences in the user culture, and this information can be integrated into the complete system engineering process at an early point of development. Such a procedure enables flexible (and module based) engineering of interactive systems with culture-specific design for the international market. MCD helps to avoid information loss in the engineering process. It supports the developer during the development process and guides him during the procedure of analyzing intercultural user information, completely from data analysis up to information integration. With usage of the intercultural variables catalogue and a user-oriented HMS engineering approach, the culture-oriented design will be a soluble challenge for the developer of interactive products. Nevertheless, an analysis of user requirements in each expected user culture is very important to guarantee user-friendly and culture-oriented products. The following is a description of an application of MCD. I used data from a cross-cultural project to explain the handling of MCD. All data are from the project Intops. Project Intops was carried out by the University of Kaiserslautern, with the aim of making statements on the influence of culture on the design of user interfaces for machines In order to clarify the question of whether European industrial standards can be employed in other culture areas, or whether there are peculiarities, project Intops was carried out. Within the framework of this project, countries have been chosen that are of interest for exports from the German engineering industry, including the following: USA, China, South Korea, India and Indonesia. Four to five industrial sites were chosen in the selected countries from the following sectors: automobile producer, automobile supplier, plastics or synthetic manufacturer, or company with products reflecting national priorities. The data in project Intops were ascertained without the MCD; this method did not exist during the term of the project. Therefore, the 94 KERSTIN ROSE¨ data analysis was not pre-structured with MCD, and it is possible that some application problems were not found. The complete project data are published in Zuhlke¨ et al. (1997).

APPLICATION OF THE MCD WITH RESULTS FROM THE INTOPS PROJECT

The MCD was developed for the specific area of human-machine-system engi- neering in the field of production automation. Project Intops (Zuhlke¨ et al., 1997) was a cross-cultural study of user differences in the field of production automation. User differences were analyzed in India, Indonesia, South Korea, Mainland China, USA and Germany. All regions are relevant German export countries. For the explanation of the MCD-handling, I chose mainland China. It is a very traditional culture and an important export market for the German machine tool industry. The MCD is based on the cross-cultural principles. Therefore, the example shows a comparison between China and Germany. Explanation starts with the definition of cultural factors, followed by description of the user-interface characteristics for both countries and the example of a definition of intercultural variables, and finishing with design recommendations.

Definition of the Cultural Factors

The first is the definition of the cultural factors. The user culture is mainland China and the developer culture is Germany. The combination is optional. The combination can change based on culture changing for developer or user. Table 7 shows a definitional description for both cultures. This description is based on

Table 7. Definition of Cultural Factors for China and Germany (Based on Hofstede, 1997).

Culture Factors China Germany

Power distance High Low Individuality Collectivism Individualism Gender Feminity Masculinity Uncertainty avoidance Uncertainty avoidance Uncertainty avoidance Time orientation Long time orientation Short time orientation The Development of Culture-Oriented Human Machine Systems 95

Hofstede (1997). A complete definition for all cultural factors can be found in Rose¨ (2002) and Marcus (2001). The description of Hofstede’s cultural factors shows a relatively great difference between China and Germany. Four of five factors show differences. Only the factor “uncertainty avoidance” is the same. These factors correspond with specific user-interface features, like those shown in Table 2 and Marcus (2001). Table 8 shows the characteristics of user interfaces for China and Germany, based on this approach. The characteristics shown in the table above provide the possibility for general design advice. There are some differences which could be directly integrated into an international approach for human-machine interaction, like error messages or choice possibilities, and profile basics. On the other hand, there are similarities for the interaction design for China and Germany, like Input and Feedback, many redundancies, and limited navigation options combined with simple controls. These similarities are based on the factor agreement “uncertainty avoidance.” Both cultures have the tendency to avoid uncertainty. These observations (information) are helpful for a general design approach and to determine the general cultural differences between China and Germany, but they are not detailed enough to support the developer during his process of developing a human-machine system for the Chinese market. The developer needs a description on the detail level of intercultural variables. As shown in Fig. 15 and described below, steps 1–3 are fulfilled successfully. This must be followed by step 4: definition of intercultural variables.

Definition of the Intercultural Variables

For a complete description of all relevant intercultural variables, there is not enough space in this section. Therefore, in the following example, a description of the MCD application principle is shown for the intercultural variable “information coding” (compare Fig. 11 and Table 5). The intercultural variable “information coding” has several main areas: one is “color,” the others are “symbols,” “labeling” and “signals.” Figure 16 shows one sheet from the intercultural variables catalogue, an overview for the intercultural variable “information coding.” With the catalogue sheet, the developer gets the following information:

“Information coding” is a direct variable and has an immediate impact on the design of human-machine-systems.
The docking place for this intercultural variable is the system level “presen- tation.” For the design of a Chinese human-machine system, this means that 96 KERSTIN ROSE¨

Table 8. User-Interface-Characteristics for China and Germany.

China Germany

Metaphor Clear hierarchy Event orientation at tasks and tools Relation and family-oriented Work-oriented, competition-oriented Representation instead of abstraction Representation instead of abstraction Traditional powerful structures Interchangeable roles and jobs Mental model Referring to data without relevance Little structure data with relevance order Role-oriented Product-oriented, task-oriented Social structures, relation-oriented Work and business structures, aim-oriented Simple mental models, clear Complex mental models, tolerance of articulation ambiguities Choice limited, binary logic Unclear (fuzzy) logic Navigation Reduction of choice and navigation Open entrance, divisible paths, areas individual paths Predefined choices Arbitrary choices Role-dependent navigation behavior Role-independent navigation behavior Multiple selection, non-uniform Limited selection, uniform navigation navigation Limited options and simple controls Limited options and simple controls Tolerance for long ways, ambiguities Referring ways, taxonomies Interaction Error messages as: “Entry Supporting error messages (With forbidden,” “You are wrong” instruction to the remedying) Wizards and Guides support the use Note maps Restricted search possibilities Keyword search Role profiles instead of user profiles User-specific behavior Practical, function-oriented Game-oriented, level-oriented Team- (cooperation-) oriented Individual-oriented Input and feedback: exact, complete, Input and feedback: exact, complete, detailed detailed Face-To Face interaction preferred Distant communication accepts Personalized news Anonymous news Presentation Formal talk Informal talk High context dependency Low context dependency Feminine colors Masculine colors (many blue tones) Terminology and pictures: simple, Terminology and pictures: simple, clear, uniform clear, uniform Many redundancies Many redundancies Arrangement of intimacy and social Concentration on the task to be group membership presented with or the product The Development of Culture-Oriented Human Machine Systems 97

Fig. 16. Information Coding Sheet from the Intercultural Variables Catalogue.

culture-specific information about user requirements must be taken into account during the engineering of the subtasks at the system level “presentation.”
The context dependency of the intercultural variable “information coding” is uncovered, and therefore good to measure.
Stability factor shows high stability of appearance. The definition of the features for the variable must be undertaken once during the product life cycle.
An on-the-spot analysis is needed as an evaluation method. The on-the-spot-analysis was carried out in the context of project Intops. Tests were carried out on color coding and icon usage. Table 9 shows a summary of results for the intercultural variable “information coding.” In Zuhlke¨ et al. (1997) and several other publications, the detailed results are presented. With that, steps 4–8 are fulfilled successfully (see also Fig. 15). During the human-machine system engineering for China, this information must be taken into account in the system level “presentation.” In detail, the color coding for China must be considered (in accordance with the project results: only a safe differentiation between the color red = emergency/problem and the color green = anything goes). Therefore, the redundancies of information presentation should be clarified with additional coding (like flashing). In general, only simple and pictorial icons (in coordination and checked with the cultural background) should be used. Technical background knowledge of the user cannot be put ahead of these considerations. Generally, all used icons should have a relation to the user’s background. Therefore the symbol/icon “sand clock” is not a good choice for China. A normal 98 KERSTIN ROSE¨ Germany Color coding in accordance to DIN60073 EN usable without any limitations. In general there is a goodthe recognition DIN-Icon. of Is limited ifhighly the abstract. icons are Culture-Specifics in China and Germany for the Intercultural Variable “Information Coding.” China coordinate with the cultural background (e.g.symbol the – sand does clock never exist – insymbol windows China, waiting and therefore its exist meaning). no mentalDIN-Icons model were for only this partly recognized. Theor developing labeling of with specific Chinese Chinese letter Icon andcontext would the have coordination a of higher these recognize with ability. the cultural Can not use the colorstate yellow for “emergency.” warning states, this colorThe were color interpreting black as is the required forWindows “standard” the colors operation were state accepted. “normal.” Achanging configurability into with a the second option “standard for mode” with pastel tones should be offered. Table 9. Icon usage Pictorial icons have to be preferred. Their recognize ability must be checked and Color coding No differentiation between the operating states, “warning” and “emergency.” The Development of Culture-Oriented Human Machine Systems 99 clock would be much easier to understand in China and would transfer the metaphor “waiting, process is going” to the mental model much better. Around the world, all users have problems with the understanding of high technical abstract DIN-icons. Simple use of these icons without testing should be avoided. The combination of icons and user language is recommended. The labels must be in the user language and should offer the possibility of change to other user languages (e.g. via soft key usage). Applications for China should offer no icons, only labels in Chinese. The Chinese language is a very complex and pictorially oriented language. The use of icons would not support the understanding of icons or easier learning. The icons are a second pictorial language for the Chinese user and therefore require a new learning effort and no real support. Some circumstances require the usage of icons, e.g. with a view to international adaptability and trade-specific icon usage. In such cases, a good labeling with clear and easy icons can be combined. This information is only a little part of the complete intercultural variables cat- alogue and the analyzed culture-specific user requirements. A lot more elements must combine for the engineering of a complete human-machine system. Only the completeness of all intercultural variable information is a basis for an optimal human-machine system. However, with that information, steps 9 and 10 can be executed. Without reference to specific intercultural variables, here are some more interesting features for a human-machine system for the Chinese market.
Restricted navigation and predefined choices, but most (various) choices on the top navigation level.
Simple, clear structures with preferably binary choices.
Consideration of possibilities for teamwork and cooperation.
High context dependency for presentation and preference of feminine colors. This information is based on the cultural factors and the user analysis. They are presented mainly to explain the handling of the MCD. The Chinese market is only an example, and therefore, the design information is not complete.

EVALUATION OF THE MCD

The information combination and information presentation carried out in the previous chapter have shown, that a small amount of information could bring a great input for design, if it is structured and easier for the developer to integrate into the design and development process of a product. With the usage of MCD, an early definition of the right and efficient analysis and evaluation methods 100 KERSTIN ROSE¨ is possible, as well as the specification of needed information. MCD is a good structuring tool for the developer of intercultural products. Previous explanations have shown a brief application of the MCD. Hopefully, the handling was recognizable. To be able to take statements concerning the completeness of intercultural variables and handling problems, much more appli- cation data must be collected. The author started with the collection of application data. That is necessary to take the step from an approach to a real method. Nevertheless, development is needed for the scales of intercultural factors (individualism etc.). Rating scales with fixed points and more details are better than polarized scales (see Osgood, 1975). More machine-oriented description for the intercultural factors should be developed. Data analysis should be improved as well. Actually there is only information about the kind of method such as: which method should I use? How do I get information? Where did I get my information? Who is the information key person? Based on the cultural structure? The next step of method development will be the construction of an information matrix. Valuable information should be: where did I get which information from which person and what method should I use for an analysis. Some of this correlated information already exists in the actual version of MCD. Missing components of the information structure are actually in development. Important for the quality of a culture-oriented human-machine system is an extensive collection of cultural data for the developer. Actually there are not enough culture-specific preferences collected; the high specification degree of human-machine systems is one reason for it. Hopefully, with the development of culture-oriented web sites, there will also be an increase of culture-specific data which are partially useful for culture-oriented human-machine systems. With MCD there exists a useful structuring approach for the integration of these data.

SUMMARY

This chapter has shown one way to specify, analyze, and integrate intercultural variables into human-machine system engineering. Based on traditional and well-known approaches from the work of Hofstede and Marcus, an additional treatment for human-machine systems was presented. The method introduced is called the “method of culturally-oriented design” (MCD). It was explained and its elements were derived over several sections, then combined. An example of its application and general use were shown. This approach and further methods represent only one possible way to design intercultural systems and products. Of special importance is the definition and treatment of intercultural variables, The Development of Culture-Oriented Human Machine Systems 101 which are the result of an integral context oriented view of system engineering, especially in an intercultural milieu.

EDITOR’S NOTE

As used by the author on pages 88–90, the term systematic refers to a plan whose main tasks and steps are listed in a hierarchical order.

REFERENCES

Bourges-Waldegg, P. (2000). Globalization: A threat to cultural diversity? In: D. Day, E. M. del Galdo & G. V. Prabhu (Eds), Designing for Global Markets 2, Second International Workshop on Internationalisation of Products and Systems (pp. 115–124) (IWIPS 2000, Baltimore, Maryland USA, 13–15 July). Backhouse Press. Bourges-Waldegg, P., & Scrivener, S. A. R. (1998, February). Meaning, the central issue in cross-cultural HCI design. In: Interacting with Computers: The interdisciplinary Journal of Human-Computer-Interaction (Vol. 9, pp. 287–309). Elsevier. Choong, Y.-Y.(1996). Design of computer interfaces for the Chinese population. Doctoral dissertation. Purdue University. USA. Choong, Y., & Salvendy, G. (1998, February). Designs of icons for use by Chinese in mainland China. In: Interacting with Computers: The interdisciplinary Journal of Human-Computer-Interaction (Vol. 9, pp. 417–430). Amsterdam: Elsevier. Coronado, J., Day, D., & Hall, B. (Eds) (2002). Designing for global markets 4. Proceedings of the fourth International Workshop on Internationalisation of Products and Systems (IWIPS 2002, Austin, Texas USA, 11–13 July). Day, D., del Galdo, E., & Prabhu, G. (Eds) (2000). Designing for global markets 2, second international workshop on internationalisation of products and systems (IWIPS 2000, Baltimore, Maryland USA, 13–15 July). Day, D., & Dunckley, L. (Eds) (2001). Designing for global markets 3. Workshop Proceedings of the third International Workshop on Internationalisation of Products and Systems (IWIPS 2001, Milton Keynes, Great Britain, 12–14 July). del Galdo, E., & Nielsen, J. (Eds) (1996). International user interfaces. New York: Wiley. de Souza, M., & Dejean, P.-H. (2000). Cultural influence on design. In: D. Day, E. M. del Galdo & G. V. Prabhu (Eds), Designing for Global Markets 2, Second International Workshop on Internationalisation of Products and Systems (pp. 69–84) (IWIPS 2000, Baltimore, Maryland USA, 13–15 July). Backhouse Press. Dong, J., & Salvendy, G. (1999). Designing menus for the Chinese population: Horizontal or vertical? Behaviour & Information Technology, 18(6), 467–471. EN ISO 13407 (1999). Human-centered design process for interactive systems (EN ISO 13407: 1999). Fernandes, T. (1995). Global interface design: A guide to designing international user interfaces.Ap Professional, Boston. Hacker, W. (2000). Cognitive modeling of design problem solving. In: Ergonomics for the New Millennium: Cognitive Ergonomics, Computers, and Communications (Vol. 1, pp. 620–622), 102 KERSTIN ROSE¨

Proceedings of the XIVth Triennial Congress of the International Ergonomics association and 44th Annual Meeting of the Human Factors and Ergonomics Society (San Diego, California, USA, July 29 through August 4, 2000). Hall, E. T. (1989). Beyond culture. New York: Anchor Books. Hofstede, G. (1997). Cultures and organisations: Software of the mind. New York: McGraw-Hill. Hoft, N. (1995). International technical communication. New York: Wiley. Hoft, N. (1996). Developing a cultural model. In: E. M. del Galdo & J. Nielsen (Eds), International User Interfaces (pp. 41–73). New York: Wiley. Honold, P. (2000). Intercultural usability engineering: Barriers and challenges from a German point of view. In: D. Day, E. M. del Galdo & G. V. Prabhu (Eds), Designing for Global Markets 2, Second International Workshop on Internationalisation of Products and Systems (pp. 137–148) (IWIPS 2000, Baltimore, Maryland USA, 13–15 July). Backhouse Press. Johannsen, G. (1993). Mensch-Maschine-Systeme (Human-machine-systems). Berlin Heidelberg: Springer-Verlag. Marcus, A. (2001). Culture(s) in user interface design for the web and mobile computing. In: D. Day & L. M. Dunkley (Eds), Designing for Global Markets (Vol. 3, pp. 23–32). Workshop Proceedings of the third International Workshop on Internationalisation of Products and Systems, IWIPS 2001, 12–14 July, Milton Keynes, UK. Marcus, A. (2001a). Cross-cultural user-interface design. In: M. J. Smith & G. Salvendy (Eds), Systems, Social and Internationalization Design Aspects of Human-Computer Interaction (pp. 502–505). Volume 2 of the Proceedings of HCI International 2001, 5–10 August 2001, New Orleans, Louisiana, USA. Osgood, C. E. (1975). Cross-cultural universals of affective meaning. Urbana Chicago, London: University of Illinois Press (p. 89ff). Plocher, T. A., Garg, C., & Chestnut, J. (1999). Connecting culture, user characteristics and user interface design. In: H.-J. Bullinger & J. Ziegler (Eds), Human-Computer Interaction: Ergonomics and User Interfaces (Vol. 1, pp. 803–807). Proceedings of the 8th International Conference on Human-Computer Interaction (Munich, Germany, August 22–26, 1999). Mahwah, NJ: Lawrence Erlbaum. Prabhu, G., & Harel, D. (1999). GUI design preference validation for Japan and China – A case for KANSEI engineering? In: H.-J. Bullinger & J. Ziegler (Eds), Human-Computer Interaction: Ergonomics and User Interfaces (Vol. 1, pp. 521–525). Proceedings 8th International Conference on Human-Computer Interaction (HCI International ’99, Munich, Germany, August 22–26). Preim, B. (1999). Entwicklung interaktiver Systeme: Grundlagen, Fallbeispiele und innovative Anwendungsfelder (Development of interactive systems: Basics, examples and innovative application area) (pp. 55–74). Berlin Heidelberg: Springer-Verlag. Rasmussen, J. (2000). Human factors in a dynamic information society: Where are we heading? Ergonomics, 43(7), 869–879. Taylor & Francis. Rose,¨ K. (2002). Methodik zur Gestaltung interkultureller Mensch-Maschine-Systeme in der Pro- duktionstechnik (Method for the design of intercultural human machine systems in the area of production automation). Dissertationsschrift zur Erlangung des akademischen Grades, Doktor-Ingenieur’ (Dr.-Ing.) im Fachbereich Maschinenbau und Verfahrenstechnik der Universitat¨ Kaiserslautern, Fortschritt-Bericht pak, Nr. 5, Universitat¨ Kaiserslautern, 2002. Rose,¨ K., Liu, L., & Zuhlke,¨ D. (2001). Design issues in mainland China and Western Europe: Similarities and differences in the area of human-machine-interaction design. In: M. J. Smith & G. Salvendy (Eds), Systems, Social and Internationalization Design Aspects of The Development of Culture-Oriented Human Machine Systems 103

Human-Computer Interaction. Volume 2 of the Proceedings of HCI International 2001, 5–10 August 2001, pp. 532–536. New Orleans, Louisiana, USA. Rose,¨ K., & Zuhlke,¨ D. (1999). Design of global user interfaces: Challenge or frustration? In: H.-J. Bullinger & P. H. Vossen(Eds), Adjunct Conference Proceedings, 8th International Conference on Human-Computer Interaction (pp. 268–269). Stuttgart: Fraunhofer IRB Verlag. Rose,¨ K., & Zuhlke,¨ D. (2001). Culture-oriented design: Developers’ knowledge gaps in this area. In: G. Johannsen (Ed.), Preprints of 8th IFAC/IFIP/IFORS/IEA Symposium on Analysis, Design, and Evaluation of Human-Machine Systems (pp. 11–16). Kassel, Germany, September 18–20, 2001. Rose,¨ K., Zuhlke,¨ D., & Liu, L. (2003). Intops 2: The machine user interface design for the Chinese market. Final Report. Projektabschlußbericht des industriegeforderten¨ Projektes, Design for Mainland China’ im Zentrum fur¨ Mensch-Maschine-Interaktion, Fortschritt-Bericht pak. Universitat¨ Kaiserslautern (in press). Schneider, G. (1997). Entwicklung und Validierung eines methodischen Konzepts zur ergonomischen Gestaltung von Bediensystemen in der Produktionstechnik (Development and validation of a method for ergonomic design of machine user interfaces in the area of production automation). Dissertation, Universitat¨ Kaiserslautern, VDI-Verlag: Dusseldorf.¨ Shneiderman, B. (1998). Designing the user interface: Strategies for effective human-computer- interaction (3rd ed.). Addison-Wesley. Stewart, E., & Bennett, M. J. (1991). American cultural patterns: A cross-cultural perspective. Yarmouth, Maine: International Press. Wahl, M. (1999). Systematische Entwicklung nutzergerechter Maschinenbedienoberfachen:¨ Ein Software-Werkzeug mit integrierten Ergonomie- und Gestaltungsregeln (Systematic develop- ment of user-oriented machine user interfaces: Development of a software tool with integrated ergonomic design guidelines). Dissertation, Universitat¨ Kaiserslautern, Fortschritt-Bericht pak, Nr. 3, Universitat¨ Kaiserslautern. Zuhlke,¨ D., Romberg, M., Eble, B., Rose,¨ K., & Krauß, L. (1997). Intops: Anforderun- gen außereuropaischer¨ Markte¨ an die Gestaltung der Maschinenbedienung. Projekt- Abschlußbericht (Request of non-european markets on the design of machine systems. Final project report), pak. Zuhlke,¨ D., Romberg, M., & Rose,¨ K. (1998). Global demands of non-European markets for the design of user-interfaces. (Proceedings of 7th IFAC/IFIP/IFORS/IEA Symposium, Analysis, Design and Evaluation of Man-Machine-Systems Kyoto, Japan 1998 – 09 – 16–18). Kyoto: IFAC, Hokuto Print: Japan, pp. 143–147. This Page Intentionally Left Blank 4. HUMAN FACTORS ENGINEERING (HFE) AND CULTURAL CALIBRATION FOR VESSEL AND OFFSHORE INSTALLATION DESIGN

Denise B. McCafferty, E. Johan Hendrikse and Gerry E. Miller

ABSTRACT

Since much of the design work for vessels and offshore installations occurs in countries other than where the vessel may operate or where the installation may be located, it is particularly important that the expected user be considered and accommodated through the design and operational phases of a project. Within the framework of engineering design and marine operations, this chapter will discuss “soft” issues, such as organizational and line management decisions and personnel selection procedures, as well as “hardware” issues related to design of living and working environments. In particular, the chapter will address how culture should be considered while identifying “user” needs and requirements.

Cultural Ergonomics Advances in Human Performance and Cognitive Engineering Research, Volume 4, 105–145 © 2004 Published by Elsevier Ltd. ISSN: 1479-3601/doi:10.1016/S1479-3601(03)04004-9 105 106 DENISE B. McCAFFERTY ET AL.

INTRODUCTION

Many seafarers currently come from Southeast Asia, and most oil and gas fields currently being discovered and developed are outside the Gulf of Mexico(GoM) and North Sea. However, the majority of new vessels and offshore installations are now being designed by engineers and technicians who are not from the countries where the vessels and offshore installations are being used or installed. Further, the hardware and software technology being used often originates in industrialized nations and is often exported for use elsewhere. Consequently, there is a significant chance that the technology will not match the expectations and experience of the user population. Maurino aptly described the problem of this mismatch by stating, “Technology is not a good traveler unless it is culturally calibrated” (Kaplan, 1995). The compatibility of the transferred technology with the capabilities and limitations of the intended workers and the expected working environment is very important, but until now, it has not been given adequate attention. Designs for vessels and offshore installations are commonly based on European or North-American standards and data. In the early stages of oil and gas produc- tion, expatriate support staff from industrialized nations may be present and these are often from western cultures; thus, the design fits them well. However, it is very common for owners/operators to choose, or be required, to employ personnel from a developing country for the long-term operation and management of a vessel or offshore facility. This means that the design of the work place may not fit these workers. For the successful transfer of the technology, it is essential that a design be “culturally calibrated.” This means the design may need to be adapted or modified to take into consideration the technological, anthropological and cultural factors of the intended user population. Human Factors (HF) aims to provide the information necessary to promote an acceptable and efficient interaction among people, the equipment they use, and the environment in which they operate. It is a body of knowledge concerned with human capabilities, limitations and other characteristics that are relevant to design (BCPE, 1999). Human Factors Engineering (HFE) is the application of this knowledge to the design of tools, machines, systems, tasks, jobs, and environments for productive, safe, and comfortable operations.

HUMAN FACTORS ENGINEERING

The notion that the human being is the source of most accidents in the workplace has existed for decades. Within the maritime industry, “human error” as the cause of, or contributor to, maritime accidents and incidents was specifically identified Human Factors Engineering (HFE) and Cultural Calibration 107 as early as the late 1950s. Yet, it was not until a series of high profile, maritime accidents occurred involving either vessels or offshore installations that the engineering community, national and international regulatory agencies, marine insurance companies and clubs, marine classification societies, and the general public became aware and concerned about these “human errors” at sea. In 1995, the Unites States Coast Guard (USCG) announced a new approach to be taken to reduce maritime accidents in US-controlled waters (USCG, 1995a,b). The new approach, that of placing greater emphasis on the reduction of human errors at sea, was certainly supported by the research data available, and certain maritime accidents that placed a spotlight on the human error dilemma. Two widely publicized offshore accidents, i.e. the capsizing and sinking of the drilling rig OCEAN RANGER off the coast of Newfoundland in 1982 and the explosion and fire aboard the North Sea platform PIPER ALPHA in 1988, resulted in 251 deaths and the destruction of hundreds of millions of dollars of physical assets. Both of these disasters were related to a series of inappropriate human actions (or lack thereof) that directly caused or contributed to these losses. One year prior to the USCG report, the Secretary General of the International Maritime Organization (IMO) issued a message marking World Maritime Day, stating that 80% of all accidents at sea are caused by human error, and concluded by saying, “if we sincerely want to stop accidents from occurring, then I think it is obvious that we should concentrate our efforts on eliminating human errors” (O’Neil, 1994). Concurrent with this industry-wide awakening of the role that human error was playing in maritime accidents and incidents was an attempt to put a name to these phenomena. The USCG referred to this area of human causation of at-sea accidents as the “human element.” Technical papers of that era also referred to “Human and Organizational Factors” (HOF for short), and the U.S. Navy coined the term “Human System Integration” (or HSI). As with the rose, the problem by any name remained the same: our vessel and offshore installations were not being designed or managed to meet the needs of the marine personnel, nor to be compatible with human physical, psychological, and social capabilities and limitations. Furthermore, the designs did not make allowances for cultural influences. The result: human errors causing major accidents and incidents on vessels and offshore installations were occurring too frequently. Based on the new attention given to the role of human error in maritime ac- cidents and incidents, some companies, especially offshore firms, began to apply HFE to the design of their new installations (as shown in Fig. 1) in the early 1990s. So a new, albeit limited, course began to be charted in the name of personnel safety and environmental protection. Unfortunately, the attitude of many of the individuals and agencies concerned with plotting this new course seemed to be that 108 DENISE B. McCAFFERTY ET AL.

Fig. 1. Deepwater Installation in Gulf of Mexico that Included HFE in its Design. more training, more procedures (or company policies), more planning for safety, and greater reliance on punishment for those who didn’t follow the procedures and policies, was all that was needed to reduce the human-element contribution to maritime accidents. There is nothing wrong with these approaches when appropri- ately applied, but there is more to the human-element problem (and solutions) than just these. In fact, based on twenty years of experience in the pursuit of reducing human error at sea, there appear to be at least eight components that can impact how safely and efficiently a person works and rests in a marine environment. Ignoring any factor may result in a potentially unsafe and inefficient mariner or offshore worker. The proposed model for addressing the eight elements contributing to human error at sea is called the “Triangle of Effectiveness” (Fig. 2). This model not only identifies and defines eight elements necessary for a successful “human- error”-reduction program, it also establishes the most effective and cost-efficient approach toward achieving that goal. This means that the basis for any “human- error”-reduction program should be the establishment of a strong management Human Factors Engineering (HFE) and Cultural Calibration 109

Fig. 2. The “Triangle of Effectiveness” for a HF Program to Reduce Human Error. commitment to, and utilization of, existing HF data in every management decision made and action taken involving those who work in a maritime setting. The next step is to design and construct the workplace to match the user’s capabilities and limitations and cultural background. Included in that design effort is the creation of the physical working environment. Each level of the triangle should then be completed in turn. By following the triangle sequence, from bottom to top, the best return, in terms of accident reduction, money invested, and return on investment, will be realized.

EIGHT ELEMENTS OF HUMAN FACTORS ENGINEERING

Management Participation

Management participation in reducing human errors in the maritime workplace is the essential base to any human-error-reduction program. This participation can be exhibited in many ways, beginning with management’s support for a formal HF program at the most senior levels. This support should be demonstrated in management decisions made involving human-element concerns. To do this, the company should base decisions on known and proven behavioral principles for creating, encouraging, and reinforcing safe and efficient human behavior in work settings. As an example, basic human physiological and behavioral terms (and requirements) such as culturally based anthropometry, circadian rhythms, cultural expectations, spatial relationships, physical capacities, punishment versus positive reinforcement, behavior modification, decision-tree support and closed-loop dynamics should be as familiar to management personnel in shipping and offshore 110 DENISE B. McCAFFERTY ET AL. companies as are economic terms like gross expenditures and net profits. These terms should be as familiar, and as important, since failure to account for these human characteristics in the workplace can surely impact both expenditures and profits, as well as personnel safety and efficiency. Some examples of these physiological and behavioral characteristics translated into daily business decisions include:

Providing uniformly enforced rules for worker conduct regardless of position or tenure in the company.
Creating and using easy-to-read and understand policies and procedures so as to increase the likelihood of employee decision making being in line with man- agement desires.
Establishing work schedules which coincide with human physiological functions (e.g. circadian rhythms).
Establishing tour duty lengths to minimize worker fatigue and burnout.
Creating a high-level risk manager with the authority to aggressively pursue safety issues within the company.
Creating a program whereby personnel identify unsafe acts and monitor them- selves to ensure those acts are eliminated from their daily work routine.
Using positive rewards rather than punishment for encouraging personnel to follow regulations, rules and policies.
Designing facilities to satisfy the workers’ needs (e.g. behavioral expectations) and to eliminate their need to take unsafe “short cuts” which can lead to accidents.
Creating and supporting a policy regarding maintenance of assets, since a well- maintained facility is a safer facility.
Soliciting inputs from personnel when establishing the work rules of the facility (i.e. vessel or offshore installation).
Selecting an approach to personnel management that best matches the cultural expectations of the workers.
Providing an overall work place designed to match the workers’ physical, mental, and social needs based on the specific populations and cultural background of those that will be present.

Workplace Design

Often this aspect of the human-error-reducing program is called ergonomics, or human factors engineering (HFE). Whatever the term, it is imperative to understand that human beings, regardless of geographic origin or ethnic back- ground, possess certain social, psychological, and physiological capabilities and Human Factors Engineering (HFE) and Cultural Calibration 111 limitations. In addition, the culture of the individual also influences expectations and behavior patterns. Therefore, when placing a person into a maritime working environment, it is imperative that the design, arrangement, and orientation of that work environment match the capabilities and limitations of the full range of workers (e.g. male and female, tall and short, young and old, expatriate and local workforce) expected to work in that environment. It is here that the worker’s cultural origin or place of residence becomes a major concern and influence on the design. If ergonomically-based designs are not provided on maritime facilities, other error-reducing elements (e.g. training, personnel selection, job aids, etc.) will cost more, and be less effective. The following statement encapsulates the importance of proper design:

Human errors induced by poor design of the workplace cannot be overcome with more training, more manuals or written procedures, exhortations to work safer, or threats of punitive actions for job accidents (Miller, 1999).

This is especially true when the worker is required to act in stressful emergency situations. Yet, this potential means for human-error-reduction through the application of ergonomics in workplace design is not well-understood by many companies. Lack of attention to effective workplace design is particularly distressing because of the eight human-error-reducing elements. The ergonomics arena is the one that is probably best suited for immediate reduction of human error in the maritime industry. Extensive research on workplace design criteria, and their impact on human behavior and performance in a work environment, has been conducted for at least the past five decades. The results of that research have been translated into several well-accepted human-based design standards or references such as those published by the American Society of Testing and Materials (ASTM F1166–2000), United Kingdom Ministry of Defence (Def-Stan 00–25, 1991), Woodson, Tillman and Tillman (1992), International Labor Office (1990), and the American Bureau of Shipping (ABS Guides–2001, 2002 and Guidance Notes – 1998 and revised 2003 edition). These, or other, human factors design standards could and are being used to improve the design of the maritime workplace whether it be on a vessel, or offshore installation. The ASTM and ABS design criteria were specifically prepared for use in the marine industry. The ASTM standard has been used as the basis for recommended design standards for commercial and military vessels and offshore installations since 1988. Thus, the maritime industry has industry-specific tools to guide the integration of ergonomics with engineering design. It is important to note, however, that the majority of the design standards noted above are principally for worker populations from Europe, North America, and 112 DENISE B. McCAFFERTY ET AL. other “Western” cultures. Similar design standards for mariners and offshore workers from cultures with physically smaller populations (e.g. Japan, Southeast Asia), and with different cultural expectations (e.g. reading from right to left, or relating the color red to mean something other than a hazard) are lacking at present. This presents a clear and significant challenge to the designer of maritime equipment, facilities, and systems to culturally calibrate the design of a vessel or offshore installation to match characteristics of personnel from other cultures. The challenge should not be abrogated just because regional/cultural data are not readily accessible. Rather, designers should be creative in culturally calibrating their design. An example of a calibration performed for a gas facility using Asian workers is described later in this chapter.

Environmental Control

Allowing the physical environment (e.g. temperature, humidity, ventilation, noise, illumination, vibration) to exceed known and established limits can increase the likelihood of an environmentally induced human error to occur on any vessel or offshore installation. Much is known about the effects of environmental factors on human behavior and performance with upper and lower limits now well established for both comfort and job safety under operational conditions. Yet, these limits may be exceeded in the maritime workplace. As just one example, safety statistics from a variety of organizations have shown that crane accidents are the second most frequent cause of injuries and fatalities on offshore installations in the Gulf of Mexico (Miller, 1999). This is not a new problem. As long ago as the 1980s, the U.S. Department of Interior, Mineral Management Service documented crane accidents and the various causes including human factors (Danos & Bennet, 1984). Traditionally, operator cabs on cranes have been not environmentally controlled. Some years ago, during a visit to an offshore installation, one of the authors of this chapter noted that a crane operator had brought an electric fan from home and had it sitting on the floor of the cab blowing on him. This was the only cooling capability existing in the cab. On the inside wall of the cab, the same operator had hung a large thermometer, also brought from home, which on this July afternoon was reading 122 degrees F (50 degrees C). It is believed that this combination of heat and July GoM humidity contributed to an operator error of task omission (i.e. the operator did not complete a task that he was supposed to have done) that day resulting in damage to the crane and platform. Yet, the accident report identified this as “human error” caused by the operator’s “inattention” to his duties. It is important to understand that human exposure to excessive temperature, humidity, noise, vibration, etc. can actually Human Factors Engineering (HFE) and Cultural Calibration 113 induce workers into committing unsafe acts which can result in serious and deadly consequences. A major step forward toward the incorporation of the proper environment for maritime work and rest was taken in 2001 with the publication of the ABS Guide to Crew Habitability on Ships and the sister publication, the ABS Guide for Crew Habitability on Offshore Installations in 2002. There is still debate over whether these standards apply to individuals from all cultures that might work on a vessel or offshore facility. Nevertheless, until research data clearly define cultural differences in response to the environmental limits now recommended by the ABS standards, it is suggested that the ABS standards be used for design of maritime facilities, regardless of the cultural backgrounds of the individuals who will staff those facilities.

Personnel Selection

Some jobs in almost every industry require special human physical, social or psychological skills not possessed by the population at large, or even by persons now working at those jobs. This has been recognized for some years in the public safety sector (i.e. police and fire) where physical and psychological screening tests are administered to eliminate those individuals who are neither physically nor emotionally suited for the rigors of being a law-enforcement officer or fire fighter. In the maritime industry, one could ask, what special physical, social or psychological skills should someone like a crane operator possess? Good depth and color perception would be desirable. Resistance to risk taking and coolness under stress might be a good psychological profile. The same question could be asked about a ship’s captain, an engineer, offshore installation manager, driller or chief mate on a vessel. Does the fact that a person has worked at any of these jobs for twenty years automatically mean that person possesses the unique emotional makeup required to be good at what that person is doing, especially when that person must respond to an emergency? The answer is no. Conversely, are there existing psychological, social or physiological screening devices available to identify potentially good or bad ship’s captain, engineer, driller, crane operator, or ship’s mate? Currently, the answer is also largely no. There are now investigative efforts underway to validate the use of a common psychological test used in the counseling field as a screening tool for identifying individuals likely to make good masters and mates on a ship for U.S. populations. Furthermore, it appears likely that this same, or a similar test could be used to screen for good crane operators or other high-risk specialties amongst American 114 DENISE B. McCAFFERTY ET AL. workers. Finally, there is a company in California which has developed and uses a test to identify potential employees who are more likely to be involved in actions and unsafe acts that could lead to on-the-job accidents than their safer counterparts. This test has been validated and is now in use by some on-shore industries, however it has only been validated for U.S. populations. There is the legitimate question of whether these personnel selection tools, developed for U.S. populations, would be appropriate for individuals from other cultures. What is not in question is that picking the right person for a particular job can pay significant dividends in terms of personnel safety and efficiency. One would expect that this would be true regardless of the culture from which an individual is selected.

Training

The traditional industrial approach to training is to use someone experienced in a job to train others who are not. The length of time devoted to the instructional effort, the training content and format is often informal and determined by the instructor based on his/her personal experiences. In many cases, little or no use is made of training specialists, or the vast amount of general knowledge that is available on how to conduct training programs. The HFE approach is different. First, the specific tasks that the trainee has to learn are identified, and the level of performance desired or needed for those tasks is established. Then, a means is created for measuring when that level of performance is achieved by the student. Only when each trainee has reached the pre-determined level of performance required is that person allowed to take on the job for which he/she was trained. Selection of the training approach, and the training equipment (e.g. part or full simulators) is made to achieve maximum training at the most affordable price. As an example, the maritime industry has expended millions of dollars on constructing full-fidelity bridge operations training simulators. However, there are research data that now show that part-task or less-than-full fidelity simulators may be as, or more, effective in teaching vessel operations, and certainly less expensive. The maritime industry may now be in a position to justify using part-task training devices for particular operations. The best approach toward training would appear to be highly influenced by differences in cultural and geographic influences. As an example, individuals from countries with less developed educational systems might be less able to learn from printed text and formal classroom instruction. Therefore, the development of training materials, and the conduct of training classes could be driven by the background of the expected worker population. Human Factors Engineering (HFE) and Cultural Calibration 115

Interpersonal Relationships

Individuals who work in the maritime industry seldom work in a vacuum, even on a vessel or installation with minimum-sized crews. They give and take directions from others, both on a peer basis as well as a supervisor/employee basis. In past years, the maritime industry functioned on a well-defined caste system. The ship’s master was, and still is, the ultimate authority on a ship. The Offshore Installation Manager (OIM) serves the same function on an offshore installation. Throughout the maritime world there has always been a one-way, top-down, chain of command. However, the maritime world is replete with accidents in which a subordinate crew member detected or suspected that an error had been made by a higher authority but did not believe that he/she had the authority, right, or duty to speak up. As a result, in the western world there is a new push to develop a “team” concept of operation. Here, each member of the crew is considered an equal when it comes to keeping the crew and facility safe. However, there are some cultures where it is considered impolite to correct a superior or to even offer a suggestion to anyone considered of higher rank. Disrespect of one’s superior is a major concern of subordinates and severely limits the input from the bottom up, even where safety is concerned. One approach toward enhancing interpersonal relationships for improved worker safety in cultures that foster equal status and cooperation between workers and supervisors is now being conducted in some shipping and offshore companies. Personnel work together to decide what are and are not acceptable safe behaviors on a particular vessel or offshore facility. The participating personnel take turns as “safety monitors” observing others in their work places and noting those acts that have been mutually agreed upon as being safe and unsafe. The emphasis however, is not on negatively reinforcing (i.e. punishing) their fellow workers for these unsafe acts as a way to eliminate them, but rather to identify and praise people for their more frequent safe acts. With this interpersonal activity, the emphasis is to have one’s peers (not management) positively reinforce personnel for working safely, not punishing them for their infrequent unsafe acts. This approach is based on the behavioral principle that it is better to reward than to punish in order to achieve a response. In doing so, attitudes about safety, not just specific acts, can be altered to make that employee a safer worker.

Job Aids

It is doubtful if any person beyond the age of childhood has not at some time in his or her work or leisure life been confronted with printed or pictorial instructions 116 DENISE B. McCAFFERTY ET AL. on how to accomplish a task and been frustrated by the poor quality of those instructions. Instructions on how to assemble even the simplest toy can sometimes be overwhelmingly complex and difficult to follow. This problem also exists in the maritime industry, and has contributed to human errors. All one needs to do is look at typical life raft launching procedures on a vessel, or read some pig launching instructions on an offshore installation, or try to follow a manufacturer’s operating procedures for what would appear to be the simplest piece of equipment to appreciate how confusing and difficult to understand these instructions can be. Communicating safety or procedural information on a vessel or offshore installation where multi-national crews are employed can be a definite problem. As an example, lack of a common language, differences in interpretation or under- standing of pictorially presented information and variations in the “meaning” of color codes can all contribute to possible miscommunication between personnel. This concern should be considered by those who design or label vessels and offshore installations where multi-cultural crews may work together. This issue is discussed in greater detail later in this chapter.

Fitness-For-Duty

Each day, each person working on a vessel or offshore installation brings to the work place psychological or physical burdens or blessings that can affect the way that person behaves on the job. None of us are the same every day, and this is not a cultural phenomenon. It’s a human one. Sometimes the burdens are externally applied, such as the constant worry over a seriously ill child, spouse, or parent, a financial debt or a marital problem. Other times the burdens are self-imposed, such as drinking or taking drugs, or going without sleep before going to the job. If the burdens prevail, or are severe enough, they can become a safety threat to the worker, and/or others at the job site. When this occurs, the affected individual may need to be removed, at least temporarily, from the job, assignment or work site. The difficulty of course is, other than the obvious substance abuse or broken limb, how do supervisors determine if one of their people is not fit-for-duty on any particular day, and what do you do with such a person if he/she is not? This is a management responsibility and task, but the existence of unfit personnel reporting for work is real and can be a major cause contributor to a human-induced accident or incident. The specific role of culture in this fitness-for-duty issue is unknown. But it can be reasonably expected that in some cultures where employment is limited and there are many who seek it, a worker may come to the job site regardless of his/her physical or mental well-being just to be sure he/she does not lose the job Human Factors Engineering (HFE) and Cultural Calibration 117 because of an absence required for treatment or recuperation. Furthermore, cases exist where job-related injuries have gone unreported (and the valuable lesson of why the injury occurred lost) because workers feared that they would be replaced, either because they felt they would be blamed for the accident, or because they would be held up for ridicule by fellow workers for being so “clumsy.” Although the Triangle-of-Effectiveness model was not initially developed as a cross-cultural model to explain the causes of human error (and how to prevent them if the eight elements in the model are considered to be “defenses” against human-induced accidents), there appears to be no reason why the model would not work for any culture. The important point to be made from the model is that when a vessel or offshore installation is designed and operated, it is imperative that those responsible for the designs appreciate and account for, the different cultures that might be present in the facilities work force. It is also important to appreciate that the cultural differences can be present in the physical, psychological, and social makeup of the crew, and that these differences should be reflected in the way the equipment and facility are designed, the workers selected, the system labeled, the personnel trained, and managed, and the procedures and policies generated and presented. If consideration is not given to the cultural influences on all eight elements of a vessel or offshore facility’s design and operation, that vessel or facility operates at a higher than necessary level of risk due to human-induced errors.

COSTS OF INTEGRATING HUMAN FACTORS ENGINEERING (HFE)

Management has a legitimate concern that HF may add unacceptable costs to the acquisition process through HFE requirements (e.g. those associated with design) and the expense of HFE personnel. However, HFE involvement in several recent offshore projects has demonstrated numerous instances where HFE inputs have reduced total acquisition costs (Miller & McSweeney, 2000). How much does HFE really cost? In no case, according to Miller (1999), who monitored the HFE costs over a nine-year period covering several offshore installation design and construction programs, did the total HF program cost more than 0.12% (and consistently ran closer to 0.08%) of the acquisition cost of the facility. On one offshore Canadian project (i.e. Sable Offshore Energy Project) according to Robertson (1999), although the original estimated cost (personnel charges only) for the HFE program was 0.07% of the facilities budget, the actual HFE cost was 0.035% of the facilities cost, or half of the estimated cost. HFE has a proven track record of returning many times the value of the initial investment throughout the operational life of a project. Miller (1999) reported that proposed HFE changes in the design of the riser tensioner system for an offshore 118 DENISE B. McCAFFERTY ET AL. facility lowered the system’s construction cost by $242,000 as estimated by the company’s engineers. This change was the result of a HFE review of the riser tensioner’s maintenance/replacement requirements. Curole et al. (1999) reported a case where HFE-suggested modifications reduced the time required for the removal of a gas turbine from a compressor package enclosure. The maintenance 1 time was reduced from approximately 10 hours to 3 2 hours and the maintenance task is now conducted in a safer manner. The overall availability or uptime of a facility is a function of the reliability and maintainability of both hardware/software and the reliability and efficiency of the personnel operating and maintaining the system. When the time required for maintenance can be reduced, especially for equipment that requires frequent maintenance or lengthy maintenance periods, as was the case with the removal of the gas turbine, it may improve the availability or productivity of a facility. The impact of each individual reduction in maintenance time may seem small. However, the sum of all maintenance time improvements in conjunction with increased system reliability, which is a function of a reduction in human error, will improve the availability of the facility. Based on historical data, Rensink and Van Uden (1998) discovered that for a typical $400 million petrochemical project, integrating HFE into a new plant design can result in a reduction of 0.25% in capital expenditure, 1% in engineering hours and 3 to 6% of the facility’s life-cycle costs. A review of company pre-and post-HFE involvement in offshore installation design also demonstrated the value of integrating HFE into new capital projects. As an example, the review of accident data (prior to the incorporation of HFE) on the offshore installations of one of the major operators in the Gulf of Mexico (GoM), revealed twenty-seven serious injuries over a seven-year period (3.8 per year) from falls down stairs and vertical ladders. This review effort resulted in a design specification for stairs, ladders and walkways based on HFE research data. This design specification was then used on subsequent offshore installations and only one stair-fall accident (0.25 per year) was recorded over the four-year period after the specification’s introduction. Curole et al. (1999) reported that operators, maintainers, and the commissioning staff stated that the extensive HFE-based labeling program (adopted by a major GoM Exploration and Production company) has been one of the most successful HFE improvements the company has incorporated. This HFE labeling program decreased the commissioning time of one facility by approximately three weeks. However, it is pointed out here and in the following sections that the largest cost advantage will be attained when HFE is incorporated in advance of or concurrent with the design of the facility. Much of the cost advantage associated with the HFE effort will be lost if the designed or constructed facility is subsequently changed to incorporate HFE recommendations. Human Factors Engineering (HFE) and Cultural Calibration 119

BENEFITS OF INTEGRATING HUMAN FACTORS ENGINEERING

As illustrated in the examples above, the return-on-investment in HFE in the offshore industry can be quite substantial. It is also reasonable to expect that HFE can likewise be applied to the shipping industry in a cost-effective manner. In general, some of the benefits that an operator might expect from a proactive HFE program include:
Improved equipment design and controls that can result in fewer accidents, proper operation of equipment, and improved maintainability. This can gen- erate increased availability and productivity, lower maintenance costs, improved personnel utilization, lower personnel exposure time and risk in hazardous areas as well as fewer accidents and incidents.
Improved layout that can result in a better flow of personnel throughout the facility. This is especially important during emergency events. Layout could make the difference in escape, muster, evacuation and rescue times.
Improved human-computer-interface design for computer-generated process, marine display and control screens. This can improve operator information pro- cessing and process control and alarm handling under both normal and upset conditions.
Improved facility design can lead to improved human performance, less physical stress and fatigue, improved quality of work and living, and a work environment which can improve worker satisfaction and morale.
Equipment that is easy to operate and maintain through the provision of properly designed and easily understandable instructions, job aids, operating manuals, and procedures. An additional benefit is the potential reduction in personnel training time requirements.
Reduced exposure to hazardous environments as a result of reduced maintenance and inspection times.

PROCESS FOR INTEGRATING HUMAN FACTORS DURING DESIGN OF OFFSHORE FACILITIES

At the 2nd International Workshop on Human Factors in Offshore Operations held in Houston, Texas in 2002, a white paper was produced on how to effectively include HFE in the design of new offshore facilities (Hendrikse et al., 2002). The white paper provides a strategy for integrating HFE as well as describing the critical HFE activities to be performed. It should be noted that the same strategy 120 DENISE B. McCAFFERTY ET AL. will work for the design of a new or retrofitted vessel. A brief summary of some of the aspects addressed in the white paper is provided below to act as a framework and help guide the cultural calibration process during the design of vessels or offshore facilities.

Typical HFE Tasks/Activities Performed by HF Personnel on a Project

Table 1 listed some typical activities performed by HF personnel, from conceptu- alization to post-commissioning and startup phases of vessels or offshore design projects. These phases will obviously be different depending on the operating company or client as pointed out in the white paper. It is during these activities that it is important to consider what the cultural background of the intended user is and how it will be accommodated during the various project phases. Examples of the types of cultural considerations-per-project phase are as follows.

Concept Phase (Define and Select) During this phase, detailed HFE design requirements that will accommodate the physiological, psychological and sociological characteristics of the anticipated user population should be included in the project specification(s) by those respon- sible for HFE on the project. The Human Factors Engineering Plan describing the scope of HFE work to be completed for the project should identify those activities to be performed that will adequately consider the anticipated user population. All analyses that allow project staff to reach decisions and formulate philosophies for the new facility regarding the health, safety and/or environment of personnel as well as the operability, maintainability and staffing levels should consider the cultural background of the intended user. This should include a type of Gross Task Analysis study or a Front-End Human-Factors-Engineering Analysis.

Detailed Engineering Design Phase During the contractor selection and award contract process, the operating com- pany, with the help of HF personnel, needs to evaluate contractor bid packages for the inclusion/integration of HFE requirements and how they intend addressing the unique cultural needs, capabilities and limitations of the intended user. HF awareness training presented to project personnel should stress physiological, psychological and social characteristics of the intended user population and provide design guidelines on how this should be incorporated during the design of new facilities. HFE design inputs to project engineers during the detail design of the workplace, habitability of accommodation facilities and when creating management and organizational systems for operating and maintaining the facility Human Factors Engineering (HFE) and Cultural Calibration 121 Operational Assessment 1. Provide HFE support during 2. Generate HFE status reports Operations and Maintenance documentation installation, hook-up and commissioning development of training devices and material on-site visits to contractor/vendor fabrication facilities inspection team Commissioning Phase Construction to Operation Phase Human Factors Engineering Activities During Each Project Phase. design documents buildings alarm, and paging system 4. Develop HFE design guidelines and checklists 4. Conduct HFE review of project 6. Conduct a Valve Criticality Analysis7. Conduct HFE review of engineering drawings and 6. Generate HFE status reports 3. Provide HFE training to project team 3. Provide HFE assistance during 8. Conduct HFE evaluation of Vendor9. bid Participate packages in Formal Safety Assessments10. (FSA) Provide HFE input to specialist11. studies Participate in 3D model reviews 12. Participate in design of accommodation13. facilities Participate in design of control room and other 14. Participate in design of fire & gas, general platform 15. Participate in labeling program 16. Generate HFE status reports 2. Determine HFE priorities 2. Conduct HFE review during 1. Contractor selection and award 1. Provide HFE to construction Table 1. Analysis Engineering Plan (HFEP) during concept design studies Project Specification(s) 5. Generate HFE status reports 5. Document all HFE design recommendations 5. Provide HFE support during 4. Conduct a Front End HFE 3. Provide HFE input and review 2. Develop a Human Factors 1. Include HFE requirements in Concept Phase Detail Design Phase 122 DENISE B. McCAFFERTY ET AL. should be based on the capabilities and limitations of the intended user. Examples of such HFE design inputs where cultural calibration is necessary will be discussed later. The evaluation and verification of compliance with HFE design requirements can be achieved by reviewing drawings produced by engineering personnel and also by participating in three-dimensional computer-aided-design (3D CAD) model by reviews, when available. In order to fully support design decisions, HF personnel may be required to conduct specific HFE analyses or studies. In some cases, infor- mation is provided on human capabilities, limitations or unique user characteristics as inputs to studies conducted by Engineering, Operations and Maintenance, or other parties. Examples of such analyses and studies include Function Allocation, Task Analyses and Job Descriptions; Hazard and Operability Studies (HAZOPs) of key areas and operating procedures; Emergency Egress, Escape, Evacuation and Rescue Studies; Material Handling Studies; Control Room Studies and Compe- tency Profiles for the intended user to facilitate their selection and training. A very important activity during this phase is the participation in the design of the alarm and labeling systems on the facility to ensure that HFE design principles have been incorporated and that cultural calibration has been successfully achieved. During procurement, HF personnel should ensure that HFE design specifications, guidelines or checklists are included in Requests for Quotation (RFQs) to ven- dors. Vendor bid submittals should be evaluated for inclusion/integration of HFE requirements and for adherence to these requirements.

Construction/Fabrication Phase At the start of the construction phase, those responsible for HFE should conduct HFE awareness seminars and special training sessions for the construction field staff (especially on-site inspectors and Health and Safety [HSE] personnel). The purpose is to create general awareness of HFE, to familiarize staff with the HFE cultural integration efforts performed during the design phase, and to introduce these personnel to HFE issues that are often created during the construction phase. HF personnel must visit fabrication yards and construction sites in order to audit compliance to HFE design standards and to ensure that the HFE requirements considered during the engineering and procurement phases have been implemented as designed.

Installation Through Startup Phases (Operation) The opportunity to be involved in the installation, hook-up, commissioning and start-up of the new facility, allows HF personnel the opportunity to perform a variety of activities including the verification of operations and maintenance procedures that were developed during design. HF personnel can also obtain Human Factors Engineering (HFE) and Cultural Calibration 123 some valuable user feedback on how well the design accommodates personnel’s capabilities, limitations and cultural characteristics.

CRITICAL SUCCESS FACTORS FOR INTEGRATING HFE

The key ingredients for the successful integration of HFE during the various phases of a capital project are:
Management commitment.
The support of people at all levels of the project organization.
Use of academically trained and experienced HF personnel.
User input during all project phases.
Early focus on known problem areas. These factors will ensure that HFE is executed effectively and efficiently at every stage of the project.

CULTURAL ASPECTS THAT RELATE TO HF IN DESIGN

While the beginning of this chapter has emphasized the integration of human factors with engineering, there are many cultural issues to be considered as a part of HF integration if a vessel or offshore installation is to be safely, efficiently, and economically operated. Important cultural aspects to consider include:
Anthropometry
Physiology and Strength
Materials Handling
Communications
Language and
Other cultural variables.

Anthropometry

There are significant differences in the bodily dimensions, body shape, weight, and proportions of people from different parts of the world. For example, a design range for 90% of the British population for a standing operation would also accommodate 124 DENISE B. McCAFFERTY ET AL.

90% of Germans but only 10% of Vietnamese. It has also been shown that even an adjustable safety helmet designed in an industrialized country cannot be fitted properly by 40% of Sri Lankans (Abeysekera & Shahnavaz, 1987). The various sizes and shapes of workers become particularly important in the selection of personal protective equipment, as is demonstrated in the example of the safety helmet. A respirator mask that may work well for one population can be ill-suited and not fit-for-purpose when used by personnel from a different region of the world. This may be due to different facial contours, the size of the face, or the presence of beards. Clothing, like coveralls and fire suits, designed to fit Scandinavians will be overly large for most Filipinos. Improperly fitted equipment is not merely inconvenient, but can lead to injuries and even be life threatening. In the case of the respirator, an ill-fitting mask will leak and could allow inhalation of noxious or deadly gases. Improperly fitted clothing may be caught in rotating machinery, or cause workers to slip, trip, or fall. One noted incident revealed that fire-fighting personnel were literally “running” out of their fire boots and stepping onto painfully hot decks because the small crew members did not fit well with the larger one-size-fits all fire boots used on a particular vessel. Considerable inconvenience, accidents, injuries and low worker productivity have been shown to be the results of incompatibilities between people and equipment. Workplaces, equipment, tools and protective clothing should fit the physical characteristics of the workers expected to use them. One means of addressing anthropometric concerns is for designers and engineers to obtain such data on the expected users of the vessel or installation. Sometimes it is necessary to use HFE standards or guidelines from industrialized nations and modify the criteria based on research data related to the expected users. For example, during a design effort completed in Houston, Texas for an offshore installation to be installed in the South China Sea, criteria from the American Standard, ASTM F-1166 (2000), were modified using anthropometric data that were available in publications by the ILO (1990), NASA (1978) and in Woodson et al. (1992). The design range for male stature provided in ASTM F-1166 is from 65 in (1645 mm) (5th percentile) to 73.5 in (1871 mm) (95th percentile). The design range for the stature of the expected user was found to be 60 in (1529 mm) (5th percentile) to 68 in (1720 mm) (95th percentile). These data highlighted the fact that the expected workers were significantly shorter than those the designers would normally be designing for on a GoM installation. It was realized that many design aspects would be affected by anthropometrical differences in the two populations. Some of the more important aspects included reach distances, access requirements, and work-surface heights. As a result, it was necessary to develop workspace design guidelines that took into consideration the physical size of the Asian workers as well as modified criteria for valve Human Factors Engineering (HFE) and Cultural Calibration 125

Fig. 3. Installation Designed with Culturally Calibrated Anthropometric Dimensions to Match Expected User Population. locations and orientations. These new criteria were included in vendor bid specifications. Applying modified criteria to design projects is easier with the use of (3-D CAD) software packages (see Fig. 3). The 3-D CAD software can be used to check that adequate space is provided for access to equipment for operations and maintenance and that work surface heights for panels, consoles, and workstations are appropriate. The 3-D CAD software often allows a “Scale Man” to be configured at any stature and size that will match the anthropometric dimensions of the intended user population in order for designers and engineers to check the proposed design against relevant anthropometric dimensions. Anthropometric dimensions also play a large role in the design of the accom- modations such as berth dimensions, and work surface heights in the control rooms or on navigational bridges. Appropriate data should be used to design offices, galleys, storage facilities, laundries, medical facilities, recreational spaces, and dining areas. Anthropometric dimensions should also be considered in the selection of survival craft, and medical stretchers (Fig. 4). 126 DENISE B. McCAFFERTY ET AL.

Fig. 4. 3D CAD Modeling Being Used to Review HFE Design Aspects Depicting a 5% Asian Male and a 95% European Male.

The importance of anthropometric dimensions and how they may vary even within the same culture is further illustrated by a study that was performed by the authors for a company operating in the Gulf of Mexico (GoM) region (See Fig. 5). The industry’s personnel weight standard, for the design of survival craft is 165 pounds with a seat width of 17 inches, as required by the United States Coast Guard (2001). Height, weight and overall size data was obtained for offshore workers from transportation records and through visits to installations. The study found that the “average” GoM offshore worker weighs 210 pounds and requires a seat width of at least 21 inches. Based on the larger physical size of GoM personnel, as compared to U.S. Coast Guard requirements, GoM operators had to reconsider the rated capacity of lifeboats utilized on installations (U.S. MMS Safety Alert 192-R 2001). Some survival craft manufacturers have subsequently down-rated the capacity of their boats when supplied to companies operating in the GoM region. In the beginning, those who were responsible for providing and/or designing workplaces, and the individual equipment, components, systems and structures that constituted those workplaces, gave little thought to creating a workplace that matched the physical size of the workers who would occupy these workplaces. Instead, the work sites were designed primarily to satisfy other engineering requirements, minimize cost and maximize production. Human Factors Engineering (HFE) and Cultural Calibration 127

Fig. 5. Typical Size of Crew Members Now Working on Offshore Installations can Prevent Boat’s Rated Capacity from Being Achieved.

At some point in the evolution of the design of the industrial workplace, the concept that humans come in many sizes, and that designing the workplace to better fit the personnel actually could increase production and profit came into being. Efforts began to collect data on the size, strength, and range of motion of the human body from populations around the world. A discipline dedicated to this effort, called Anthropometry, evolved and eventually significant data on men, women, and children from many nations around the world existed. As a result, it has become generally accepted that human workplaces, and the individual equipment, components and structures that constitute the workplace, should be designed, arranged, and configured to fit the physical size of the full range of personnel who will occupy these workplaces. During the early industrial period, when anthropometry data was used in the design of the equipment and overall workplace, a mistake was made where the arithmetic mean of the body measurements, contained in such anthropometrics databases, was used as “the design criteria.” Thus was born the erroneous concept of designing for the “average man.” It was believed that by designing for the average, most workers would be accommodated or adjust. Today, the approach is to design to fit a range of potential users, typically identified as the 5th percentile to the 95th percentile. A person with 5th percentile body dimensions will serve as the lower design limit since only 5% of the population group to which that person belongs will be smaller in size than the 95th percentile person. In contrast, a 95th percentile dimension means that only 128 DENISE B. McCAFFERTY ET AL.

5% of the population group is larger than that person. Consequently, designing to fit the 5th percentile to 95th percentile population range means that a particular design will fit 90% of the total potential user population. There are three primary ways to design a workplace and individual equipment, components, structures and systems in the workplace, to make them useful and safe for a large range of people. First, design the equipment, structures and systems to be operable and maintainable by a pre-determined “worldwide” population. The anthropometric data created to identify this “worldwide” population is arbitrarily set to include as many potential workers from around the world as is deemed feasible. This has been attempted before, with the notable example of an anthropometrics standard published by the International Organization for Standardization (ISO) (DIN/ISO 3411) to be used for the design of earth-moving machines usable by any popu- lation worldwide. To demonstrate how daunting this task is the anthropometrics standards for standing and seated head heights in DIN/ISO 3411 for the 95th% male would be too short for operators in France, North America, Northern Europe and Australia. Likewise, the leg clearance (i.e. buttock to knee length) for these same 95th percentile males, equipment operators from North Africa, France, North America, Northern Europe, Central Europe and Australia would not be inadequate. Finally, using the ISO workstation dimensions would exclude most of the female population of the world from being earth-moving equipment operators. Therefore, using a single anthropometrics standards created to include “most of the world” is not always practical nor a technically sound approach toward designing workplaces or individual equipment and components that go into these workplaces. The second approach is to design the equipment, structures, and systems to be “adjustable” (e.g. seats move forward/backward, up and down, control panels move in and out, valves somehow rotate to be lower or higher, equipment components to slide together or apart for improved access, etc.). Some of these ad- justments are practical, and are in fact now used (e.g. seat adjustments). However, most adjustments needed to make the equipment structures and systems fit the full range of user anthropometrics as depicted in Table 2 are not feasible, technically or economically. A third technique is to design the workplace, equipment, structures and systems to fit expected user populations. This is achievable if the user population is known, is relatively limited in terms of geographical and cultural makeup, remains constant throughout the lifetime of the equipment, and attention and effort are given during the design phase to matching the hardware and systems to the user population. Fitting the workplace and equipment to the expected workers is made more difficult when little data exist on the user population. Such is the case where Human Factors Engineering (HFE) and Cultural Calibration 129 Males (Dimensions in Millimeters). 190 200 90 100 90 110 50 70 40 80 5th 95th 5th 95th 5th 95th 5th 95th 5th 95th Comparison of Anthropometrics Data From Six Regions of the World For 5th percentile and 95th percentile & Bottom Regions Table 2. RegionNorthern EuropeNorth AmericaNear EastWest AfricaSouth East AsiaLatin America 1710Difference Between Top Stature 1670 1910 1900 1620 1530 1560 900 1520 Sitting Height 890 1800 1720 1790 1000 1710 Forward Reach 830 990 790 760 820 810 Sitting 790 Eye 940 Height 900 880 930 900 910 780 730 790 Knee 770 Height 730 760 860 820 900 870 820 730 680 650 860 505 720 820 780 780 505 600 800 600 485 465 490 465 565 525 570 520 130 DENISE B. McCAFFERTY ET AL. vessels or offshore installations are built for use in countries where minimal or no effort has been directed at creating the kinds of databases needed to fit the equipment, structures and systems to the potential workers who reside in those countries. It is also complicated when the workplace and equipment are used by a succession of worker populations. It is common practice in the maritime industry to build a vessel or offshore installation for an initial user population, and then have it sold several times over its useful life to a variety of other owners who exist in other parts of the world and possess workers with different anthropometric dimensions than the original user population. Difficult as it may be, however, it is incumbent upon the designer to attempt to design the workplaces and individual equipment to fit the potential workers as best as possible. In an attempt to at least provide designers with accurate anthropometric data on populations likely to work at sea, the International Labor Organization (ILO) produced its International Data on Anthropometry in 1990. In an attempt to identify the differences in human anthropometric dimensions, ILO divided the world into twenty (20) geographical/cultural regions. To demonstrate the differences in body sizes from among the world’s population, five body measurements important to the design of a workplace or the equipment, components and structures within those workplaces, from six of the worlds’ 20 regions, are shown in Table 2. The information contained in Table 2 displays data for the region with the physically largest population in the world (Northern Europe), the physically smallest population (Latin America) and several regions with population sizes lying between the two extremes. There are several factors to note from the above table that could impact design of the workplace or the hardware that could be used by personnel from all of the geographic regions:

(1) The difference in size between the largest and smallest populations increases as the individuals increase in size (i.e. go from the 5th percentile to the 95th percentile). This places greater emphasis on using the worst-case situation (e.g. largest potential user population for overhead clearances, berth size, clearance between equipment for maintenance access, and smallest potential user popu- lations for maximum weight lifting and carrying, overhead reach, and forward vision over equipment or obstacles) in establishing workplace and equipment design requirements. (2) Whereas the difference in stature in the standing position is 190–200 mm (7.0–7.75′′) between the tallest and shortest populations, this difference decreases to only 90–100 mm (3.5–3.9′′) when those same populations are measured in a seated position. This means that it could be more difficult to design a workstation or individual piece of equipment requiring the user to Human Factors Engineering (HFE) and Cultural Calibration 131

stand than for a workstation or equipment designed for a seated operator where the workers could come from a multi-cultural workforce. (3) The significant difference in physical size between different cultures clearly mandates a design that does not rely on a one-size-fits-all approach. Provisions should be made for “fitted” personal protective equipment, lifeboat seating, fire extinguishers, lifted or carried equipment, tools and spare parts, and any other item that depends on a proper fit between the hardware and personnel.

It admittedly is difficult to design workstations, and equipment, components, structures and systems that go into the workplaces, to physically fit all of personnel that may eventually use the workplace. Nevertheless, it is important for the designers of such workstations to appreciate the importance of designing to fit a range of personnel so as to maximize worker productivity and safety. This is especially true since most of the vessels and offshore installations currently are being designed in one location for use by personnel from other regions and cultures. It is imperative that designers appreciate the significance of cultural dif- ferences on human body dimensions, strength, and range of mobility and that these differences be considered and accounted for in the design of maritime workplaces and equipment.

Physiology and Strength

Physiological differences such as oxygen consumption rate and working capacity differ among people in various geographical regions. There are significant dif- ferences in work capacities and strength capabilities between people of different ethnic groups and between people in industrialized and developing countries (Abeysekera & Shahnavaz, 1987). Manual materials lifting research provides some limits and modifiers for determining what is reasonable for different populations. Typically, these modifiers are based on a percentage of the maximum acceptable weight of lifts for the North-American worker population. Many North-American companies have set single-person lifting limits of between 20 and 23 kg (45–50 lbs) for male workers and 17 kg (37 lbs) where men and women are employed. These limts are often doubled for two-person lifts. To determine lifting limits for other populations, a modification factor or multiplier can be applied. For example, to compute lifting limits for Asian males, a multiplication factor of 0.80 would be applied to the North American lifting limit (ABS, 2003). The product of this computation would be the weight an Asian male would be expected to lift safely, without injury to himself or damage to the item being lifted. Similar research and modifiers are also available for the carrying of loads. 132 DENISE B. McCAFFERTY ET AL.

With an awareness of the limits and modification factors, it is easier for designers and engineers to plan for manual lifting of loads and also to determine where assisted lifting may be necessary. Realizing that there are differences in physiology and strength between people from different regions, adjustment of work standards, working hours and work rates can be considered as alternative means of protecting personnel. Depending on their country of origin, workers may also be affected by climatic conditions, especially in cases, where changes in location may be frequent. Special attention may be required to ensure that workers stay cool or warm under various conditions, such that their work performance is not adversely affected by indoor or outdoor work duties.

Materials Handling

As mentioned above, the limit of a load to be lifted by a North-American worker is often restricted to a maximum of 23 kg (50 lbs). This limit, however, assumes that such a load would be located between a person’s hips and shoulders and that the load would not be carried for a long distance. When considering materials handling for any population, many factors have an influence. As well as the height of the surface where a load is located prior to lifting, the place where a load will be set down after it is lifted or carried is also important. Other concerns related to personnel performance while lifting include:
Age
Gender
Strength (e.g. male vs. female, weight-to-strength ratios)
Physical condition (e.g. cardiovascular and muscular endurance)
Flexibility (range of motion)
Health habits (e.g. smoking, alcohol consumption, and fitness level)
History of previous injury or predisposition for injury
Materials handling knowledge or training. For some of these variables, modification factors, beyond those previously men- tioned for Asian males, are available (ABS, 2003). Table 3 provides lifting and carrying multipliers. Knowing that human lifting capacities are limited, a materials handling study should be conducted for any vessel or offshore installation to identify the equipment that would need to be lifted by personnel, as well as the mechanical handling devices necessary for performing such assisted lifting activities. The study can use data about bodily dimensions for reviewing the space to be allocated around equipment for maintenance access. Lifting limits and modification factors Human Factors Engineering (HFE) and Cultural Calibration 133

Table 3. Lifting and Carrying Multipliers.

Lifting or Carrying Situation Multiplier

Age – for personnel over 50 years old 0.8 Asian female – Where the lifting or carrying population is Asian females 0.86 Asymmetrical lifting – When a worker must twist the torso 45◦ or more 0.70 Handholds – In the absence of handles or with poor handholds 0.90 Lifting or carrying load size – If the depth of the object exceeds 61 cm (24 in) 0.66 If the depth of the object exceeds 91 cm (36 in) 0.50 If the depth of the object exceeds 122 cm (48 in) 0.33 Limited Headroom – where headroom is limited (e.g., where personnel must 0.65 remain “bent” at the waist) Obstacles – If a lower protruding shelf or other obstacle limits the lifter’s approach 0.66 to the desired surface Temperature – for temperatures greater than 32 ◦C (90 ◦F) 0.88 Work duration – for every hour worked over 8 hours 0.96/hour can be used to determine what items can be lifted and carried by personnel versus what items will require lifting aids. These devices include a pedestal crane, bridge cranes, beam clamps and hoists, A-frames, heat exchanger tube bundle pullers and an assortment of trolleys for transporting smaller units of equipment. Once again, many of these decisions can be reviewed during the 3-D CAD model review, evaluating the space provided around equipment for removal and lay-down using designated assisted-lifting devices. If the staffing level aboard the vessel or installation is expected to be low, there is even more reason for special attention to be given to manual materials handling and the availability of suitable assisted lifting devices. During the previously mentioned design effort in Houston, Texas, a materials handling study was performed. One problem that came to light was that the weights of the proposed portable fire extinguishers were nearing the acceptable lifting limits of the Asian personnel. To ensure ease of handling by the smaller workers during fire fighting, smaller, lighter fire extinguishers were procured for the installation (and their numbers were increased).

Communications

It has long been recognized that ineffective communication has been a contributing factor in many accidents. Potential problems increase in cases where multiple languages are used. The International Maritime Organization (IMO) highlights 134 DENISE B. McCAFFERTY ET AL. the potential for communications problems in the Maritime Safety Committee’s (MSC) Circular 813. This Circular highlights the fact that “low standards of crew training and a lack of knowledge and ability to use a common or working language on board a ship are a key concern.” One maritime industry report (De la Campa- Portela, 2003) states that approximately 20% of maritime accidents and incidents involve communication problems. A breakdown of the communications-related accidents data is as follows:
22.4% were due to language problems
18.4% were due to technical problems
59.2% were considered “problematic” communications. Problematic communications included situations where there was poor commu- nication between a ship’s master and a pilot, problems within a working team, or misuse of communications equipment. Communication problems in the maritime industry are not limited to those between masters and crew but may also occur with other vessels, port personnel, regulatory authorities or the shipping compa- nies’ own shore-based staffs. Communications problems are not limited to verbal communications and are sometimes due to receiving or attempting to use written instructions. As just one example of the need for a cross-cultural communication system, the following warning sign was found in the Hydraulic Pump Room on a Ro-Ro vessel that had experienced a succession of three different owners from three different countries (i.e. Japan, Saudi Arabia, and Greece) and three obvi- ously different languages before being acquired by yet a fourth owner (i.e. the U.S. Government) with an English-speaking crew.

“CAUTION OF ONE PUMP RUNNING When either pump in parallel running is stopped abnormally (or intentionally), the other pump is stopped too. Close six stop valves (1–10) on the side of broken pump and the delivery valve (2–7) or (3–7) of broken pump on valve panel before starting the normal pump only again” To complicate this warning and instructions there were actually four pumps in the space, none of which was identified, and there were no valves identified or numbered. Perhaps not surprisingly, there were warnings and labels found on this vessel in all four languages of the four countries in which the vessel had been registered. While the above accident data and real world examples relate to the maritime industry, similar concerns exist for offshore installations. With more drilling and production occurring in developing nations, mixed-language crews are becoming Human Factors Engineering (HFE) and Cultural Calibration 135 the norm rather than the exception. Some of the tools that HFE can offer to assist owners and operators in the area of communications include the use of standardized protocols for verbal communications, constrained language lists of approved terminology for written procedures, and the use of symbols and pictures rather than printed text for hazard warnings and instructional information.

Language

The huge diversity of languages in the world leads to translation problems, misin- terpretations, misunderstanding and confusion in verbal and printed information. Appropriate language should be used in operation and maintenance manuals to accommodate the users of the technology. Very often the language of the developing countries does not have the technical terminology or it cannot be directly translated into the language of the intended user population. For example, in some of the ethnic languages of Southern Africa, the same word is used for the colors blue, green and black. Cultural misinterpretations and language problems can lead to serious confusion and unsafe situations. Such problems may be associated with the identification of hazards, notification of mandatory safety requirements, identification of prohibited actions, presentation of general information such as life boat launching instructions, identification of safety equipment and evacuation routes, and the location of fire-fighting and other hazard-control equipment. Inspection of vessels and offshore installations clearly shows how variable this information traditionally is between, and even within, companies. Even a cursory review of company, national and international regulations covering these items shows how confusing and inconsistent they have become. This in spite of research that shows that most information (including safety information) is received by an individual through vision. That is why it is so important to communicate safety and procedural information in a visual format that can be understood by all workers on a vessel or offshore facility, regardless of their educational levels or their language or cultural origins. To circumvent the language barrier a good hazard identification and procedural information-presentation system should be based on the following:
Use of pictograms instead of printed text to present the information to eliminate the confusion caused by a language unknown to the user of this information.
Presentation of the information in an easy to see and read format.
Placement of the pictograms so that if they show equipment or any real world hardware the pictogram and real world equipment are both spatially related when viewed by the viewer of the pictogram and the equipment or hardware. 136 DENISE B. McCAFFERTY ET AL.

Company-created pictograms for those hazards, procedures, or situations unique to the company or facility.
Assurance that the pictograms accurately depict the real world equipment or hardware so the reader does not have to “interpret” or “fill in” the pictograms in order to understand the hazard or procedure presented.
Limiting the number of adjacent posted information relating to hazards or pro- cedures to approximately three.
Compliance with existing National and International regulations such as those created by the American National Standards Institute (ANSI), European Union (EU), International Standard Organization (ISO), and International Maritime Organization (IMO).
Compliance with an accepted Human Factors Engineering (HFE) design stan- dard such as the American Society for Testing and Materials (ASTM) “Standard Practice for Human Engineering Design for Marine Systems, Equipment, and Facilities” (ASTM F1166–2000).
Flexibility so new hazards or procedures can be added or removed from the system or temporarily installed or removed once the hazard has been eliminated.
Site-specific and hazard-specific warning notices and labels are needed to prevent exposures to hazards. This means that specific hazards must be identified and hazard warnings individually prepared for each specific hazard at each specific location. Such effort provides the potential for warning notices and labels to be maximally effective.
Use of shapes and colors (e.g. orange triangles, circle with diagonal line, green square, etc.) to identify whether the information provided is an identified hazard, mandatory regulation or requirement, prohibited action, general information, or location of safety equipment or fire-fighting equipment. These shapes and colors are usually defined in a national or international regulation.
Locating the signs and pictograms properly so they can be easily found and seen by workers under all conditions and situations (e.g. low illumination levels, need for rapid identification of emergency response equipment), that necessitate the use of the hazard or procedural information.
Use of materials and mounting techniques so the hazard or procedural informa- tion will be easy to repair or replace, resistant to environmental factors, and have a long (i.e. 7–10 year) life span.
Provision of well-marked escape or evacuation routes for use in normal and low ambient illumination levels. These escape markers should be placed so there is at least one visible from any location on the vessel or offshore installation so as to give a worker immediate information on how to start on the escape or evacuation route. Human Factors Engineering (HFE) and Cultural Calibration 137

Although establishment of a comprehensive hazard and procedural communica- tion system may seem an excessive effort during the design and operation of a vessel or offshore facility, experience has shown its value during both routine and emergency operations. One international oil drilling and service company initiated just such a communication system in 1997 after a lost-time accident investigation revealed that the employee involved did not fully understand the language used on existing safety and warning signs (Van Rooyen & Lace, 1998). Later discussions within the company indicated that this was a common problem, not an isolated event. Consequently, the company undertook a company-wide effort to create a new standard for the preparation, manufacture, and installation of hazard warnings and procedural information designed to cross national language and cultural boundaries reflected in their multi-national workforce. The issue of language is also important when it comes to the design, placement, and orientation of individual equipment component identification labels. It is acknowledged that there is cost associated with the development of labels, especially if they are to be printed in two different languages. As a result, in many cases, only one language is selected. If the “official” vessel or installation language is to be English but the expected user is not fluent in English, then there are several factors to consider. These include the following: Consistency of label format and design – The design and layout need to remain consistent from one label to another; although the message may change, the format should not, so that users develop an expectation and familiarity with a standard presentation method. Consistency of label content – The content of the message should be clear and easy to understand. The words used should be simple. Jargon and colloquialisms should be avoided. Labels should avoid words that have multiple meanings or words that may be unclear to non-native language speakers. Consistency of placement – The placement should also remain consistent. Limited use of abbreviations/acronyms – The use of abbreviations and acronyms is widespread in English-speaking vessels and installations. It is important to re- member that other languages will not use the same abbreviations or acronyms so they may not appear comprehensible to the expected users. Nomenclature – The description of equipment or components should relate to function. The use of colorful terms or nicknames (e.g. pig trap) should be avoided since they may be misinterpreted by or offensive to non-English speaking readers. 138 DENISE B. McCAFFERTY ET AL.

Psychology

Psychological differences that affect a person’s task performance include cogni- tive differences, variation in skills, and differences in perception and cognitive models. There can be a large gap from existing knowledge and skill levels to those required for successful operation and maintenance of the new technologies found on today’s vessels and offshore installations. Training should be adapted to local practice in order to establish the actual level of knowledge on which to build the newly required technical skills, and to train the skills themselves (Kaplan, 1995). People from different countries may have different cognitive models employing different operational images and stereotypes. For example, a water tap may work in the clockwise direction to increase flow in some countries, but the opposite direction in other countries. The meanings of colors also change from one region of the world to another. For example, red may mean “danger” in some cultures under some conditions, while in other parts of the world, red may be interpreted as “safe.” (Abeysekera & Shahnavaz, 1987) In other cases, workers may perceive colors of red, green, and yellow to have meanings similar to Western expectations, but people from these same cultures have different perceptions and expectations about colors such as orange, blue, and white. The way people think, feel and react is affected by their culture. Culture consists of traditional ideas and their attached values. Peoples’ cultural backgrounds determine their attitudes to work, technology and organizations, work habits, work group dynamics, religious beliefs and customs. The effects of cultural variables can have wide implications.

Other Cultural Variables

There are numerous cultural variables that may affect the workplace beyond those mentioned previously. Those from Western cultures sometimes do not recognize that such differences exist and bring their views and ideas to workplaces outside their own countries and then wonder why work-related problems occur. Given the number of companies from the United States and Canada that have facilities in Mexico, an attempt was made via a website (www.mexconnect.com) called Mexico Connect to educate others to certain cultural differences. A table from the website has been reproduced here (see Table 4), to highlight how some variables may be different from nations that are geographically close, yet culturally different. Human Factors Engineering (HFE) and Cultural Calibration 139

Table 4. Cultural Comparisons.

Aspect Mexico Canada/USA

Family Family is the first priority Family is usually second to work Children are celebrated and sheltered Children often minimally parented, are independent Wife fulfills domestic role Wife often fulfills dual roles Mobility is limited Mobility quite common Religion Long Roman Catholic tradition Mixed religions Fatalistic outlook. “As God wills” “Master of own life” outlook Education Memorization Analytical approach Emphasis on theoretical Emphasis on the practical Rigid, broad curriculum Narrow, in-depth specialization Nationalism Very nationalistic (U.S.) Very patriotic Proud of long history and traditions Proud of “American way of life” Reluctant to settle outside Mexico Assumes everyone shares his/her materialistic values (Canadian) Less than U.S. Often has more “World” view Personal Sensitivity Difficulty separating work and Separates work from personal relationships emotions/personal relationships Sensitive to differences of opinion Sensitivity seen as weakness Fears loss of face, especially publicly Tough business front Shuns confrontation Has difficulty with subtlety Etiquette “Old world” formality Formality often sacrificed for efficiency Etiquette and manners seen as “Let’s get to the point” approach measure of breeding Personal Appearance Dress and grooming are status Appearance is secondary to symbols performance Status Title and position more important Money is main status measure than money in eyes of society and is reward for achievement Aesthetics Aesthetic side of life is important No time for “useless frills” even at work Ethics Truth is tempered by need for Direct Yes/No answers given and diplomacy expected Truth is a relative concept Truth seen as absolute value 140 DENISE B. McCAFFERTY ET AL.

As shown in the Cultural Comparisons provided in Table 4, education is one area where different standards and educational methods may exist from one culture to another. Educational standards for similar professions in different countries can differ greatly. The knowledge of an engineer in an industrialized country can, for example, be much broader than that for an engineer from the same discipline in a developing country, whose knowledge may be more in depth in that specialization (Abeysekera & Shahnavaz, 1987). The whole emphasis of education may be different, as well, with one culture stressing theoretical constructs while another stresses practical application and problem solving. Beyond education, familiarity with industrialization may have an impact on understanding the concepts of compatibility, display-control relationships and the operation and maintenance of modern equipment. In areas where people have not had exposure to technology or automation, building workplaces that require high levels of technical skill can result in problems. Differences in the standard of living affect attitudes towards work and life in general which may have a negative impact on technology transfer. Many of these are highlighted in the Cultural Comparisons table. Workers in developing countries may view work unfavorably due to heavy domestic responsibilities, inadequate infrastructure and transport systems, and poor wages. Often, the primary purpose of work is to survive and job satisfaction can be generally low. Workers may seek the sense of self not through work but through family and social relationships. Low wages may drive people to seek additional sources of income (Abeysekera & Shahnavaz, 1987). Managerial and organizational methods adopted in industrialized countries may be inappropriate when applied in developing countries. The different human, cultural, social and behavioral patterns of the people in developing countries demand that management systems of industrialized countries require prior adjustment or adaptation before they are transferred. Power and hierarchical decision making, individual vs. collective practices, and the perceived roles of men and women can differ and may require consideration when technology is transferred (Abeysekera & Shahnavaz, 1987). Once again, the Mexico Connect website (www.mexconnect.com) outlines Comparative Management Styles, as summarized in Table 5, that may be found between Mexico and the United States and Canada. Lack of attention to differences in cultural attitudes or practice may lead to workplace problems. For example, if Western designers do not consider certain cultural or religious requirements for workers from other regions of the world when designing galleys and food preparation and storage areas, difficulties may arise. It may also be difficult for expatriate managers to determine the source of discontent among the crew if they expect that a complaint will be verbalized to them. Human Factors Engineering (HFE) and Cultural Calibration 141

Table 5. Comparative Management Styles.

Aspect Mexico Canada/USA

Work/Leisure Works to live Lives to work Leisure considered essential for Leisure seen as reward for hard full life work Money is for enjoying life Money often end in itself Direction/Delegation Traditional manager is autocratic Managers delegate responsibility Younger managers starting to and authority accept and delegate responsibility Executives seek responsibility Subordinates used to being and accept accountability assigned tasks, not authority Theory vs. Practice Basically theoretical mind Basically pragmatic mind Practical implementation often Action-oriented problem-solving difficult approach Control Still not fully accepted Universally accepted and Sensitive to being “checked practiced upon” Critical feedback expected and Sensitive to giving and receiving discussed critical feedback Loyalty Mostly loyal to superior (person Mainly self-loyalty rather than organization) Performance motivated by Beginnings of self-loyalty ambition Staffing Family and friends are preferred Relatives usually barred – No due to trustworthiness nepotism here! Promotions based on loyalty to Promotion based on performance superior Competition Avoids personal competition; Enjoys proving self in favours harmony at work competitive situations Training & Development Training highly theoretical Training concrete, specific Few structured programs Usually structured programs Time Relative concept Literal imperative Deadlines flexible Deadlines and commitments are “Manana can mean not today” firm What is happening now is more What is happening now is only important than the future important if it contributes toward the future Planning Mostly short term because of Mostly long-term in predicted uncertain environment and sense environment of “now” 142 DENISE B. McCAFFERTY ET AL.

CULTURAL CALIBRATION

There are different approaches that may be used for a new vessel or offshore installation to be operated in an area other than where it is designed or where it is likely to be operated by a culturally diverse crew. One means is to design from first principles using data available about the target population. This approach is possible for most Western or developed nations. However, when designers are undertaking a project for a developing nation, data on the population may not be comprehensive. In this case, it may be necessary to recalibrate an existing design to accommodate new users. This will require using any data available for the target population, then modifying an existing design based on that data. The greatest benefit of this approach is that the original concepts and compromises that were documented about the existing design can be reviewed in light of differing cultural requirements. Regardless, the resulting designs should be checked at each available opportunity to ensure that any errors are not built in as a product of the redesign efforts. The resulting product should also be verified by representatives of the end user population prior to being put into use. When attempting to create a new design or modify an existing one that takes into account cultural needs, one problem that designers and engineers may encounter is a lack of adequate anthropometric and physical strength data for target populations. The data that are available may only include a few basic dimensions, and often information on women is absent. The recent publication of new anthropometric and strength data from the U.K. Department of Trade and Industry, including the book entitled, Adult Data, has made this task easier. Also, in the 2003 revision of the ABS Guidance Notes for the Application of Ergonomics to Marine Systems, information has been provided to allow calibration for different populations. One section of these Guidance Notes provides methodologies for determining manual materials handling limits for various populations and genders. An Appendix provides principles and data related to modifying anthropometric values to meet the requirements of different populations throughout the world. Another source of anthropometric data is the International Labour Organization’s book entitled “International Data on Anthropometry” (1990). Even with these data sources, acquiring information for different populations of workers can be difficult. Obtaining information on social factors and management styles can be even more difficult than finding anthropometric data. Given the scope of the task of designing for different world populations, if consideration of the effect of cultural factors occurs late in a project’s timeline, it may be difficult, if not impossible, to locate reference material in time to be put to practical use. As was pointed out by Kaplan, “reports and writings linking cross-cultural and national considerations to problems requiring resolution in a broadly conceived ergonomics and its ancillary domains are scattered nationally and worldwide in Human Factors Engineering (HFE) and Cultural Calibration 143 widely disparate sources” (Kaplan, 1995). Furthermore, appropriate keywords and descriptors of cultural issues do not bring up items in current databases, and data are extremely difficult to find.

CONCLUSIONS

Not all of the preceding cultural aspects are necessarily applicable or relevant in all cases. In fact, there are many factors that have not been discussed. In many cases, technology is engineered in an industrialized country to be operated and maintained by workers from a developing country. However, many of the same cultural aspects should be considered for technology being transferred between industrialized countries. The same cultural aspects exist when a product is built or produced in a developing country for the market in an industrialized nation. Regardless, for technology transfer to be successful, it should suit the culture of the user. Implementation of cultural calibration, however, is difficult due to the lack of comparative information on ethnic or cultural variability. To be successful, consideration of cultural calibration should occur in the very early stages in any project that involves technology transfer. Some suggestions that can assist designers and engineers during “cultural calibration” include the following:
Use whatever information is available to amend culturally-specific design guide- lines and standards to include culturally appropriate data. It may be necessary to develop new design guidelines or requirements if none are available. In any case, assumptions made should be documented.
People from the intended user population should be involved in the design throughout the facility life cycle.
The design of vessels and installations should be optimized whenever possible. Relying solely on training and/or selection of the right person for the task to com- pensate for design limitations is risky since differences may exist in educational and training standards between different regions. In some cases, the operator or maintainer that the designer has in mind may not exist in another part of the world.
The implications of language need to be considered very carefully and special accommodations may be necessary to find alternative methods to convey infor- mation to multi-national or mixed-language crews.

REFERENCES

Abeysekera, J. D. A., & Shahnavaz, H. (1987). Ergonomics of technology transfer. International Journal of Industrial Ergonomics, 1, 265–272. 144 DENISE B. McCAFFERTY ET AL.

American Bureau of Shipping (1998, revised 2003). Guidance notes for the application of ergonomics in marine systems. Houston, TX: Author. American Bureau of Shipping (2001). Guide for crew habitability on ships. Houston, TX: Author. American Bureau of Shipping (2002). Guide for crew habitability on offshore installations. Houston, TX: Author. American Bureau of Shipping (2003). Guidance notes for the ergonomic design of navigation bridges. Houston, TX: Author. American Society for Testing and Materials (2000). Standard practice for human engineering design for marine systems, equipment and facilities (ASTM F 1166–2000). West Conshohocken, PA: Author. Board of Certification in Professional Ergonomics (BCPE) (1999). Certification policies, procedures and practices. Candidate Handbook – 1999 Edition. Bellingham, WA: Author [also see www. bcpe.org]. Curole, M. A., McCafferty, D. B., & McKinney, A. E. (1999). Human and organizational factors in deepwater applications (OTC 10878). Proceedings of 1999 Offshore Technology Conference, Houston, Texas, May 3–6. Danos, W., & Bennet, L. E. (1984). Risk analysis of crane accidents. OCS RPT MMS 84–0056. Metaire, LA: U.S. Dept. of Interior. De la Campa-Portela, M. (2003). Communication breakdown. Marine Engineering Review (Dec/Jan), 34–35. Hendrikse, E. J., McSweeney, K., Atkinson, P., Miller, G. E., Connor, G., Poblete, B., Meyer, R., O’Connor, P., & Heber, H. (2002). Effectively including human factors in the design of new facilities. Proceedings of the 2nd International Workshop on Human Factors in Offshore Operations. Houston, TX, April 8–10. International Labor Organization (1990). International data on anthropometry. Occupational Safety and Health Series: No. 65. Geneva: Author. International Organization for Standardization (2000). Earth-moving machinery-Human physical dimensions of operators and minimum space envelope (DIN/IS0 3411:1995). Geneva: Author. Kaplan, M. (1995). The culture at work: Cultural ergonomics. Ergonomics, 38(3), 606–615. Miller, G. (1999). Human factors engineering: What it is and how it can be used to reduce human errors in the offshore industry (OTC 10876). Proceedings of 1999 Offshore Technology Conference. Houston, TX, May 3–6. National Aeronautical and Space Administration (1978). Anthropometric Source Book (NASA- Reference Publication 1024). Washington, DC: Author. O’Neil, W. A. (1994). Better standards, training, and certification – IMO’s response to human error – A message from the Secretary General of IMO on World Maritime Day. IMO News (3), i–ix. Robertson, N. (1999). Starting right: Sable offshore energy project’s HFE program (OTC 10877). Proceedings of 1999 Offshore Technology Conference, Houston, TX, May 3–6. Rensink, H. J. T., & Van Uden, M. E. J. (1998). Human Factors Engineering in Petrochemical Projects Part I. PTQ Summer 1998 and Part II. PTQ Spring 1999. U.K. Ministry of Defence (1991). Human factors for designers of equipment, Part 4: Workplace Design (DEF STAN 00-25 (Part 4/Issue 1). Glasgow: Ministry of Defence, Directorate of Standardization. U.S. Coast Guard, Department of Transportation (1995). Prevention through people. Washington, DC: U.S. Government Printing Office. Human Factors Engineering (HFE) and Cultural Calibration 145

U.S. Coast Guard, Department of Transportation (1995). Coast Guard Regulation 46, Sub- chapter I. 160.035-2. General requirements for lifeboats (Code of Federal Regulations, 46CFR108.035-2). Washington, DC: U.S. Government Printing Office. Van Rooyen, J. F., & Lace, N. (1998). A new language in rig communication and safety signs designed to cross national language and cultural boundaries, enhancing rig safety information and hazard. Proceedings of 1998 Society of Petroleum Engineers Conference, Caracas, Venezuela, June 7–10. Woodson, W. E., Tillman, B., & Tillman, P. (1992). Human factors design handbook Information and guidelines for the design of systems, facilities, equipment and products for human use (2nd ed.). New York: McGraw-Hill. This Page Intentionally Left Blank 5. CROSS-CULTURAL FACTORS IN AVIATION SAFETYଝ

Ashleigh Merritt and Daniel Maurino

ABSTRACT

This paper presents the safety case for the consideration of cross-cultural factors in aviation by focusing on cultural interfaces, those situations where members of one culture encounter people or artifacts from other cultures. Global aviation is strongly influenced by the USA and Western Europe as the largest manufacturers and largest customers; hence almost all cultural interfaces are weighted in favor of the dominant users. The challenge for safety is not to ignore or eliminate these interfaces but to manage the potential threats they pose. To move forward, there is a role for those inside and outside the dominant model. INTRODUCTION

Unlike no other technology before or since, aviation is responsible for creating the “global village.” It is now possible to reach any part of the world by air, and to do so in previously unimaginable time. Whereas international travel had previously been a privilege of the wealthy elite or the adventurer, today’s world is accessible to more and more people. Business, leisure, even religious pilgrimages are now achieved with the help of air travel. Aviation has changed the way we think about the world and what is possible in the world.

ଝ The contents of this chapter were originally developed to be included in a Human Factors Digest for the International Civil Aviation Organization. Cultural Ergonomics Advances in Human Performance and Cognitive Engineering Research, Volume 4, 147–181 © 2004 Published by Elsevier Ltd. ISSN: 1479-3601/doi:10.1016/S1479-3601(03)04005-0 147 148 ASHLEIGH MERRITT AND DANIEL MAURINO

Civil aviation is the global success that it is today because of its dedication to improving safety. This pursuit of improved standards has advanced the industry in several directions. First and foremost, brilliant minds have created ever more sophisticated machinery – today’s aircraft are extraordinary marvels achieving speeds and load factors never before imagined. A second avenue of improvement has focused on the human in the aviation system. Hawkins’ SHEL model placed the human at the center of the aviation enterprise, and reminded us that the best machinery must still be handled by humans (Hawkins & Orlady, 1993). Crew Resource Management (CRM) expanded the Human Factors horizon from individuals to teams (Wiener, Kanki & Helmreich, 1993), and Reason’s model of Organizational accidents went one step further to show how policies and activities at the management level can impact safety-related activities throughout an airline, including in the cockpit (Reason, 1997). This dedication to improving safety has led to the investigation of cross-cultural factors in aviation safety. There is a prevailing way of doing business in global aviation that has been shaped in large part by the manufacturers and the largest customers. The industry’s standards and practices have been shaped via competitive deregulation, professional interest groups and resource-rich investment in technology. The outcome of this influence and investment is the outstanding safety record that civil aviation now enjoys. Despite this overall success, some regions of the world do not enjoy the same high safety record as others, prompting the inquiry as to why. International standards and practices should be equally relevant, equally applied, equally enforced, and equally affordable around the world. However, the Safety Oversight Audit Program of the International Civil Aviation Organization (ICAO) has clearly demonstrated this is not the case (ICAO, 2000). Regional accident rates vary dramatically around the globe, suggesting that the prevailing model of aviation practice may not be equally applicable around the world. Under- standing these variable safety statistics more completely was the impetus for this document. This paper presents the safety case for the consideration of cross-cultural factors in aviation. It does so by focusing not on cultures per se but on cultural interfaces, those situations where members of one culture encounter people or artifacts from other cultures. Put at its most simple, as long as one stays within the bounds of one’s culture, all of the advantages of cultural membership hold; fellow members and the environment are predictable, thereby making daily routines easier and quicker. But as soon as we encounter members or artifacts (aircraft, procedures, regulations) from other cultures, these cultural efficiencies are challenged and the opposite occurs: The environment becomes less predictable, more uncertain, and requires more cognitive effort. In time, and with sufficient exposure, new habits Cross-Cultural Factors in Aviation Safety 149 emerge to deal with the cultural interface. In civil aviation today, cross-cultural contact is the norm rather than the exception. In such a global context, cultural interfaces are a daily reality. To illustrate the safety case surrounding cultural interfaces, the paper builds upon three existing industry conceptual models. First, Hawkins introduces the notion of interfaces and notes their relevance for Aviation Human Factors. Second, Reason’s model of organizational accidents broadens the Human Factors horizon to include organizational factors distant from yet influential upon the cockpit. The safety case surrounding cultural interfaces in aviation seeks to broaden the horizon even further, showing how members of one culture can incur confusion, misunderstanding, and misapplication when encountering members or artifacts of another culture. In this sense, cultural interfaces can be seen as latent conditions. Lastly, the Threat and Error Management (TEM) Model (Helmreich, Klinect & Wilhelm, 1999) provides a framework for “seeing” the cultural interface in the operating context. It can determine which types of cultural interfaces are the most problematic for a specific context; it can also study the threat management strategies employed by aviation personnel within specific contexts to manage these interfaces. Successful solutions can be shared with the industry.

CULTURE, CONTEXT, AND CULTURAL INTERFACES IN AVIATION

In a broad sense, culture can be defined as the ongoing interaction of a group of people with their environment. The environment shapes the responses of the people, and the people’s responses in turn modify the environment. A culture develops and changes due to three processes – technological and physical changes in the environment; changes in the internal dynamics of the social system, and historical circumstances that are fortuitous or serendipitous (Fig. 1). A simple but significant example of technological change affecting a culture is the invention of the mechanical clock. The first town clock was installed in a town in France in 1314; many public clocks appeared throughout Europe shortly thereafter. Prior to this invention, the concept of time was simple, general, and loose. With the widespread use of the mechanical clock, time began to be concep- tualized as a succession of measurable units. The “working day” was invented and announced by the bells of the town clock. The invention of the airplane has been no less profound; it has changed the way people conceptualize the world and their “neighbors.” 150 ASHLEIGH MERRITT AND DANIEL MAURINO

Fig. 1. How a Culture Develops and Changes.

Some Definitions of Culture A shared system of beliefs (what is true), values (what is important), expectations, and behavior meanings (what is implied by engaging in a given action) developed by a group over time in order to meet the requirements of living and operating in a particular (geographical) niche. (Psychologist) It’s what I expect of myself and what I expect of others in the groups in which I live and work. (Medical practitioner) It’s the human-made part of the environment. (machines, buildings, technology) (Anthropologist) It’s the way we do things around here. (customs, procedures) (Business CEO)

The longer a group stays together, works together, or shares common goals to- gether, the more its members will discover and share acceptable solutions to com- mon problems with each other. The more the members share solutions, the more they will start to think and act alike. This ongoing adaptation is the basis of culture. This logic holds whether it’s a national culture, an organizational culture such as an airline, or a professional culture such as pilots. Because a culture is shaped by its environment and evolves in response to changes in that environment, culture and context are really inseparable. This rather broad definition is deliberately adopted here to allow the greatest scope in understanding culture and cross-cultural influences on aviation safety.1 Cross-Cultural Factors in Aviation Safety 151

Culture and Context

There are many layers of context in which a culture is embedded, including the political, the physical, the social, and the economic. Figure 2 shows the four contexts in which all aviation endeavors occur. Any airline finds itself influenced by these contexts; its daily operations are the solutions to the problems and demands created by these contexts.

Economic and Political Context This layer includes several macro-level features, including:
National wealth, per capita income, and tax base
Population size and density
Stability of economic and political systems
Laws
Education Together these factors decide the size of the aviation customer base in a country and the general affordability of air travel for people in that country. It also decides the size and stability of government support for the aviation infrastructure. The level of education in a country determines the population’s familiarity with technology and their ability to work with it or create their own technology.

Fig. 2. Civil Aviation: A Complex System. 152 ASHLEIGH MERRITT AND DANIEL MAURINO

Geographic and Physical Context Perhaps the easiest to see is the physical context. The features that define this context include:
Geography and complexity of terrain
Climate and weather
Population dispersion and accessibility Combined with other factors such as customer demand, these factors decide flight routes and schedules.

Social Context The third layer is most closely related to the people and their customs. They include:
Socially agreed upon ways of making sense of the surrounding environment
A group’s preferred ways of thinking, acting, and interacting
Values – what is considered important
Behaviors – what is considered normal These features develop over time, handed down from one generation to the next, evolving into a social system of shared meanings and coordination among a culture’s members. These shared understandings make living and working more ordered and predictable. The social context is what allows people to manage their daily lives so efficiently by providing many short cuts based on familiarity. One need only be a tourist in a new land, driving on the opposite side of the road or buying groceries, to recognize how mentally demanding the unfamiliar can be. Similarly, flying into a new airport or dealing with Air Traffic Control (ATC) in foreign countries can be very demanding compared with flying in local familiar airspace. Humans function more efficiently in their own social context because generally they know what to expect from others and they know what is expected of them. And usually, as long we stay within our cultural bounds, we are correct in our expectations. Values – what is important, what is correct – are also shaped within the social context. Values are held so deeply that most people are surprised, shocked, and even offended when they encounter people with values different from their own. Because they are held so deeply and emotionally, values tend to be unquestioned and resistant to change. Some values that drive daily aviation operations throughout the world, and which likely differ throughout the world, include:
How fast should be the pace of life and work? How fast is too fast; how slow is too slow? Cross-Cultural Factors in Aviation Safety 153

How important is technology vs. humans? Who or what has the greater authority, flexibility, “wisdom?” Do I trust the machine or do I trust the person?
Similarly, how important are rules and regulations vs. human discretion? Who or what has the greatest authority, the greatest “wisdom?”
What is the acceptable level of safety and related risks? Is one accident per million departures reasonable given the level of safety in other transports and industries? Should it be higher? Should it be lower?
What is the appropriate way to use and interact with authority? Does everyone have a legitimate voice, or only those with experience and seniority? The social context is what people most associate with culture because of the customs, norms, values, and beliefs. Historically derived and shaped over time in response to environmental demands and changes, the social context is every bit as important as the economic, political, and physical contexts in determining how people will act and react.

Airline Operating Context The features of the airline operating context are familiar to all aviation personnel. This layer includes:
The international and local regulations governing aviation traffic
The airline’s policies and procedures
Aircraft
Manuals and other documentation
Maintenance, and ground equipment
Air Traffic Control
Airports
Passengers and their expectations Figure 2 shows that this context is impacted by all three outer layers discussed above, yet many of the features in this layer are similar the world over. Airlines tend to fly the same planes and adhere to the same procedures and regulations despite the fact that they operate in different economic, political, social, and physical environments. This is an important point that speaks to the heart of this paper. Why do airlines the world over look at least superficially similar? What are the safety consequences of this apparent similarity? This issue will be discussed at length later in this paper. To summarize this section, a culture can be considered the ongoing interaction of a group of people with their environment. It can be understood more completely by paying attention to economic, political, physical, and social factors. 154 ASHLEIGH MERRITT AND DANIEL MAURINO

Cultural Interfaces

There are many advantages to being a member of a culture. Fellow members and the environment are more predictable, thereby making daily routines easier and quicker. There is security in this knowledge, and this confidence furthers one’s efficiency in the culture. But as soon as we encounter members or artifacts from another culture, these cultural efficiencies are challenged and the opposite occurs. The world becomes less predictable, more uncertain, and requires more mental effort. We need to “think on our feet” to make sense of the surrounding context, rather than taking it effortlessly for granted. In time, and with sufficient exposure, new habits emerge to deal with the interface. Understanding new habit formation and its implications in human interactions is at the heart of cross-cultural endeavors.

Cultural interface Members of one culture come into contact with members or artifacts of another culture. Artifact Any human-made object, e.g. airplanes, navigation aids, SOP’s, training manuals, regulations

By its very nature, aviation is a cross-cultural endeavor. Pilots fly in foreign airspace, transporting passengers and cargo around the world. Even when flying within their country of origin, many pilots fly aircraft designed and built in another part of the world. The cultural interfaces in aviation are many and diverse. Table 1 provides a non-exhaustive list of these cultural interfaces, highlighting the fact that it is not only when people interact with people, but also when people interact with products of other cultures that uncertainty can be introduced.

Moderating Factors Given the daily occurrence of cross-cultural interactions in aviation, there are several factors that moderate the experience, including cultural distance, resource match, and experience at the interface. Cultural distance explains how some countries can be culturally similar, despite the geographic distance, because of a common language, shared history, similar religion, or political systems, while other geographically close countries remain culturally distant. For example, Australia and the USA are culturally closer than the USA and Mexico. The larger the cultural distance between two countries, the greater the uncertainty there will be at the cultural interface. Hence a Chinese regulator trying to understand a (translated) Australian accident report is likely to Cross-Cultural Factors in Aviation Safety 155

Table 1. Cross-Cultural Interactions: A Daily Occurrence for Aviation Personnel.

Members of Culture 1 Members & Artifacts Interactions at the Interface of Culture 2

Pilots People Flight Attendants Pilots Multicultural cockpits, airline mergers, alliance/contract pilots Engineers ATC Language exchanges in foreign airspace Airline & Airport Personnel Flight Attendants Different training, goals, and expectations Accident Investigators Ground Personnel Different priorities; confusion around standards Local Regulatory Body Passengers Different expectations for service Trainers Training in another country on new aircraft or as a cadet Investigators International investigation teams Artifacts Manuals Manuals written by personnel in one part of the world are translated or interpreted in other parts of the world. Aircraft Equipment specifications that make sense in one part of the world may not be so “logical” in other parts of the world Maintenance & engineering proficiency as defined and schooled in one part of the world may not be possible in other parts of the world Regulations Regulations created for a specific economic, political, and social context are interpreted and applied in other parts of the world Legal Systems Different paradigms of accident investigation and accountability

Note: Language difficulties apply across all domains. experience more uncertainty about its meaning than a Canadian reading the same document. A second difference relates to resource similarities in the operating context. To the extent that the social and economic context differs between two cultures, the greater will be the mismatch at the interface. For example, the latest safety advance from the aircraft manufacturer may bemuse a pilot from a developing country, given that the local infrastructure is in no position to support it. The great talent of humans is their ability to adapt. The more experience at the interface, the less the interface will appear challenging. A pilot flying for the first time into a new airport in a distant country experiences the interface more 156 ASHLEIGH MERRITT AND DANIEL MAURINO

Table 2. Four Options at a Cultural Interface.

ASSIMILATION INTEGRATION A > BA+ B Members of Culture A learn and adopt Culture Members of Cultures A and B learn each other’s B’s ways, in whole or in part ways & compromise ASSIMILATION SEPARATION B > A A|B Members of Culture B learn and adopt Culture Members of Cultures A and B ignore each A’s ways in whole or in part other’s ways & do not change acutely than a pilot who regularly travels in and out of that airport. Experience at the interface builds familiarity and reduces uncertainty. Hence pilots flying internationally for multicultural airlines are able to develop the necessary familiarity while maintaining a heightened awareness of cultural differences. With experience and exposure comes adaptation. However, adaptation is mostly based on cosmetic behaviors. It is a commented fact that cosmetic behaviors crumble under stress and a reversion to native behaviors takes place. The safety concern about dealing with cross-cultural interfaces through adaptation is obvious; in stress situations adaptation may become ineffective.

Behavior at the Cultural Interface Broadly speaking, there are 4 possible responses when members of one culture encounter members or artifacts of another culture (Berry, 1990). These are summarized in Table 2. The first two options are mirror images – either members of Culture A learn and adopt the ways of Culture B, or members of Culture B learn and adopt the ways of Culture A. Both options reflect a willingness or necessity to accept the ways of the other culture. Even within a single national culture, for example, in airline mergers, we see that the smaller airline usually ends up adopting the practices and policies of the larger airline, even where the smaller airline has a better way of doing things than the larger airline. A third option for interaction, Integration, allows members of Cultures A and B to learn something of each other’s ways, and find compromises that will be the most effective for that interface. In this approach, the two merging airlines might form pilot committees to review documentation and procedures and create a new (and improved) set of standard operating procedures suitable for pilots from both airlines. The final option, Separation, acknowledges that in many cases, members of Cultures A and B choose to ignore the ways of the other culture, and maintain their own cultural preferences. This model has been observed in some airline mergers, Cross-Cultural Factors in Aviation Safety 157 in so much as many years after a merger, some pilots will still identify with their former airline, drawing a distinction between themselves and the pilots of the other “dominant” airline, refusing to fully accept their connection with the other airline. Bearing in mind the moderating factors of cultural distance and resource match, and the motivation arising from the perceived benefits of adopting other cultural practices, the learning that takes places at the interface can be of varying quality. The understanding and adoption of practices from another culture can be:
Conscious or unconscious – Members may not even recognize they are adopt- ing a different way, yet they do it nonetheless. Unconscious adaptation can be problematic when the social norms in the two cultures vary because the behavior may look the same, yet be driven by different underlying logic. As the surprised tourist can attest, a gesture in one country may have a totally different meaning in another!
Complete or in part – Some practices may be easier (and more natural) to adopt than others. For example, the need for communication in the cockpit as advocated by CRM is appreciated globally, while the “best” form of command authority differs as a function of leadership preferences in different countries.
In good faith or otherwise – There may be a genuine effort to adapt, or in the face of confusion, there may be a superficial effort at adopting the other’s ways. The logic of some international regulations may be obscure to some developing countries, yet they feel compelled to adopt them as is.
Successful to varying degrees – Success is when members of one culture know how to act as if natives of the other culture. They understand the organizing principles of the culture sufficiently to predict the behavior of others in the new culture. In truth, it is very difficult for most people to achieve this level of cultural knowledge and awareness. This section introduced the concepts of culture and cultural interfaces. Table 1 listed some examples of the confusion that can arise at the intersection of two cultures and showed that cross-cultural interactions are a daily occurrence in aviation. The success of these interactions varies depending on familiarity and motivation. Because of the potential for confusion, misunderstanding, and misapplication, these cultural interfaces deserve closer scrutiny. The safety case for managing cultural interfaces is presented in the next section.

THE SAFETY CASE FOR CULTURAL INTERFACES IN AVIATION

This section draws on three models of human performance to argue the safety case for cultural interfaces in aviation. Hawkins’ SHEL model, adopted by ICAO 158 ASHLEIGH MERRITT AND DANIEL MAURINO as the conceptual model to explain Aviation Human Factors, is used to justify the basic safety case underlying cultural interfaces in aviation. Reason’s model of organizational safety is used to describe cultural interfaces as latent conditions in the system which if not managed can lead to active failures. Finally, the University of Texas Threat and Error Management (TEM) Model is proposed as an operational tool to deal with cultural interfaces as potential threats that must be managed in the same way as other environmental threats.

SHEL Model

The SHEL model2 (the name being derived from the initial letters of its four components) was the first model in aviation safety to directly address the risks latent at operational interfaces (Hawkins & Orlady, 1993). The interfaces were the interaction of Liveware (humans) with Hardware (machines), Software (procedures, symbology, etc.), other Liveware, and the Environment. The model placed the human at the centre of the four interfaces; a quick review of each of these interfaces will show the presence of culture. Liveware. Some of the factors affecting the performance of individuals include physical (the individual’s physical capabilities to perform the required tasks), physiological (those factors that affect the human’s internal physical processes), psychological (those factors affecting the psychological preparedness of the individual to meet all the circumstances that might occur during a flight), and psycho-social factors (external factors in the individual’s social system that influence them in their work environment). Though not directly mentioned in the original model, it is reasonable to include cultural factors with other physiological, psychological, and psycho-social factors influencing basic human performance issues such as workload, attention, memory, communication, and so forth. Liveware-Hardware (L-H). This interface looks at how the human interfaces with the physical work environment, e.g. displays to match the sensory and information processing characteristics of the user, of controls with proper movement, coding and location. This interface has received the most attention, as researchers and designers struggle to understand how humans will adapt to new technology, e.g. automated aircraft. Given that manufacturers are unlikely to build aircraft for different regions and populations, it is incumbent on them to build the most universally acceptable aircraft – by no means an easy task. The extent to which pilots have been exposed to similar technology previously will affect their ability to adapt. Nonetheless, technology designed with thoughtful consideration of cross-cultural issues involved in the transfer of technology can go a long way to neutralize cultural differences. Cross-Cultural Factors in Aviation Safety 159

Liveware-Software (L-S). The L-S interface is the relationship between the individual and all the supporting systems found in the workplace; e.g. the regulations, manuals, checklists, publications, standard operating procedures, computer software design, and so forth. It includes such “user friendly” cultural issues as format and presentation, vocabulary, clarity, symbology, etc. Unlike equipment, which cannot be so easily customized, software is an area with potential for manufacturers to work with their customers, to produce culturally calibrated documentation to ease the cultural interface. Liveware-Liveware (L-L). The L-L interface is the relationship between the individual and other persons in the workplace. In aviation, the advent of Crew Resource Management (CRM) (Wiener, Kanki & Helmreich, 1993), resulted in considerable focus on this interface, focusing on group performance activities such as leadership, crew co-operation, and teamwork. Adding culture to the mix expands the possibilities for confusion and misunderstanding as preferred modes of social behaviour interact across cultures in multicultural cockpits and international flights. As with early aviation Human Factors, this is the interface that so far has been the focus of cross-cultural endeavours. Liveware-Environment (L-E). This interface involves the relationship between the individual and both the internal and external environments. The internal workplace environment includes such physical considerations as temperature, ambient light, noise, vibration, air quality, etc. The external environment (for pilots) includes such things as visibility, turbulence, terrain, weather, etc. Further, the aviation system operates within a context of broad political and economic constraints, which in turn affect the overall corporate environment. Included here are such factors as the adequacy of physical facilities and supporting infrastructure, the local financial situation, regulatory effectiveness, etc. The contexts discussed in the previous section are clearly implied here. In sum, while culture was not on the Human Factors radar back in the 1970s when the SHEL model was first proposed, the idea of interfaces has been with us for 30 years. Extending the model to acknowledge cross-cultural realities is a logical step.

Reason’s Model of Latent Conditions

The SHEL model placed the human at the center of the aviation system to empha- size its most critical role. In so doing, it might inadvertently have generated a situ- ation that held the individual accountable for any failures in that system. Professor James Reason provided a broader perspective on aviation safety (Reason, 1997). Reason explained that seldom does an error by one individual operating in isolation precipitate an accident; typically, several causal and contributing factors 160 ASHLEIGH MERRITT AND DANIEL MAURINO converge in time and space to create a situation that is particularly vulnerable to one or more unexpected unsafe acts. Examples of such catastrophes include the accidents at the Three Mile Island (Pennsylvania, USA, 28 March 1979) and Chernobyl (Ukraine, USSR, 26 April 1986) nuclear power plants, the Challenger space shuttle (Florida, USA, 28 January 1986), and the double B–747 disaster at Tenerife (Canary Islands, Spain, 27 March 1977). Analysis of major accidents in technological systems indicated that the preconditions to disasters can be traced back to identifiable systemic deficiencies. It is typical to find that a number of undesirable events, all of which may have contributed to an accident, evolved through an “incubation period ”of perhaps years, until a triggering event, such as an abnormal operating condition, precipitates a disaster. Reason refers to these preconditions as latent unsafe conditions. Latent unsafe conditions have certain characteristics:
Latent conditions are to technological organizations what resident pathogens are to the human body. Like dormant bodily pathogens, which may lead to human disease, latent organizational conditions may be present for many years before they combine with local circumstances and active failures to penetrate the system’s many layers of defense.
Latent conditions are present in all systems. They are an inevitable part of organizational life where compromises between conflicting goals, e.g. production and safety, lead to shortfalls somewhere in the system.
Unlike active failures, which usually have relatively immediate and short-lived effects, latent conditions can lie dormant for a long time until they interact with local circumstances to defeat the system’s defenses.
Latent conditions may only become evident once the system’s defences have been breached. They may be present in the system well before an accident and are generally created by decision-makers, regulators and other people far removed in time and space from the accident.
Latent conditions include poor design, gaps in supervision, unworkable or ambiguous procedures, clumsy automation, shortfalls in training, and less than adequate tools and equipment.
Those at the human-machine interface, the operational personnel inherit any defects in the system, such as those created by poor equipment or task design; conflicting goals (e.g. on-time service vs. safety); defective organization (e.g. leading to poor internal communications); or bad management decisions (e.g. deferral of a maintenance item). Reason argues that Safety efforts should be directed at identifying and mitigating these latent unsafe conditions on a system-wide basis, rather than localized efforts to minimize unsafe acts by individuals, as unsafe acts are only the proverbial tip of the iceberg. Cross-Cultural Factors in Aviation Safety 161

Using Reason’s logic, cultural interfaces can be considered a latent condition, unless explicit attention is directed to them. They exist in the system, lying dormant, and operational personnel are expected to manage them. No one claims responsibility for them, yet they may one day foster unsafe acts leading to an active failure. They will foster these unsafe acts by virtue of the uncertainty and potential for misunderstanding they inject into the system. Otherwise competent professionals will “trip” at a cultural interface at a time when other defences are weak, perhaps when they’re under stress, and the results will be disastrous. As with other latent conditions, safety efforts should be directed at identifying and mitigating the potential adverse safety consequences of these cultural interfaces if left unchecked. The more these interfaces are actively understood and managed, rather than being allowed to continue unchecked as a pathogen in the system, the fewer unpleasant safety surprises they will generate.

Cultural Interfaces and the Threat and Error Model

The University of Texas Threat and Error Management (TEM) Model (Helmreich, Klinect & Wilhelm, 1999) provides a way of translating cultural interfaces from latent conditions into more visible and immediate concerns seen in an airline’s operating context. The TEM model, which evolved out of work with Line Operational Safety Audits (LOSA),3 posits that a typical line flight involves inevitable and mostly inconsequential threats and errors. Some errors are due to flaws in human performance (e.g. selecting wrong frequencies, acknowledging incorrect read-backs, mishandling switches) and others are fostered by external threats such as a late gate change, difficult terrain, a dispatch error, or equipment malfunction.

Threat An event or error that occurs outside the influence of the flight crew but requires their attention and management to maintain adequate safety margins.

Threats are different from latent conditions in that they are not necessarily weaknesses in the aviation system; rather they are conditions or events that increase the complexity of flight operations and therefore hold the potential to foster error. Flightcrew threats include environmental threats such as weather, ATC, terrain, and airport conditions, and airline operating threats such as dispatch, cabin, ramp, and maintenance problems. Successful flight depends on successful threat management. 162 ASHLEIGH MERRITT AND DANIEL MAURINO

Cultural interfaces fit the definition of threat in several ways:
Threats can be detected and successfully managed; they can also be ignored, leading to consequential or inconsequential outcomes. This is also the case with cultural interfaces.
Cultural interfaces are part of the operating environment, as are weather, terrain, airport conditions, and so forth.
Cultural interfaces, like other threats, require attention and management to maintain adequate safety margins.
Not every cultural interface produces a threat, but it does contain the potential of a threat. Viewed from this perspective, cultural interfaces are an inevitable component of the aviation environment, along with environmental threats and airline operating threats. Acknowledging the existence of cultural interfaces raises awareness of their threat potential – an important first step. To summarize the safety case for cultural interfaces: Hawkins introduced the notion of interfaces in the 1970s and noted their relevance for Aviation Human Factors. Reason brought the model of organizational accidents to aviation in the 1990s and broadened the Human Factors horizon to include organizational factors distant from yet influential upon the tip of the arrow, i.e. the flight deck, the ATC room, the maintenance hangar. Cultural Interfaces broaden the horizon even further, showing how members of one culture can incur confusion, misunderstanding, and misapplication when encountering members or artifacts of another culture. The Threat and Error Management Model provides a framework for “seeing” the cultural interface in the operating context. Bringing these interfaces to the forefront allows for systematic investigation. This investigation can determine which types of cultural interfaces are the most problematic; it can also study the threat management strategies employed by aviation personnel to manage these interfaces. Successful solutions can be shared with the industry.

The aim is not to eliminate culture or make us all the same. The goal is to recognize and manage the potential threats posed by different cultural interfaces.

A DOMINANT MODEL IN AVIATION: SOME CONSEQUENCES

Up to this point cross-cultural interactions have been conceptualized as the interaction of two relatively equal parties. The members of one group decide the Cross-Cultural Factors in Aviation Safety 163 extent to which they want to adopt or negotiate the practices of another culture. The reality of global aviation suggests that many interactions are not so balanced, that they are usually weighted in favor of a more dominant group or culture. This section explores the evolution of these weighted interfaces and some of the global consequences. Manufacturing, standards, research and technology, and global accident rates will be discussed.

Manufacturing

The first plane designed to carry fare-paying passengers was built in 1927 – the Ford Trimotor had 12 passenger seats. Boeing built the more comfortably advanced 247 in 1933, and Douglas Aircraft Company improved upon the concept with the 21-seat DC–3 in 1936, the first aircraft that enabled airlines to make money by carrying passengers rather than cargo. The major aviation innovations of World War II – radar and jet engines – occurred in Europe; production was the chief goal of the United States. There were fewer than 300 air transport aircraft in the United States in 1939; by the end of the World War II, U.S. manufacturers were producing 50,000 planes a year. The three major manufacturers were Douglas, Lockheed, and Boeing, and with 80% of the commercial airplane market, these companies established the United States as the leader of postwar civilian aircraft production. A new generation of jet airliners arrived in 1963 using the fuel-saving technology of the turbofan engine. Several models produced in Europe and the United States found popular support in the market: the Boeing 727 and 737, the French Caravelle, the BAC 111, the DC–9, the Fokker F–27 Friendship, and the de Havilland Twin Otter. Boeing continued its hold on the market with the 747. Airbus was established in 1970 as a European consortium of French, German and later, Spanish and U.K. companies. Airbus’ first aircraft, the A300B, was launched at the 1969 Paris air show. Ten years later, the consortium had expanded its family of aircraft to 16, and now Airbus shares the market for new aircraft equally with Boeing. (As Boeing has been established for longer, its aircraft outnumber those of Airbus by a factor of six to one.) The conclusion is obvious: Despite the former USSR’s formidable aviation industry, and enterprises in countries including Australia, Canada, Brazil, and Indonesia, North America and Western Europe are the major suppliers of aircraft equipment for civil aviation.

Markets and Standards

The U.S. government played a pioneering role in the liberalization and deregula- tion of the global airline industry during the post World War II era. At the Chicago 164 ASHLEIGH MERRITT AND DANIEL MAURINO

Convention of 1944, the United States pushed for “unrestricted international operating rights” in foreign airspace but was met with intense resistance from European nations fearing that an open skies regime would lead to domination by U.S. airlines. However, since America was developing as the world’s largest and most lucrative air transportation market, European governments were essentially forced into deregulating their airline industry to gain access to American skies. Rapid liberalization of the airline industry in the early 1980s resulted in a significant increase in the demand for air transportation between countries, causing intense competition among national carriers for profitable long haul flights. With airlines from one country now having access to another country’s airspace, national airlines formed multinational airline alliances to strengthen their international position and prevent further encroachment by foreign competitors. With almost 70% of the world’s aviation departures occurring in either the United States or Western Europe, these regions are in a strong position to influence standards for airline operations. These standards were first formalized by ICAO through the Chicago Convention of 1944 and its eighteen supporting annexes. In spite of the multi-national nature of ICAO and its harmonized rule-making process, the United States and Western Europe have wielded considerable influence in the development of the rules and regulations of international aviation standards throughout the latter half of the 20th century. This occurred not only through ICAO but also through other international organizations such as the International Air Transport Association (IATA). However, the most important influence has been wielded through the goods and services (training, SOP, safety practices, operational standards and so forth) prescribed by American and European aircraft manufacturers such as Boeing, McDonnell, Douglas, Lockheed, Airbus, and British Aerospace. The conclusion is obvious: North America and Western Europe are the largest markets for aviation. Professional and special interest groups within these countries have influenced global aviation standards.

Case Study USA: The Largest Aviation Customer and Supplier Based on passenger figures from the year 2000, 42% of all air travel originates in the USA, 26% originates in Europe, and 23% originates in the Asia-Pacific region; the remaining 11% occurring in diverse locations. The aviation industry in the United States contributes 9% of the Gross Domestic Product, making it the largest industry in the most industrialized nation in the world. Cross-Cultural Factors in Aviation Safety 165

To put this number in perspective, the money generated by civil aviation in the United States is larger than the entire Gross Domestic Product of Canada, a country with the 8th largest GDP in the world.4 How did this situation come to exist? How did the United States become the largest customer and the largest supplier in global civil aviation? There are several factors that together explain this outcome. Their origins can be traced to three of the contexts discussed in Section I. Socio-economic and political context Population: The population of the United States exceeds 250 million; a large population makes an expensive mass-transit system viable. Historic: The U.S. emerged from World War II as the primary supplier of airplanes. Within the US, manufacturers competed by building ever more successful planes in terms of size, fuel economy, and passenger comfort. Political: The U.S. has enjoyed a stable government system throughout the twentieth century. This has allowed government structure to grow, stabilize, and become predictable in terms of controls and support. Tax Base: A corollary of this political stability has been a relatively high and stable tax system that provides consistent government support for the aviation infrastructure. The tax base also funds aviation research. Economic System: Capitalism and the belief in a free-market economy led to deregulation of the aviation industry in 1978. Competition has created survival of the strongest, with little sympathy or assistance for those airlines or manufacturers who cannot compete. Education: 90% of the population have gone to high school; more than 75% enroll in some form of tertiary education. An educated population provides the demand and the human resources for research and development of new technology and products. Economic Resources: The Gross Domestic Product of the USA is the largest in the world, twice that of the next largest country Japan. High per capita income makes air travel affordable. Geographic and physical context Geography (the USA): The U.S. is a vast country with the majority of the population living on the East and West coast, making air travel more of a necessity than a luxury. Air travel is the primary means of public transportation between cities in the U.S. today. Geography (the world): The U.S. is separated from most of the world by large oceans. Air travel is the only fast way of reaching these countries. 166 ASHLEIGH MERRITT AND DANIEL MAURINO

Social context Ideas & Authority: The U.S. is characterized as having low power distance relationships relative to many other countries (Hofstede, 1980, 2001). This attribute, combined with a highly educated population and a market-driven economy results in ideas, especially practical (money-making) ideas, taking precedence over status. Anyone can have a good idea, and ideas propel change. Pace of Life: People in the United States have a more urgent sense of time than people in many other countries. This urgency is well serviced by faster air travel and convenient schedules. Risk: In a market-driven democracy, the flying public decides the level of acceptable safety. The absence of civil strife, plus ever-improving health statistics such as longevity rates drives the perception of acceptable risk ever downward.

While a similar case might be made for Western Europe as a whole, no other single country has quite the same confluence of history, geography, economy, population, politics, and national values; it is not surprising then that the United States has emerged as the largest customer and the largest supplier of commer- cial aviation in the world.

Research and Technology

The aviation industry owes its existence to modern technology. Most of this technology has been developed by researchers working in government and industry-funded institutions in politically stable, resource-rich countries such as North America and Western Europe. Like anyone else, these researchers think and work in an environment that is influenced by larger economic, political, cultural, and physical concerns. For example, the popularity of air travel in North America and Western Europe has shifted the notion of acceptable risk. As more and more people use airlines as a common and cheap mode of transport, so too the public now expects air travel to be as safe as riding a bus. This demand for safer and safer air transport drives the research for safety-enhancements. At the same time, the economic context determines what is possible financially, and in what form. Research is typically undertaken to solve local issues with tools supported by local conditions. In North American and Western Europe, technological solutions are usually expensive, but they are validated and compensated by strong customer demand. These solutions are a good fit for the problems generated Cross-Cultural Factors in Aviation Safety 167 in high traffic-density situations supported by a strong and stable infrastructure with modern fleets and enforced maintenance and training standards. These characteristics are a reliable part of the operating context in these countries and are part of the researchers’ starting assumptions when they start to shape solutions. But these solutions are then transferred to other parts of the world where the infrastructure is not equally strong, fleets are older, and training and maintenance are as good as they could be given the local constraints. Different contexts create different priorities and problems. They require different solutions. Assumptions about what problems are important and what solutions are viable drive all research efforts. Technology is a cultural artifact because it is driven by:
What is perceived as a problem in the current environment
The suitability of tools that are brought to bear on a problem, and
The goodness-of-fit between solutions and the most likely users. Technology is the product of people solving problems that they see; hence, it is less of a problem when used within the culture of its origin. In such circumstances, the technology may be the perfect alignment of problem, tools, and user-customers. But when members of one culture encounter technology created in another culture, the cultural interface may introduce potential mis-matches. The mis-match may occur in the definition of the problem, the suitability of the tool, or the cultural norms and preferences of the user-customers. Conclusion: Researchers based in North America and Western Europe have developed technology that has led to significant global improvements in aviation safety and efficiency. However, it’s a matter of record that these improvements have not been equal around the world.

Global Consequences of Weighted Interfaces

The preceding sections suggest that North America and Western Europe have exerted strong influences on aviation through manufacturing, the size of their markets, and the technological advancements they have funded and developed. As a result, when we talk of cultural interfaces in aviation, most of the time we are talking about aviation personnel encountering aircraft, equipment, training, manuals, and standards that originated in the economic, political and cultural context of North America and Western Europe. Some examples of these weighted interfaces are listed in Table 3. Further evidence of these weighted interfaces may be found in regional accident rates. In the USA, Australia and Western Europe, the civil aviation accident rate is less than one accident per million departures. In other parts of the world, the rate 168 ASHLEIGH MERRITT AND DANIEL MAURINO

Table 3. Some Unintended Consequences of the Dominant Model in Aviation.

Issue Example

Identification of problems Researchers in North America and Western Europe are developing solutions to reduce the accident rate from one per million to one per 10 million departures. In some parts of the world, the rate is closer to one accident per 100,000 departures. These regions have different, possibly more essential problems that need to be identified. Solutions The high-technology response to traffic problems is to install radar, GPWS and other expensive solutions. In certain parts of the developing world, this is not viable due to a smaller customer base and reduced national wealth. When flight support services are unreliable in this context, the low-technology solution is for pilots to rely on other pilots in the airspace for updated information. Another example: An international safety group developed an ALAR Tool Kit based on extensive research. They translated the product into several languages and copied it on to one efficient compact disc for global distribution. When distribution of the product started, it was realized that pilots in many developing countries do not have compact disc readers. Language The Lingua Franca of aviation is English. This standard favors pilots in some areas of the world more than others. Nonetheless, all are expected to speak English as the normal way of doing business in international aviation. A software company experimenting with English vs. native language computer interfaces discovered that pilots committed more ATC readback errors when the interface was English and not their native language. Despite the higher error rate, the pilots still said they preferred the English computer interface, as it was important to them to “be like other pilots.” Training Airlines in some countries buy training “off the shelf” from North American and Western European suppliers, often with little or no customization for local conditions. This has been particularly problematic with CRM training, which is a very culturally influenced tool because of its focus on communication and command authority. Technology transfer In an industrially developing society, new technology must be dealt with in the absence of prior experience of its earlier stages because there is no tradition in the use of earlier, related technologies. There can be a significant gap from existing knowledge and skills to those required for successful operation and maintenance of the to-be-implemented new technology. As a general example, it is easier for a person who has worked with Microsoft Office 95, 98 and 2000 to work with the latest upgrade than it is for a person who has never before worked with Microsoft software. Cross-Cultural Factors in Aviation Safety 169

Table 3. (Continued )

Issue Example

There is also the problem that the technology implicit in the design of the new technology can clash with the culture of the user. Regulator A region imported another country’s civil aviation regulations as the best (and cheapest) way to comply with the international standards. They even used automatic translation. It was not successful because there was no adaptation to the local context. For example, the inspectors are traditionally paid considerably less than airline personnel in this region and it is possible to circumvent the inspector because of his low status. The respect accorded to inspectors to do their job in this region is considerably lower than in the culture in which the civil aviation regulations originated. varies from 2 to 10 accidents per million departures. There is a negative correlation between accident rates and the extent to which regions posses the same economic, political, and social contexts as North America and Western Europe. The richer the country, the more politically stable, and the more economically viable that air travel is in that region, then the lower is its accident rate. In other words, the closer the fit to the dominant model as influenced by North America and Western Europe, the less consequential the outcomes at the interface appear to be.

Interactions at Weighted Cultural Interfaces

Section I described four possible ways to behave at a cultural interface. They were: Assimilation whereby members of Culture A learn and adopt Culture B’s ways, or conversely, members of Culture B learn and adopt Culture A’s ways; Integration whereby members of Cultures A and B learn each other’s ways & compromise; or Separation whereby members of Cultures A and B ignore each other’s ways & do not change. In light of the discussion re the dominant influences on many of the interfaces in aviation, the model can be updated as shown in Table 4. Option 1, Assimilation, remains the same as in the previous model. The more a country shares similar political, economic and social factors with the dominant model as shaped by North America and Western Europe, the more chance it has of fully assimilating into the dominant model and succeeding. This option has much popular support in the guise of “One Size Fits All” approved practices and regulations. This is an attractive option because the majority of research and technological advancements have been pored into the dominant model. Hence, for those who can assimilate, the safety and productivity benefits are substantial. 170 ASHLEIGH MERRITT AND DANIEL MAURINO

Table 4. Four Options at a Weighted Cultural Interface.

Option 1 Option 3 ASSIMILATION INTEGRATION Culture A learns and adopts dominant Culture A interacts with proponents of Dominant Model model to understand and modify approaches One size fits All Local solutions Option 2 Option 4 COSMETIC COMPLIANCE MARGINALIZATION Culture A gives the appearance of Culture A unable to adopt dominant model; no option adopting dominant model Face Value Isolation

Option 2, Cosmetic Compliance, acknowledges the unlikelihood that members of the dominant model will assimilate into other cultures. Instead, this option addresses the difficulty of full assimilation. Cosmetic compliance applies to those cases where members of Culture A want to emulate the dominant model and its practices, but full assimilation is not possible, usually because they do not share the same contextual traits as mentioned previously (infrastructure may be weaker, fleets older, customer base smaller, technicians and civil servants may come from different educational backgrounds). The result is likely to be superficial adherence based on insufficient resources, incomplete understanding, and/or mis-matched tools. This option invariably fails under high-stress situations as the more fundamental logic of one’s primary culture takes hold and directs behavior. The first two options show the strength and influence of the dominant aviation model as shaped by North America and Western Europe. The majority of airlines around the world have adopted the dominant model, or tried to, hence the real and apparent similarity observed across airlines. These options support the absolute notion that “An Airline is an Airline; A pilot is a pilot” regardless of background and that therefore there is no need to address cross-cultural factors in aviation. As stated previously, the advantages of adopting the dominant model are numerous; nonetheless, there are serious safety concerns inherent in cosmetic compliance. Skipping ahead to Option 4, Marginalization is the result of members of Culture A being unable to assimilate into the dominant model, yet having no substantive alternative. The differences in context and resources as so profound as to make Assimilation virtually impossible. At the same time a viable alternative is not known. Advocates of the dominant model may offer help, but the help is nearly always based on the contextual assumptions of the dominant model, hence the members of Culture A are back where they started, unable to use the “help.” The result is regionally isolated groups with poor performance and high accident rates as currently seen in some parts of the world. Cross-Cultural Factors in Aviation Safety 171

The last option to be discussed, Integration, offers the greatest hope for successful management of cultural interfaces and the potential threats they pose. The concern for global safety and the recognition of threats in common airspace is sufficient reason to consider the Integrationist perspective. It requires members of Culture A to be able to identify and articulate the problems they see in trying to adopt the dominant model to their local conditions, and it requires members of the dominant model to consider culturally calibrated modifications to their model. Integration is the only way to develop accurate and efficient solutions to local problems in areas that do not share the same economic and cultural features inherent in the dominant model. ICAO and other international organizations have played significant mediating roles in this respect.

Summary

The United States emerged from World War II with the economic resources and mass-production capabilities to dominate aircraft manufacturing. Western Europe followed shortly thereafter, and technological innovations in the following decades further strengthened their position as leaders in civil aviation. With large customer bases and secure infrastructures, these countries were able to research and develop safety enhancements that were quickly absorbed into an evolving dominant model. Decades of significant financial investment have sustained and developed the dominant aviation model to such a level that a serious competitive alternative seems unlikely. Global aviation is a constant cross-cultural endeavor. However, the emergence of a strongly dominant model has meant that most cultural interfaces, certainly those involving technology, are now weighted in favor of the dominant model and its practices. In light of this dominance, the options that promote the greatest global safety are full Assimilation into the dominant model and Integration of the dominant model with local conditions to produce efficient local solutions. Full Assimilation is possible only when the contextual factors in the receiving model are similar to those found in the dominant model. In reality, some form of cosmetic compliance is often the case. Integration offers the most hope for crafting local solutions for local problems; however, it also requires the most effort, as explained in the next section.

THE WAY FORWARD

This paper has presented the safety case for cross-cultural factors in aviation by focusing not on cultures so much as on cultural interfaces, those situations where 172 ASHLEIGH MERRITT AND DANIEL MAURINO members of one culture encounter people or artifacts from other cultures. In a global industry such as aviation, interactions at these interfaces are an every-day occurrence.

Cross-cultural issues in aviation can only be resolved with joint effort. This is not something that “they” (the other cultures) have to fix – there is a role for people on both sides of the interface, for members of the dominant model as well as people outside the dominant model.

At this time, there is almost no systematic information available about cultural interfaces. We don’t know which interfaces are problematic; nor do we know the extent of successful assimilation or integration in different regions of the world. In order to move forward, two broad and interdependent strategies are proposed. The first strategy is to raise awareness of cultural interfaces and their threat-potential amongst various aviation personnel via training and analysis. This paper hopes to facilitate this awareness. The second strategy is to collect systematic data in the operating context of different regions around the world to quantify the threats posed by different cultural interfaces and to understand local adaptations to the dominant model. The specific details of each strategy are described below.

Raising Awareness: It Takes Two to Tango

Training to “See” Culture One of the interesting things about culture is that everyone is a member of at least one culture, yet people can rarely explain their culture very clearly to others. That is because like fish unaware of the water in which they swim, people in their own culture tend to be unaware of the principles and values that guide their behavior. Historically derived, the culture is handed down from generation to generation as folk wisdom. The trainers teach the cadets, the Captains shape the First Officers, the “old guys” in the hangar teach the new guys “how things are really done around here” and so the cultural beliefs are passed on to the new members. It is natural for people to believe that the culture in which they were socialized is the “way of the world” and that practices and beliefs are similar the world over, i.e. “the same as mine” (Ward, Bochner & Furnham, 2001). However, in today’s world of global media and global travel, this ethnocentrism cannot be logically sustained. To believe that one culture is inherently superior to another is arrogance or ignorance or both (Stewart & Bennett, 1991). Helping people move beyond their own cultural boundaries to accept other cultures without becoming defensive is something that can be addressed through Cross-Cultural Factors in Aviation Safety 173 training (Bennett, 1998). Being able to accept other cultures requires a person to look objectively and with some distance at their own culture. Once a person can see the strengths and shortfalls in their own culture, they can start to see the pluses and minuses in other cultures. Next, the person can then start to see how something that works in their culture may not work in another culture, for good reason. The final step is to be able to see how an idea or tool from one culture can be modified to make sense in another culture. This is the end goal – to see how culturally produced artifacts and procedures can be modified and adapted for other cultures. The personnel who could benefit from such training include:
Those who study human performance
Those who design aviation tools and equipment
Those who write operating manuals
Those who train personnel from other regions
Those who interact directly with personnel from other regions (multi-cultural cockpits, pilots and flight attendants working international flights) Training should include a mix of culture-general theories and culture-specific practices (Landis & Bhagat, 1996). There is research that shows that people from different cultures differ in their preferred way of communicating, interacting, learning, thinking, reasoning, and problem-solving (Helmreich & Merritt, 1998; Hofstede, 2001; Nisbett, 2003; Samovar & Porter, 1998). These fundamental human processes underlie all design and training issues. As mentioned earlier, this type of training is not just for those who want to assimilate into the dominant model. By way of example, English is the lingua franca of aviation. Does this mean that only non-native speakers of English must study and perfect their standard English usage? The answer is most definitely “No.” The dominant model favors certain nationalities because of the English requirement. The challenge for these English-speakers is to remove all jargon, jokes, excess words, complex sentence constructions, and non-standard phraseology from their professional communication in order to communicate standard English to those from other backgrounds. The majority of Anglos are mono-linguists who have never faced the challenge of communicating in a second or third language. Anglo pilots and Anglo ATC need to appreciate the necessity for slow, short, and clearly articulated words, especially if stress or comprehension is an issue.

Analysis to “See” Culture Whether one advocates Assimilation (“One size fits all”) or Integration (Informed Local Solutions) as the best model of global air safety, the first step is the same – to analyze and explain the underlying logic and intent of the dominant model and its tools. Researchers, designers, manufacturers, and regulators need to recognize 174 ASHLEIGH MERRITT AND DANIEL MAURINO the assumptions upon which their work is based, and make the logic available for scrutiny by others. The contextual framework introduced in Section I can be used to explore the factors that shaped the thinking of those who shaped the dominant model. Even if advocates of the dominant model ethnocentrically believe that Full Assimilation is the best global solution, they are still obliged to clarify the model, its tools, and its assumptions so as to allow clear and unobstructed adherence. An example of such an analysis can be seen in the Case Study on LOSA below. Notice how the case study focuses on the cultural assumptions and beliefs that underlie LOSA; it does not address “other” cultures, it simply explains the designers’ logic. It is the obligation of aviation personnel outside the dominant model to assess the extent to which conformity with the dominant model is plausible. In order to do that, a contextual analysis similar to the one just mentioned but applied to local conditions would be necessary. These two analyses could form the basis of an In- tegrationist dialog aimed at identifying mis-matches in the operating environment regarding problem definition, tools, and/or customer and user-preferences.

Case Study Cultural Assumptions Underlying Line Operations Safety Audits (LOSA) LOSA is an aviation safety tool originally developed by the University of Texas Human Factors Research Project (see LOSA Manual, ICAO Document 9803). LOSA is endorsed by the International Federation of Airline Pilots’ Associa- tions (IFALPA), ICAO, and IATA, and has been used by airlines in the United States, Asia, Australia-Pacific, and Europe. As LOSA becomes better known, more airlines are asking what is it, and should we be using it? The following analysis should help airlines decide if LOSA is right for them. The analysis ad- dresses three areas – how the problem was defined for which LOSA was seen as the answer, the tools that were brought to bear on the problem, and the cultural norms and preferences of the user-customers. The analysis is not presented as “proof” that LOSA should be adopted universally; on the contrary, its logic is laid bare to allow interested parties to determine if a LOSA or some modifica- tion would be a useful tool in their local environment. (Note: Many of the factors mentioned earlier regarding U.S. norms and preferences are also relevant here.) The problem in context Every airline asks itself, “Are we safe enough?” In many airlines, the answer was originally determined by looking at accident and incident data. Researchers studied the data, introduced new technology and other solutions to address the hypothesized deficit, assessed its effectiveness, and modified where necessary. This led to reduced accidents (and reduced accident data to analyze). Still, Cross-Cultural Factors in Aviation Safety 175

the public wanted less risk/more safety – a single accident could mean the demise of an airline – and seeking a competitive edge in a market-driven industry, airlines looked for more sophisticated and more proactive means of improving their safety records. The problem was how to stay ahead of the public’s perception of acceptable risk, reduce costs associated with adverse events, and still make a profit. Tools to address the problem LOSA is the joint product of a research group located in the Psychology Department of an American University and several major American airlines. Cultural biases are therefore unavoidable. One such culturally fostered belief among the University of Texas researchers (an Anglo-Saxon group), like most American and Western European scientists, is that “Truth” is established via the scientific method – a method that proposes you can take a sub-group of the population, and by systematically observing that sub-group, you can draw inferences about the whole group. Statistical analysis of the observations allows probability-based inferences to be made. There is a belief that the scientific method has a purity and lack of bias that makes it superior to individual opinion, regardless the individual. Rigorous scientific technique ensures unbiased outcomes, i.e. greater truth. The scientific method is one way to acquire knowledge, but it carries many underlying assumptions about em- piricism and objectivity; other forms of inquiry acknowledge the inevitability of subjective “truth” and include ethnographic field studies and interviews. There is also a strong belief, supported by cultural practices that data can be confidential (in other words, not shared with inappropriate others) and anony- mous (the respondent’s identity is protected), and that these characteristics will further the unbiased nature of the outcome. It is very common to conduct surveys in the USA and Western Europe based on the promise of confidentiality and anonymity, so much so that respondents accept it as standard and proper practice. This faith in protection against exposure is not necessarily the case everywhere. As psychologists, the researchers focus on human rather than machine performance, and look at antecedents and consequences of that performance. Being from an American university, the research group obviously subscribes to many American values. There is very low power distance within the group, and all ideas are welcomed from the most junior student upwards. There is a sense of time urgency and a need for constant innovation and change that propels the group to constantly evolve its research ideas and products. Pragmatism – always asking if something will work or if it will be useful – is a guiding 176 ASHLEIGH MERRITT AND DANIEL MAURINO

principle that motivates the researchers to constantly test their ideas in the field with end-users An idea only has value if it can be used. The urge for activity (“doing” more than being or thinking) also motivates constant evolution and testing of tools in the field. Ongoing funding by the Federal Aviation Administration has enabled this research evolution. The money is generated through taxes and allocated annually. That is, there is an assumption that aviation research is valuable and should be funded by the general population. This is part of a greater belief in science and technology as valid paths to a more valued future. As part of the greater community of aviation researchers located in resource- rich, technologically enabled countries, the group has had ongoing access to new theories and ideas. The concept of error as an inevitable and healthy aspect of hu- man behavior – first expounded in WesternEurope and later exported to America – has greatly influenced the group’s work. Several years of discussions, confer- ences, and published papers (in English) have enabled the researchers to advance their ideas in a community of like-minded scientists. The Threat and Error Man- agement model discussed in Section II evolved out of this intellectual context. In sum, the tools that were brought to bear on the problem included empiricism and its belief in “Truth through Data,” a focus on human performance, years of research backed by stable funding and enhanced through interaction with like-minded researchers, and a time-urgent pragmatism mixed with a belief in constant improvement. Norms and preferences of the user-customers LOSA was developed with the help of personnel working in large Anglo-Saxon airlines. As a result, many of the features of the LOSA “solution” reflect their norms and preferences. LOSA is defined by ten operating characteristics: (1) Jumpseat observations during normal flights (2) Joint management/pilot association sponsorship (3) Voluntary crew participation (4) De-identified, confidential, and non-disciplinary data collection (5) Targeted observation instrument (6) Trained and calibrated observers (7) Trusted data collection site (8) Data cleaning roundtables (9) Targets for enhancement (10) Feedback results to the line pilots An examination of these characteristics reveals several cultural assumptions and beliefs. First, there is a strong reliance on empiricism and the scientific Cross-Cultural Factors in Aviation Safety 177

method (#4, 5, 6, 7, and 8 above); there is also a strong egalitarianism and sense of low power distance that places pilots, unions and management on equal footing for the purposes of #2, 3, and 10. Point # 9 presupposes that the reason for conducting LOSA is a desire to improve (not simply to get a “check”). Fi- nally Point #1, on which all of LOSA is predicated, presupposes that pilots will allow confidential, anonymous, and objective observation of their performance. To establish this credibility with the pilots, the proposed solution requires that pilots observe pilots. They may be retired pilots, first officers, training captains, or line pilots; nonetheless the airline must be prepared to pay these people to observe others working. As such, LOSA can be an expensive tool. LOSA was developed in response to a problem defined by airlines operating within the dominant model, and it was shaped to suit the preferences of the users, namely airlines operating within the dominant model. Other airlines will need to decide:
If the problem as defined is relevant to them and shares the same sense of priority as other safety concerns in their local environment. Resources might be more effectively applied elsewhere in the system to improve safety, there may not be the same level of public outcry for safety, and there may not be a market-driven, competitive need for improvement.
Whether the empirical data-driven tools brought to bear on the problem would be logical and acceptable to pilots and management in their local context, and
If they would need to modify the tool in any way to make it more acceptable to their pilots or management. The University of Texas researchers endorse the Integrationist approach of local solutions for local problems. As such, they invite airlines to enter into a dialog with them about how LOSA might be modified to meet their different needs.

In sum, raising awareness of cultural interfaces via training and analysis is an essen- tial step toward recognizing and managing them. Those who shape the dominant model as well as those who try to adapt it need this awareness.

Data-Driven Research

Another way to focus attention on cultural interfaces is through data-driven research. Focusing on cultural interfaces rather than cultures allows culture to be taken out of the head and put into the operating context where it can be studied more dispassionately and with greater operational focus. Cultural interfaces deserve a place on the aviation safety research agenda. 178 ASHLEIGH MERRITT AND DANIEL MAURINO

TEM – A Methodology for Investigating Cultural Interfaces The University of Texas Threat and Error Management model (TEM) (Helmreich, Klinect & Wilhelm, 1999) divides the operating context into environmental and airline operating threats, error, undesired aircraft states, and their respective management patterns. As discussed previously, cultural interfaces fit the definition of a threat and can be incorporated into the model. Jumpseat observations during normal operations (the LOSA methodology) will proactively identify problematic interfaces and their management. It will also identify those interfaces that are not problematic, and most importantly, it will identify successful local adaptation and integration strategies to manage problematic interfaces. Crew interviews will further the understanding of these interfaces. For example, it may become apparent from flying in certain parts of the world that certain procedures are unworkable, that communication with ATC is problematic due to accents or equipment, that constraints make some maintenance activities unviable at some stations, that navigation aids are unreliable, that weather notifications are based on outdated or incomplete data, that passengers are likely to carry hazardous materials, and so forth. Culture and cultural interfaces, while interesting in and of themselves from an anthropological perspective, are only interesting here in how they impact safety. Thus, tying cultural interfaces to their management in the operating context is the bottom line. Manufacturers, legislative bodies, individual airlines, airline alliances, and cross-cultural teams of Human Factors researchers could conduct the research. “Seeing” some of the interfaces may be initially challenging due to reasons mentioned above; hence, the presence of one or more cultural mediators (see below) on the research team would be a great advantage.

Cultural Mediators: An Indispensable Component

Some people are more suited to understanding cultural interfaces than others. Of particular importance are those people who have lived and worked effectively in two or more cultures and are able to explain the practices and logic of one culture to members of another culture in a way that makes sense to both sides (Bochner, 1981). These cultural mediators act as bridges between cultures. Such people exist in aviation coincidentally, but more are needed. To that end, the design of programs to specifically develop more cultural mediators is proposed as a solution. These programs, in the form of internships and exchange programs, could occur within airlines, airline alliances, research institutions, civil aviation authorities, accident investigation services, and manufacturers. There are existing programs similar to those advocated here, but they are initiatives of specific organizations rather than a Cross-Cultural Factors in Aviation Safety 179 concentrated, industry-wide endeavor. Making cultural interfaces an explicit (not implicit), often-discussed topic would be part of these programs. Some interfaces need to be challenged, and that will only come with increased confidence and understanding on both sides. Hence, any programs, exchanges, or internships that would accelerate the number of cultural mediators in aviation would be very welcome.

SUMMARY AND CONCLUSION

Cultural interfaces will always be a part of global aviation. The challenge for safety is not to eliminate these interfaces but to manage the potential threats they pose. Systematic research of the interfaces in different regions around the world will uncover strengths and weaknesses in the global system. These successes and failures will focus future efforts to improve the interface. To move forward, there is a role for those inside and outside the dominant model.

The Role of Advocates and Insiders of the Dominant Model

Continue to advance safety in the industry utilizing all resource-rich advantages
Recognize that inevitable cultural boundaries surround many solutions and ad- vances
Recognize full assimilation into the dominant model is not always possible due to contextual differences
Accept integration and adaptation as valid safety strategies and open the door for integration (ideas from outside the dominant culture) as a way to build a stronger global industry
Support systematic research of cultural interfaces in different operating contexts
Establish programs that develop cultural mediators (internships, exchange)
Enlist cultural mediators in the development of tools, policy, and procedures
Encourage manufacturers to work at the level of Standard Operating Procedures with their customers.
Provide awareness training for those who operate at cultural interfaces

The Role of Cultural Groups Outside the Dominant Model

Recognize contextual differences with the dominant model
Acknowledge constraints in complying with external standards 180 ASHLEIGH MERRITT AND DANIEL MAURINO

Articulate local problems and mis-matches with the dominant model
Recognize cultural limits of some solutions advanced by the dominant model
Encourage systematic research of cultural interfaces in the operating context
Encourage personnel to become cultural mediators
Develop the knowledge and confidence to challenge the dominant model
Develop the knowledge and confidence to advance local solutions
Provide awareness training for those who operate at cultural interfaces As the first and greatest of all global industries, it is time for aviation to acknowledge cross-cultural factors in aviation safety. With combined effort, one of the last frontiers in aviation safety can be confronted, investigated, and managed.

NOTES

1. Anthropologists, psychologists, political scientists, and sociologists all differ in their definitions of culture. However, culture is bigger than any one discipline, and much can be learned from accepting all definitions simultaneously. 2. The SHEL concept was first developed by Professor Elwyn Edwards in 1972, with a modified diagram to illustrate the model developed by Frank Hawkins in 1975. 3. For a full discussion of LOSA, see ICAO Doc. 9803. 4. GDP statistics issued by the International Monetary Fund.

REFERENCES

Bennett, M. J. (1998). Basic concepts of intercultural communication. Intercultural Press. Berry, J. W. (1990). Psychology of acculturation. In: J. Berman (Ed.), Cross-Cultural Perspectives (Vol. 37, pp. 201–234). Nebraska Symposia on Motivation 1989. Bochner, S. (1981). The mediating person: Bridges between cultures. Boston, MA: G. K. Hall. Hawkins, F., & Orlady, H. (1993). Human factors in flight (2nd ed.). Aldershot: Ashgate. Helmreich, R. L., Klinect, J. R., & Wilhelm, J. A. (1999). Models of event, error, and response in flight operations. In: R. S. Jensen (Ed.), Proceedings of the Tenth International Symposium on Aviation Psychology (pp. 124–129). Columbus: Ohio State University. Helmreich, R., & Merritt, A. (1998). Culture at work in aviation and medicine: National, organizational and professional influences. Brookfield, VT: Ashgate. Hofstede, G. (2001). Culture’s consequences (2nd ed.). Beverly Hills: Sage. International Civil Aviation Organization (2000). ICAO safety oversight audit manual, Doc 9735. Montreal, Canada: Author. Landis, D., & Bhagat, R. (1996). Handbook of intercultural training. Thousand Oaks, CA: Sage. Nisbett, R. (2003). The geography of thought: How Asians and Westerners think differently ... and why. New York: Free Press. Reason, J. (1997). Managing the risks of organizational accidents. Aldershot: Ashgate. Cross-Cultural Factors in Aviation Safety 181

Samovar, L., & Porter, R. (1998). Intercultural communication (5th ed.). Belmont, CA: Wadsworth. Stewart, E., & Bennett, M. (1991). American cultural patterns: A cross-cultural perspective. Intercul- tural Press. Ward, C., Bochner, S., & Furnham, A. (2001). The psychology of culture shock. London: Routledge. Wiener, E. L., Kanki, B. G., & Helmreich, R. L. (Eds) (1993). Cockpit resource management. San Diego, CA: Academic Press. This Page Intentionally Left Blank 6. THE ROLE OF NATIONAL CULTURE IN ENHANCING TRAINING EFFECTIVENESS: A FRAMEWORK

Shreya Sarkar-Barney

ABSTRACT

The focus of global training has primarily been on preparing employees to work effectively in other cultures, such as in expatriate training, accultura- tion training, and training for technology transfer. One issue that has been ignored is the implication of using training systems that are developed in a specific cultural context and then deployed globally. This chapter proposes a framework to show the influence of culture on one aspect of training effectiveness, the transfer of newly learned skills to the job. Specific relation- ships are proposed, using Baldwin and Ford’s (1988) transfer of training framework as a guide, and also by synthesizing findings from areas such as cross-cultural psychology, human resource management, education, and technology management.

INTRODUCTION

Training in organizations is a strategic human resource function (Dipboye, 1997). It is used in small and large corporations to develop appropriate knowledge, skills, and abilities in employees to enable them to perform important tasks and ultimately deliver organizationally relevant outcomes (McGehee & Thayer, 1961). The success of any training function depends on the relevance of the content to job

Cultural Ergonomics Advances in Human Performance and Cognitive Engineering Research, Volume 4, 183–213 © 2004 Published by Elsevier Ltd. ISSN: 1479-3601/doi:10.1016/S1479-3601(03)04006-2 183 184 SHREYA SARKAR-BARNEY requirements and timeliness of the training delivery (Brinkerhoff & Gill, 1994). To meet these goals, training systems should be adaptable to changing internal organizational and external marketplace conditions. In the past decade, one such change has been increase in transnational business. Market expansions to global locations and relocation of manufacturing centers to low-cost labor markets has created a need for considering the influence of culture on human performance and developing culture-specific organizational practices. This is evidenced by the increased focus on understanding global training practice by researchers and practitioners (see Black & Mendenhall, 1990; Kemper, 1998; Rhinesmith, 1996; Weech, 2001). However, this focus has been one-sided, and the interest has mainly been on training employees to work efficiently in other cultures, such as in expatriate training, acculturation training, and training for technology transfer (see Rhinesmith, 1996). The growth of multinational corporations has necessitated the training of em- ployees located in different geographic and cultural environments. In research and practice, there has been very limited attention paid to the implication of training employees with different cultural characteristics and whose work environment may differ from the home country of the transnational business. This gap is reflected in the results of a survey designed to understand the variety of issues organizations would be facing in the coming years. Among the respondents, 64% indicated that training of international employees was going to be a major issue for their organization (Petrini, 1995). The performance of employees in diverse geographic and cultural locations has the potential to influence firm performance just as much as the performance of those within the home country (Erez & Earley, 1993; Ghosal & Bartlett, 1988). This point becomes even more salient when one considers that there are an estimated 37,000 transnational corporations with as many as 207,000 affiliates worldwide (“The Economist,” 1994, p. 57), and specifically in the United States, a third of the profits of U.S. industries are derived from international trade (Casio, 1989). Consider for example, a hypothetical U.S.-based multinational organization that specializes in manufacturing electronic devices. In this firm, electronic components might be designed in a U.S. office, produced in a plant in Taiwan and then shipped for assembly to a corporate unit located in Malaysia, where labor costs are lower. Employees in all three locations require effective development, but training programs designed in the U.S. may not work equally effectively for employees in all locations. Ineffective training in any location has the potential to cause defects in the system design (USA), assembled product (Malaysia) and/or delays in production (Taiwan), and result in lost revenue. To improve firm performance, these transnational corporations have to rely on training to develop the necessary job-required skills in their worldwide employees. The Role of National Culture in Enhancing Training Effectiveness 185

While some organizations are still in the initial stages of considering training of international employees, others have implemented training modules across multiple countries (see examples in Odenwald, 1993). However, little is known about the relative effectiveness of these approaches. In reviewing the literature, one finds many studies examining training effectiveness issues in U.S. organizations, but they provide little guidance regarding what issues should be considered when implementing training programs in other cultures and national locations. Anecdotal evidence from instructors involved in global training suggests that it is important to take into account culture and language differences when delivering training in other cultures (Kemper, 2000). Employee development practices based on the home country’s values are not likely to transfer easily and need to be adapted for these to be accepted and made usable in other locations (Sarkar-Barney, 2001). Many researchers have also acknowledged the need for considering the context in which work occurs for the purpose of training (Kaplan, 1998). In introducing the idea of Cultural Ergonomics, Kaplan (1998) argues that “whenever we focus upon a particular example of work, we may speak of the milieu in which it occurs,” as culture, among other things, is likely to influence “ways of behaving and performing the task.” Additionally, there is empirical evidence of cross-cultural differences in other human-resources practices such as leadership (e.g. House et al., 1998), conflict resolution techniques (e.g. Ohbichi, Fukushima & Tedeschi, 1999), negotiation (e.g. Gelfand & Dyer, 2000), and performance management (e.g. Davis, 1998). This suggests that there are likely to be cultural influences in training practices. To better understand issues related to training employees in various cultural settings, there is a need for a framework that incorporates current research on training effectiveness along with culture theory. Such a framework has the potential to illuminate the applicability of a training-effectiveness theory across multiple cultural contexts and thereby advance science (Berry, Poortinga, Segall & Dasen, 1992). Further, practitioners may make use of this framework to guide their work to ensure that their training interventions are humane and effective. Thus, the purpose of this chapter is to present a framework dealing with the influence of culture on one aspect of training effectiveness, transfer of training. Transfer of training in this present chapter refers to the use of knowledge, skills and abilities learned in the training situation transferred onto tasks in the workplace. It is an important training outcome that indicates the level of success of a training initiative (Kirkpatrick, 1994). It is estimated that corporations worldwide allocate a 0.5–32% of the wage bill on training of employees (Gehin & Jobert, 2001; Wilkins, 2001; Xie & Wu, 2001). However, reports from trainers suggest that the level of transfer of most training efforts is typically low (Georgenson, 1982; Newstrom, 1986). To justify the return on investment related 186 SHREYA SARKAR-BARNEY to training, a better understanding of factors that influence transfer of training can greatly improve training effectiveness. The following sections will review the literature on cross-cultural psychology and transfer of training (TOT) and present propositions about the influence of culture on TOT variables.

INFLUENCE OF CULTURE: VALUE ORIENTATION

Cross-cultural researchers use value orientations to understand and explain dif- ferences among members of different cultures. Although values are by no means a universal lens for viewing culture and cultural differences, a culture-values approach has proved useful within psychology, in part because of the field’s long historical attention to the measurement and analysis of values (e.g. Rokeach, 1973). Values can be described as an overarching concept that spans a variety of situations and actions. Values influence the way people perceive and evaluate the world around them; values also help people to select appropriate actions for interacting with their surroundings. Hofstede (1984, p. 18) defines a value as a “broad tendency to prefer certain states of affairs over others.” Schwartz (1992, p. 2) has provided a more elaborate definition, suggesting that values are “desirable states, objects, goals, or behaviors, transcending specific situations and applied as normative standards to judge and to choose among alternative modes of behavior.” Theorists suggest that values do not directly influence people’s behavior (Schwartz, 1999). However, values do have an indirect influence on behavior through attitudes and goals. Additionally, the values of members of a group can influence societal structures (Hofstede, 1980) and ecological factors, such as the technological advancement of a society (Gidden, 1984). Because values link socio-cultural factors to individual behaviors, they can be considered at the individual level as well as the cultural or societal level. At the individual level, values serve as guiding principles in life that motivate and direct activities toward certain goals (Rokeach, 1973; Schwartz & Bilsky, 1987, 1990). Researchers have identified a taxonomy of values that appear to exist universally in individuals (see Schwartz, 1992; Smith & Schwartz, 1997; Triandis et al., 1985). However, not all values take equal importance in any given person’s life. Researchers propose that the nature of relationships among the different values is that of compatibility and conflict. If there exists a set of values that serves as “guiding principles” in one’s life, then there may also exist elsewhere a set of values that is opposed to the person’s fundamental beliefs and preferences (Schwartz, 1992). The basic hypothesis of cultural-value studies is that people from different cultures differ normatively in their value orientations, which ultimately causes differences in the overt behaviors of many of the people much of the time. Values The Role of National Culture in Enhancing Training Effectiveness 187 that are compatible with one another are more likely to reside in a single individual as opposed to values that are incompatible. For example, a person with high individualism, when working in a group setting may emphasize his or her own goals (e.g. promotion, pay raise) over the overall group’s goals (e.g. increasing the group’s productivity by helping a co-worker). It is unlikely that two opposing values (e.g. strongly individualistic and strongly collectivist; Triandis et al., 1985) would reside in the same individual at the same level at any given time. Cultural or societal level values represent implicit and explicit ideas shared amongst a group about what events, characteristics, and behaviors are good, right, and desirable in society (Williams, 1968, 1970). Such values are said to influence cultural norms in the society and a society’s tendency to manifest certain types of organizational or governmental structures (e.g. capitalism vs. socialism, flat vs. hierarchical organizational structure; Hofstede, 1980; Schwartz, 1994). Societal and cultural values also influence organizational structuring and practices. From an organizational-science perspective, societal values are important because they have been proposed to have a causal relationship with organizational practices by framing the nature of relationships among employees and the way activities and goals are prioritized at work (Aycan et al., 2000; Schwartz, 1999). For example, cultures that value collectivism are more likely to have worker-owned enterprises, whereas individualistic societies may prefer enterprises that operate on the principles of performance-based profit sharing (Erez, 1997).

Overview of Schwartz’s (1992, 1994) Culture Value Framework

Schwartz’s (1992, 1994) value taxonomy is useful to make predictions about the influence of culture on organizational practices such as training. Although overviews of culture-value frameworks have noted a number of similarities between Schwartz’s framework and other popular culture value frameworks, Schwartz’s framework includes explicit measurement strategies and construct def- initions for individual level values where other frameworks are more ambiguous on these points (see Smith & Schwartz, 1997). Schwartz has proposed and tested two separate taxonomies, one to measure culture values at the individual and the other to measure culture values at the societal or culture level. Schwartz’s work was originally based on previously identified values and on literature from various religious texts of the world and philosophical writings from different cultures.

Individual-Level Values One of the basic assumptions of Schwartz’s individual-level value taxonomy is that the content of each value should be determined by the motivational goals they 188 SHREYA SARKAR-BARNEY

express. Schwartz’s classification system consists of ten motivationally distinct types of values: (a) Self-direction – need for control and independence; (b) Stimulation – need for variety, arousal and excitement; (c) Hedonism – pleasure seeking and need for sensuous gratification; (d) Achievement – need for comparing personal competence with others; (e) Power – need to control others and show dominance; (f) Security – need for maintaining safety and harmony of self and others; (g) Conformity – need to control one’s actions to comply with social expectation and norms; (h) Tradition – a commitment to the practices and beliefs of the group; (i) Benevolence – concern for the welfare of other members in the group; (j) Universalism – being sympathetic and considerate of others and concern for the welfare of people and nature. According to Schwartz, the ten motivational goals can be grouped into two sets of bipolar/higher order values for a more parsimonious representation. One of the bipolar dimensions contrasts (a) Openness to change (includes self-direction and stimulation) with conservation (includes conformity, tradition, and security). It refers to “...the extent to which they (the values) motivate people to follow their own intellectual and emotional interests in unpredictable and uncertain direction vs. to preserve the status quo and the certainty it provides in relationships with close others, institutions, and traditions” (Schwartz, 1994, p. 43). The other bipolar dimension contrasts (b) Self-enhancement (achievement and power) with self- transcendence (universalism and benevolence). It refers to a focus on the interests of the self over others vs. a desire to promote the welfare of all others and of nature, even if it requires foregoing one’s own needs and interests. In line with Schwartz’s (1994) idea of the simplistic view of the value structure, the higher-order bipolar values will be used in generating propositions about the role of culture in TOT. Schwartz has developed a 56-item instrument to measure individual-level values. The scale has been validated using samples of teachers and students (N = 44,000) from 54 nations (Schwartz, 1992). Studies using Schwartz’s scale have obtained significant correlations between individual-level value orientation and certain attitudes and behaviors. A group of studies examining the relations between Schwartz’s value types and individual commitment to religion found a strong relation (r = 0.45 to r = 0.61) between the value of tradition and religios- ity (Huismans, 1994; Schwartz & Huismans, 1995). These findings supported the validity of Schwartz’s (1992) culture value scale as well as the view that individual-level values provide a suitable lens for understanding culturally-driven individual behavior. The Role of National Culture in Enhancing Training Effectiveness 189

Culture-Level Values Schwartz’s culture-level values differ from his individual level values. According to Schwartz (1992), individual-level values reflect “the psychological dynamics of conflict and compatibility that individuals experience in the course of pursuing their different values in everyday life.” In contrast, culture-level values reflect “the different solutions societies evolve to the problems of regulating human activities, the different ways that institutional emphases [sic] and investments are patterned and justified in one culture compared with another” (Schwartz, 1994, p. 92). Schwartz’s culture-level dimensions were obtained by performing Similarity Structure Analysis (SSA) on the mean value scores of each country. Therefore, the results of the culture level statistical analyses were independent of those at the individual-level. Based on the data obtained from teachers in 47 countries, seven values were obtained which were clustered into three sets of opposing values, each addressing a specific societal issue (Schwartz, 1999). Schwartz rank-ordered the countries based on their relative value (using mean country score) on each of the seven dimensions. This provided an alternative approach to that used by Hofstede for studying culture-level values. A notable advantage of the Schwartz approach is that it includes western and non-western values and the proposed structure has been cross-validated using different samples (Smith & Schwartz, 1997). A description of the values follows. Conservatism verses autonomy. The first set of values refers to the societal issue of defining the nature of the relationship between individuals and the group. Using the multidimensional data-analytic technique, Schwartz identified three values (Conservatism, Intellectual autonomy, and affective autonomy) that were related to this issue. Similar to the structural organization proposed for individual-level values to represent the conflict-compatibility relationship among values, he conceptualized the three values as located in the two polarized ends of a factor. This proposed factor is similar to Hofstede’s (1980) idea of individualism/collectivism. One end of the pole is labeled “Conservation,” it refers to an emphasis on interdependence and finding meaning in collectivities. The other pole is labeled “Openness to change.” It includes the values of intellectual and affective autonomy and refers to cultural emphasis on indepen- dence, pursuit of intellectual ideas, as well as experience of emotional pleasures (Schwartz, 1994). Hierarchy verses egalitarianism. This category of values refers to the societal issue that concerns itself with “being able to guarantee responsible behavior that will lead to the preservation of the social fabric” (Schwartz, 1994). It includes the value of hierarchy and refers to a motivational orientation that emphasizes enhancing of self-interest, and a tolerance for unequal distribution of power and wealth. The other pole includes the value of egalitarianism. It emphasizes 190 SHREYA SARKAR-BARNEY equality in the treatment of others, promoting the welfare of others and allocating resources based on the principles of equality. Mastery verses harmony. This category refers to the societal issues relating to humankind’s relationship with nature and the social world. Mastery refers to a cultural emphasis on self-enhancement and tolerance of actions that promote the self over others. Harmony refers to a focus on factors that go beyond self-interest, such as protecting the environment and promoting world peace. The identification of cultural-value orientations has been an important develop- ment in the social sciences. This development has moved the field beyond simply documenting cross-national differences and toward an understanding of why such differences exist. In his seminal work on cultural values, Hofstede (1980) found several correlates of culture that predicted differences among social, economic, and political status of a nation. One such finding relates to the association between individualism and personal freedom (r = 0.49). In a meta-analytic study of Asch-type conformity experiments, Bond and Smith (1996a) found that conformity was higher in cultures that have high scores in collectivism (␤ = 0.35) and conservatism (␤ = 0.32). In studies looking at variables more central to work behavior, Peterson et al. (1995) found that role overload was higher in high power-distance cultures (r = 0.42) and role ambiguity was greater in low power distance cultures (r =−0.42). Cultural differences have also been noted in human resource practices. Ryan et al. (1999) found that companies located in high uncertainty-avoidance cultures used interviewing as a selection tool (r = 0.37, p < 0.001) more frequently than companies in low uncertainty-avoidance cultures (Ryan et al., 1999). The study also found that companies in high power-distance cultures were less likely to use peers as interviewers (r = –0.38, p < 0.001) than those located in low power-distance cultures.

BALDWIN AND FORD’S TRANSFER OF TRAINING FRAMEWORK

Baldwin and Ford (1988) reviewed the literature of and provided a comprehensive framework for understanding TOT. Their model has served as a guiding framework for other researchers and has inspired additional investigations in the area of TOT (e.g. Noe & Wilk, 1993; Rouiller & Goldstein, 1993). In fact, the large number of studies conducted on TOT following the publication of their framework prompted a second review (see Ford & Weissbein, 1997). Baldwin and Ford’s (1988) model identified three sets of input elements that influence learning and transfer. The model states that input elements, namely, trainee characteristics (ability, motivation and self-efficacy, personality), work-environment (support & opportunity to use), and training design (content of training, learning principles) directly cause learning The Role of National Culture in Enhancing Training Effectiveness 191 and transfer of learning. Each of these linkages is described below along with research evidence.

Trainee Characteristics

This input factor refers to specific attributes of the trainee that he or she brings to the training situation. The attributes include variables such as personality, motivation, ability, skill, and self-efficacy. The general hypothesis is that trainees who are motivated to learn and who have a high aptitude for learning are more likely to learn and make use of their training. A variety of trainee characteristics (e.g. self-efficacy, locus of control, conscientiousness, job involvement, and motivation to learn) have been studied as antecedents to learning and TOT and shown to be effective predictors of training outcomes (Colquitt, LePine & Noe, 2000). Ryman and Biersner (1975) found that in Navy diving programs trainees’ confidence in their ability to successfully complete the training program was related to success in the course (r = 0.30) and drop-out rate. Noe and Schmitt (1986) found that high job-involvement was associated with greater learning following training (r = 0.45). Additionally, those with higher learning scores received higher on-the-job performance ratings (r = 0.61), suggesting that learning mediates the relation between job involvement and TOT. In recent studies, researchers have taken an additional step toward understand- ing the role of individual characteristics by looking at its interaction with the work environment and training design. Stevens and Gist (1997) conducted a study to test if high self-efficacy and low self-efficacy trainees performed differently in a salary negotiation task (for which they were trained) under conditions that emphasized performance (how well they used their skills) rather than mastery (opportunity to practice trained skills). Their results showed that the performance of the low self- efficacy trainees was lower than the high self-efficacy trainees in the performance condition ($40,200 vs. $43,200, ␤ = 0.39, p < 0.05), but there was no difference between the two in the mastery condition ($43,700 vs. $43,200, ␤ = 0.11, n.s.). Ford, Quinones, Sego and Sorra (1992) reported similar results in their study using Air Force technical training candidates. They found that trainees with higher self-efficacy sought out greater opportunities to practice their skills at work.

Training Design

This component refers to the content and instructional design aspects of a training program. Researchers in this area have incorporated literature on learning- principles and training to understand how the delivery format of instructional 192 SHREYA SARKAR-BARNEY material influences learning and TOT. According to Baldwin and Ford (1988), the key learning principles utilized in training research are: (a) identical elements; (b) teaching of general principles; (c) stimulus variability; and (d) various conditions of practice. The first two principles propose that similarity between the training condition (either in stimulus material or the general principles), and the work environment will enhance TOT to the work setting. The stimulus variability principle suggests that increasing the variety of different ways in which stimuli (i.e. learning materials) are presented will enhance learning. Providing a variety of examples is one of many ways this principle can be used. “Conditions of practice” refers to how the material is presented, in parts or all together, how many times the material is practiced, etc. In general, research results are supportive of all the learning principles. However, results suggest that the principles vary in their relative efficacy, primarily depending on the nature of the task. For example, it has been found that for simple and moderately complex tasks, performance is higher if the training material is presented in segments, also known as distributed training and practice (Briggs & Naylor, 1962; Naylor & Briggs, 1963). On the other hand, for complex tasks, it is found that exposing the trainee to the entire task (massed practice) and then allowing opportunity to practice separate segments works better (Holding, 1965).

Work/Transfer Environment Characteristics

This refers to aspects of the work environment that influence the trainees’ opportunity to use their training on their job and is often referred to as “climate” for TOT (see Rouiller & Goldstein, 1993). The construct includes factors such as supervisor support, coworker support, opportunity to use training, and negative consequences or positive reinforcement following the use of training at work. Baldwin and Ford (1988) consider this to be one of the most crucial factors for TOT on the job. At the time of their review, only a handful of studies had investigated this aspect of training. More recent studies have found support for the role of the work environment. Rouiller and Goldstein (1993) found that the combination of learning during training and organizational transfer climate explained 54% of the variance in transfer behavior. In another study using Airforce cadets, Teachout, Sego and Ford (1997) found that transfer opportunity (number of times training was used on the job) was significantly related to training effectiveness (r = 0.66). As presented above, the framework is built upon findings from existing research, which lends credibility to the model. However, it should be noted that the majority The Role of National Culture in Enhancing Training Effectiveness 193 Ford, 1988). Extending a Transfer of Training Framework to Include the Influence of Cultural Values (Adapted from Baldwin & Fig. 1. 194 SHREYA SARKAR-BARNEY of the research is based on studies from a single culture (i.e. the U.S.), which restricts the generalizability of the model (Triandis, 1994). Based on previous research, we know that culture influences people’s preferences and the way they interact with the world (see Hofstede, 1980; Schwartz, 1994; Smith & Schwartz, 1997). Because a major aspect of TOT relates to how people learn and interact with their colleagues and with technology in their work setting, it is important to understand the role of culture in global training. The following observation makes this point even more salient. In examining the applicability of Japanese training and development practices in Singapore, Putti and Yoshikawa’s (1984) research led them to conclude that competitive and individualistic Singaporeans are likely to find the Japanese “groupism” based approaches undesirable (as cited in Latham, 1988). The following sections consider each of the variables included in Baldwin and Ford’s (1988) transfer of training framework individually and integrates them with relevant literature from cross-cultural Psychology to generate specific propositions about how the two might interact (see Fig. 1). As is customary in the social science literature, the propositions immediately follow the section that discusses the variables on which they are based. Note that the proposed relationships are exploratory and need to be empirically tested. This approach to developing cross-cultural hypotheses or propositions is consistent with the paradigm proposed by Lytle et al. (1995) for testing the generalizability of a mid-range theory across cultures (One-way approach). According to Lytle et al. (1995), the one-way approach follows three steps. The first step is to select a mid-range theory or model that has been developed and validated in one culture. Alongside, a theory of culture should be identified that can help explain the cross-cultural differences in the model. The second step is to combine the two theories to “develop inductively an explanation of how culture affects the model.” The third step is to derive hypotheses about cross-cultural differences deductively, based on cultural explanations and the cultural profile of the groups to be studied.

THE INFLUENCE OF CULTURE ON TOT

Trainee Characteristics

Previous research has identified three individual difference factors (motivation, personality, and ability) that can influence TOT (Baldwin & Ford, 1988). Each of these findings is based on certain assumptions about human nature. To understand if the findings generalize to other cultures, it is important to contrast The Role of National Culture in Enhancing Training Effectiveness 195 the assumptions behind the findings with the differing assumptions that infuse motivation and behavior in different cultures. If the assumptions of the model are specific to a particular culture, then the underlying beliefs of other cultures need to be considered to fully understand the cross-cultural universality of the relationships the model proposes among variables of interest. The paragraphs below integrate literature about individual value orientations with literature on TOT to generate specific propositions about how culture might influence the relations proposed in Baldwin and Ford’s model (also see Fig. 1). Note that because trainee characteristics are individual difference factors and the goal is to explain individual behavior, consistent with multi-level theory, the propositions are stated at the individual level of analysis (see Klein, Dansereau & Hall, 1995).

Motivation Research on motivational factors suggests that individuals with high levels of motivation are more likely to learn and transfer training to the workplace. Moti- vation is typically measured by assessing a trainee’s confidence level in learning a task (see Ryman & Biersner, 1975) or the trainee’s desire to learn the training material (see Colquitt, LePine et al., 2000; Tubiana & Ben-Shakhar, 1982). The assumption is that motivation influences the choice of expending effort, as well as, directing and maintaining effort on a particular task (Kanfer & Ackerman, 1989). In a training situation, individuals who are motivated to learn will expend efforts on learning and find ways to practice and use their new knowledge (Colquitt et al., 2000). However, cross-cultural research suggests that the same attributes may not be equally motivating in all cultures (i.e. interest in learning may not be the most important sign of motivation in all cultures). Cultural values are likely to influence the interpretation and meaning assigned to various motivational techniques (Erez, 1994), which may cause each culture to assign different levels of importance to various aspects of work (Ronen, 1994). Motivation to learn can reflect an individualistic value orientation that emphasizes achievement and growth as opposed to group harmony and benefits (Casio, 1989). In a study contrasting the work preferences of groups of collectivist and individualist cultures, Meindle, Hunt and Lee (1989) found some support for the differential motivational orientation. Their data revealed that collectivist societies placed lesser emphasis on aspects of work related to achievement, independence, and growth. Com- paratively, the collectivist members placed greater emphasis on pay, benefits, security, recognition, and work conditions. The opposite pattern of relationships was obtained for individualistic societies. In another study, Haruki, Shigehisa et al. (1989) compared the effect of reinforcement on learning between Japanese and North American participants. Their findings showed that reinforcement of 196 SHREYA SARKAR-BARNEY correct responses resulted in greater learning in both groups. However, when a special condition of reinforcement was introduced in which the experimenter received a reward for the correct performance of the participants, learning was considerably higher among the Japanese participants. The findings highlight the cultural specificity in the motivating potential of a reward. The authors’ explained the differences in findings based on the cultural differences between the Japanese (a highly collectivist culture) and the North Americans (a highly individualistic culture). The collectivist focus of the Japanese may have caused them to see the experimenter as one of their own group and hence the increase in performance. Based on the above discussion, the relationship between motivation to learn and TOT best applies to cultures that stress achievement, growth, and success. Therefore, cultural value orientation of the trainee is likely to moderate the motivation-transfer relationship. It is likely that individuals who have high motivation and who also endorse values such as openness to change and self-enhancement will use training as a means to an end (e.g. to achieve higher levels of job performance). Hence, they will be more likely to transfer training to their job. Conversely, for those individuals high on motivation and who also endorse conservatism and self-transcendence values, training may be a way to enhance social relationships and maintain the harmony at work. Therefore, they may not direct their efforts toward learning and show less interest in using their new knowledge on the job. Also, for such individuals, changing the way they do work may be too disruptive of the existing harmony. This is likely to be reflected in their intentions to transfer fewer skills from the training context to their jobs. Using Schwartz’s value classification, individual values of openness to change and self-enhancement come closest to measuring a focus on personal growth and achievement. These values are opposed to the values of self-transcendence and conservatism that stress status quo and maintaining social order. Proposition 1a. Openness to change and self-enhancement values will moderate the motivation to learn and the TOT relationship such that high scores on these values will enhance the motivation-transfer relationship. Proposition 1b. Self-transcendence and conservatism will moderate the motivation to learn and the TOT relationship such that high scores on these values will attenuate the motivation-transfer relationship.

Job Involvement Another motivation-related variable that has been found to influence TOT is job involvement. Job involvement is a belief about one’s relationship with his The Role of National Culture in Enhancing Training Effectiveness 197 or her job. It is determined by the extent to which one’s job satisfies one’s needs (Kanungo, 1982). Lowered job involvement is often accompanied by job alienation, an emotional detachment from the job. In the training context, it has been proposed that individuals who have high job involvement are more likely to learn and transfer learning (Noe & Schmitt, 1986). Several researchers have proposed and tested the specific influences of job-involvement on TOT. Noe (1986) initially proposed that job involvement would influence trainees’ motivation to learn. Empirical tests of the proposed relationships have not been supportive (see Colquitt et al., 2000; Mathieu, Tannenbaum & Salas, 1992; Noe & Schmitt, 1986). These studies found a significant correlation between job-involvement and motivation (r = 0.23), but this weak relationship washed out in the presence of other predictors. However, there is some evidence that job-involvement may have a direct relation with training outcome measures. Noe and Schmitt (1986) found that trainees with high job-involvement were more likely to demonstrate greater learning following training (␤ = 0.45). Additionally, those with higher learning scores received higher on-the-job performance ratings (r = 0.61) suggesting that learning mediates the relation between job involvement and TOT. In searching the literature, no systematic studies examining cross-cultural differences in levels of job involvement were found. Considering that job- involvement is related to motivation (Mathieu, Tannenbaum & Salas, 1992; Noe & Schmitt, 1986), the hypotheses are similar to those proposed for motivation to learn. In line with Noe and Schmitt’s (1986) findings, job-involvement is proposed to have a direct influence on TOT as opposed to being mediated by motivation to learn. Proposition 2a. Openness to change and self-enhancement values will moderate the job involvement and TOT relationship such that high scores on these values will enhance the job involvement-transfer relationship. Proposition 2b. Self-transcendence and conservatism values will moderate the job involvement and TOT relationship such that high scores on these values will attenuate the job involvement-transfer relationship.

Personality Personality is an individual characteristic that the trainees bring to the learning situation. A number of personality variables such as conscientiousness, locus of control, need for achievement, and anxiety have been found to be predictive of training outcomes (Colquitt et al., 2000). Cross-cultural comparisons of the personality frameworks suggest that the models are universal (McCrae & Costa, 1997; McCrae, Costa, delPilar, Rolland & Parker, 1998; Paunonen & Ashton, 198 SHREYA SARKAR-BARNEY

1998). However, cross-cultural researchers have expressed concern with this perspective. Because people operate in a “cultural medium” that influences their thoughts and actions, it is difficult to assume that traits do not covary with culture (Fiske, 1995). The intersection of personality with culture has only recently begun to receive research attention (Lee, McCauley & Graguns, 1999). Efforts to identify culturally unique personality dimensions have yielded new factors not captured by the “Big Five,”1 the most popularly used model of personality. One such factor, is the identification of the “interpersonal relatedness” factor in studies that have originated in the Chinese culture. This factors refers to relationship orientation, seeking inner peace and internal harmony, as well as interpersonal behavior directed toward enhancing one’s face and avoiding losing face (Cheung et al., 2001). The importance of social relationships in self-construal has also been discussed in reference to other cultures. Okeke, Draguns and Sheku (1999) synthesized quantitative and qualitative information related to culture and personality in Africa. They describe the notion of the “self” of an African person to be that of a “sociocentric person” that is influenced by the person’s social relationships and attachments. Similar observations about a social self-concept have been reported in the Mexican culture (Diaz-Loving & Graguns, 1999). Other culture-specific personality constructs such as a physical dimension of personality (Diaz-Loving & Graguns, 1999), have also been suggested but have yet to be investigated empirically. To understand how personality differences across cultures can influence learning and training transfer, one must consider the mechanism by which personality operates. The influence of personality is motivational. It creates individual differences in self-set goals and influences the subsequent availability of cognitive resources for task engagement. This leads to differences in learning and transfer behaviors (Kanfer, 1991). Because personality and value both have a motivational effect, they are likely to have congruent influences on the criterion when matched. Similar to self-enhancement values, Conscientiousness, need for achievement and locus of control (internal) are all personality variables that emphasize self-enhancement goals such as self-discipline, achievement striving and reliance on personal efforts. However, individuals who are high on the interpersonal-relatedness factors are likely to emphasize social-goals (similar to self-transcendence values) over self-oriented goals (Okeke et al., 1999). This qualitative difference in goal characteristics is likely to result in differential attention to the learning task. Those who are self-enhancement oriented will devote more focused attention on the learning task, as it is likely to lead to valued outcomes for the self (e.g. praises from the manager, career advancement, and raises). Conversely, individuals who are socially oriented will divide their The Role of National Culture in Enhancing Training Effectiveness 199 attention between the training task and off-task issues (e.g. maintaining social relations). This is likely to result in lowered levels of training-related outcomes. Proposition 3a. Individuals who score high on the self-transcendence value dimension will also score high on the interpersonal-relatedness personality dimension. Such individuals will demonstrate lower training transfer. Proposition 3b. Individuals who score high on the self-enhancement value dimension will also score high on the personality factors of: (a) conscien- tiousness; (b) need for achievement; and (c) internal locus of control. Such individuals will demonstrate higher training transfer.

Cognitive Ability Although researchers debate about the relative importance of general and specific cognitive ability in predicting training outcomes (Barrett et al., 1999; Ree & Earles, 1991), the findings are overwhelmingly in support of the importance of cognitive ability in learning, an antecedent to training transfer (Colquitt et al., 2000). A variety of propositions have been made about the relationship between cognitive ability (as measured by cognitive test performance) and culture. Some have proposed that tasks requiring higher cognitive complexity are likely to reveal cultural differences (see Van De Vijver, 1997). Others have proposed that measurement inaccuracies are likely to lead to cross-cultural differences in test scores (Jensen, 1985; Mercer, 1984). Still others have suggested that cognitive ability differences are associated with education level and socio-economic status (Jensen, 1977). A common thread among these perspectives is a focus on score differences in cognitive ability tests between ethnic groups or cultures, as opposed to the presence of a differential construct of cognitive ability across cultures. Support for the differential argument would suggest that different aspects of cognition should be considered when training in diverse cultures. While some have argued for this perspective, it has not received empirical support. Helms (1992) proposed that cognitive ability tests that emphasize social relations and reasoning in social context are likely to lead to higher scores in African-American test takers. However, studies comparing score differences in cognitive ability tests using social items (e.g. there are 3 mangoes that are to be shared equally between two brothers. How many mangoes does each brother get?) found differences to persist between black and white participants (DeShone et al., 1998). Based on this review, it can be suggested that the predictive relationship between cognitive ability and learning and transfer is most likely to be universal. Proposition 4. There will be no systematic influence of culture in the rela- tionship between cognitive ability and TOT. Cognitive ability will be similarly predictive of training transfer in all cultures. 200 SHREYA SARKAR-BARNEY

Training Design

Training design refers to the design and delivery of the instructional material. Published literature in this area has primarily focused on the application of learn- ing principles (i.e. identical elements, teaching of general principles, stimulus variability, and conditions of practice) to the training situation. In general, the use of these principles in designing training has been found to lead to greater learning and transfer (Baldwin & Ford, 1988). In this section, I outline the mechanism by which one of the learning principles (identical elements) might explain differ- ences in training transfer across cultures. In addition, I discuss the importance of considering cross-cultural differences in cognitive styles as well as cultural variations in trainer-trainee relationships and their influences on training design and TOT (see Fig. 1). It should be noted that the propositions related to training design are generated at the culture-level. Due to high costs associated with developing training programs, it is cheaper to tailor training to meet the needs of a group rather than an individual. Therefore, the proposed relationships dealing with training design are intended to apply to cultures as distinct from individuals (Klein et al., 1995; Smith & Schwartz, 1997).

Principles of Learning/Training Content The identical elements principle of learning suggests that similarity in stimuli (or principles learned during training) between the training environment and the work setting are more likely to cause transfer of learning to the workplace (Thorndike & Woodworth, 1901). Combining the principle of identical elements with literature on transfer of technology in a multination setting can help us to understand its cross-cultural implications for training. In their theory of transfer of technology across nations, Kedia and Bhagat (1988) proposed that any technology that is transferred from a supplier to a recipient organization can be classified into three types: that is, product-embodied, process-embodied, and person-embodied. Kedia and Bhagat (1988) referred to product-embodied technologies as the transfer of actual products such as com- puter components or machinery. Process-embodied technology transfer includes “transfer of blue-prints or patent rights of the actual scientific process ...” (Kedia & Bhagat, 1988, p. 561). Person-embodied technology transfer refers to a supplier organization’s efforts to develop the relevant technical competencies among members in the recipient organization “...to implement and diffuse the imported technology” (Kedia & Bhagat, 1988, p. 562). They suggested that, compared to product-oriented technologies, transferring process and person-oriented technology should be harder. This is because the latter two forms of transfer The Role of National Culture in Enhancing Training Effectiveness 201 require considerable interaction between supplier and recipient organizations, which can be inhibited due to cultural differences (Kostova, 1999). Similar to the principle of identical elements in learning theory, lack of similarity reduces the possibility of transfer between the originating and receiving culture. They reason that this is because of the differential absorptive capacities of each culture. Cultures that are open to change and innovation have higher absorptive capacities than those resistant to change (Kedia & Bhagat, 1988; Kostova, 1999). There are many similarities between the process of technology transfer and transfer of training back on the job. Both involve the implementation of new knowledge, skills, and behaviors that are generated/learned in one setting and are put to use in a different work setting (Kedia & Bhagat, 1988; Newstrom, 1986). Similar to the classification of the types of technologies, organizational training content can be placed into the bins labeled product-focused (learning about a new product such as an automated teller machine), process-focused (learning to use a new software), and person-focused (learning interpersonal skills such as negotiation techniques). To take advantage of economies of scale, many multinational organizations develop training programs in their headquarters and implement standardized training modules in all their global locations (Odenwald, 1993). However, using the paradigm created by Kedia and Bhagat (1988) it may be supposed that TOT will be inhibited with increasing “dissimilarity” between the training content’s originating culture and the recipient culture. This is because trainees in the recipient culture may find the training content less relevant or applicable to their situations and therefore be less willing to apply the new knowledge. However, based on the idea of “absorptive capacities,” recipient cultures that are “open to change” will demonstrate greater implementation of training knowledge than those valuing traditional attitudes. Note that the proposition related to the influence of culture on training content is not based on specific culture values but on cultural dissimilarity between the supplier and recipient country. Cultural dissimilarity refers to a pattern of difference across all seven culture level values proposed by Schwartz (1994).

Proposition 5. TOT will be highest in the supplier country and will progres- sively decline as the cultural dissimilarity increases between the supplier and recipient country. This decline in TOT will be highest for person-oriented courses followed by process-oriented courses and the least for product- oriented courses.

Cognitive Style Effective training design is an important element in determining learning outcomes (Baldwin & Ford, 1988). Training design influences what content is taught and 202 SHREYA SARKAR-BARNEY how it is taught. Research has helped identify a variety of training design factors that are likely to increase learning. To promote concept formation during learning, experts (see Ackerman, 1991; Anderson, 1989; Anderson & Krathwohl, 2001) recommend organizing the training content using the dimensions of knowledge (factual, declarative, procedural, and metacognitive). Additionally, structuring training content and delivery using principles such as the identical elements theory, transfer through principles, sequencing and stimulus variability have been found to lead to better learning and transfer (see Baldwin & Ford, 1988; Goldstein & Ford, 2001). Recent research in cross-cultural and cognitive psychology suggests that additional factors such as culture-specific cognitions should be considered, to understand the learning process (Nisbett et al., 2001; Norenzayan et al., 2000). Previously, cognitive processes such as learning, reasoning, and making inferences were assumed to be “absolute” or invariant across individuals. New research suggests that socialization practices and formal training can influence reasoning and logical thinking. According to sociologists, Inkeles and Levinson (1969), socialization influences what information we pay attention to and how we deal with information, which is likely to lead to cross-cultural variation in cognitive processing. Research by Nisbett and his colleagues provides some support for the “culture variation” view. According to Nisbett et al.’s (2001) findings, Asians use a “holistic” pattern of thinking. They tend to consider all aspects of a situation and understand the behavior of the objects by assigning causal relations among them. They are more likely to use prior information than logic for this. In contrast, West- erners are likely to be more “analytic.” They put greater emphasis on the objects rather than the context, and focus on categorizing information using rules and logic to understand the behavior of the objects. Several other studies have found empirical support for this cultural distinction in cognition. Strohschneider and Guss (1999) conducted a study to compare problem-solving strategies between Indians and Germans. They found that Germans asked more questions related to the specific aspects of the scenario, thus demonstrating analytical thinking. In contrast, Indians asked more questions pertaining to the background of the scenario, which suggests that they engaged in holistic thinking. Okeke et al. (1999) in their qualitative analysis of the African culture, found a prevalence of holistic thinking. They report that Africans show a penchant toward synthesis rather than analysis, which is demonstrated by their emotional and intuitive approach to reasoning. Nisbett et al.s’ (2001) findings related to holistic and analytic cognition have implications for training design. In one study Norenzayan, Nisbett, Smith and Kim (2000) taught American (Asian and European) and East Asian participants to The Role of National Culture in Enhancing Training Effectiveness 203 classify animals using an exemplar-based strategy (to promote holistic thinking) and a rule-based strategy (to promote analytic thinking). The study found no differences between the Asian and American participants in the exemplar-based approach. However, Asian participants made more errors in classification and had a slower speed in the rule-based condition. This suggests that matching training design elements with the culture of the trainees can lead to higher learning outcomes. Failure to do so can place one group at a disadvantage if asked to learn material using a strategy that does not match their dominant learning style. Previous investigation of cognitive style differences across cultures has been based on between-country comparisons as opposed to cultural-value differences. To better understand which aspect of culture is likely to lead to these differences, it is important to consider the role of values (Hofstede, 1980). Therefore, in presenting the proposed relationships, general terms such as Asian or Western culture have been replaced by the dominant cultural values represented by countries in these geographic regions. Using Schwartz’s (1994) mean culture- value scores, Asian (e.g. Taiwan, Malaysia) cultures tend to score higher on the values of conservatism and harmony, whereas, developed western cultures (e.g. Germany and United States) score higher on the values of autonomy and mastery.

Proposition 6. Trainees in cultures that are high on conservatism and harmony will demonstrate greater learning and transfer when the training design evokes holistic thinking. In contrast, trainees in cultures that are high on autonomy and mastery will demonstrate greater learning and transfer when the training design evokes analytic thinking.

Trainer-Trainee Relationship In cross-cultural training situations, the relation between trainers and trainees is an additional factor that relates to training design, which can impact learning and TOT. Practitioners who are involved in global training have found that the cultural background of the students or trainees determines the type of instruction format to which they are most receptive (Kemper, 1998). Using approaches that are successful in one culture does not imply that they will be effective in another culture. For example, the experiential approach to teaching, which is popular in the U.S., is found to be best suited for individualistic and low power distance cultures (Francis, 1995). In these cultures, trainees are more open to new experiences, and the social order between the trainer and trainee is small (low power distance). Hence, there is free exchange of ideas. Collective societies with greater power distance may prefer a more didactic approach, where the student’s role is more passive (Francis, 1995; Hofstede, 1986). 204 SHREYA SARKAR-BARNEY

The differences in perceptions of effective teacher-student communication due to cultural orientation were demonstrated by one empirical study. It compared communication styles of Chinese and Australian students with their university lecturers and found that each of the student groups had different ideas about effective communication (Gallois et al., 1990). The Chinese students preferred passive forms of communication, which they perceived as showing respect toward the teacher. In contrast, Australian students and lecturers believed that, compared to the passive students, the assertive students were more talented because they were willing to challenge the teacher. Additionally, the Australian lecturers indicated a greater willingness to help the assertive students than those who were passive. Findings such as these suggest that designers of training systems should not only focus on the content of the training program but also pay attention to aligning the delivery format and trainee-trainer interactions with the cultural orientation of the learner (Hofstede, 1986).

Proposition 7. Trainees in hierarchical cultures (Schwartz, 1992) will demon- strate greater learning and transfer when training involves trainer-centered approaches compared to trainees in egalitarian cultures. This relationship will be opposite for trainees in egalitarian cultures, where trainee-centered approaches are likely to lead to greater learning and TOT.

Work Environment

The work environment refers to the job context in which training participants operate on a regular basis. Practitioners have stressed the importance of a sup- portive work environment to facilitate implementation of training knowledge in the workplace (Eddy et al., 1967; Georgenson, 1982; Newstrom, 1986). Findings from empirical studies have been congruent with this assertion. A supportive TOT climate where the social environment is conducive to implementation of new learning, and the nature and design of the job provides opportunities to practice new knowledge has been linked to positive TOT (Baumgartel & Jeanpierre, 1972; Baumgartel, Reynolds & Pathan, 1984; Huczynski & Lewis, 1980; Noe & Wilk, 1993; Rouiller & Goldstein, 1993; Tesluk et al., 1995; Tracey et al., 1995). Given that the relationship between the trainee and the work environment involves interpersonal relations, especially between people of different hierarchical status, socio-cultural factors are likely to play a major role in the work-environment- training outcome relationship (Hofstede, 1980, 1991; Schwartz, 1992; Tayeb, 1988). Additionally, recent literature on TOT has gravitated toward considering various aspects of the work environment related to training under the single The Role of National Culture in Enhancing Training Effectiveness 205 construct called “climate for transfer.” For the purpose of understanding cross- cultural influences on the work environment-TOT relationship, it is more useful to look at the separate components of transfer climate because there may be distinct influences of culture on each of the factors. Also, differentiating between the various factors (i.e. relative importance of manager, co-workers, peers, the work itself) can be beneficial for developing targeted organizational interventions. The paragraphs below offer specific propositions about how culture might influence each of the work environment variables listed in Baldwin and Ford’s (1988) framework (also see Fig. 1).

Supervisory Support Western academic literature has identified several supervisory behaviors that can be linked to positive TOT in trainees. Such supervisory behaviors include openness to ideas, providing freedom of work, and actively supporting implementation of new knowledge (Baumgartel & Jeanpierre, 1972; Baumgartel et al., 1984; Huczynski & Lewis, 1980; Tesluk et al., 1995; Tracey et al., 1995). A common assumption among all of these studies is that supervisors as a group agree on the positive benefits of training. A recent cross-cultural investigation of managerial values provides information that contradicts this assumption. Aycan et al. (2000) examined managers’ cultural values and beliefs about work and human resource practices (related to job design, supervision and control, and performance-reward contingency) in 10 countries. Striking differences were found between nations with respect to managers’ values and their beliefs about work. Some of the key findings were that in paternalistic cultures (where the supervisor takes on a role similar to parents), such as Japan and India, managers were more likely to engage in employee-empowering practices. In contrast, managers from cultures characterized as fatalistic (believing that that the future cannot be changed, whatever happens must happen) such as India and Russia, were less in favor of employee participation and also were opposed to the view that employees’ behaviors are malleable. The latter finding has several cross-cultural implications with respect to managerial support for employee development activities. Managers who perceive that people’s behavior cannot be changed are less likely to have faith in training or other forms of change intervention. Note that the construct of fatalism may be considered similar to Schwartz’s (1992) cross-cultural factor contrasting harmony and mastery. The mastery dimension emphasizes “...getting ahead through active self-assertion (ambition, success, daring, competence)”; conversely, the harmony dimension emphasizes accepting “...the world as it is, trying to preserve rather than to change or exploit” (Smith & Schwartz, 1997, p. 100). Based on these definitions, fatalism and harmony may be considered as similar constructs in the 206 SHREYA SARKAR-BARNEY sense that they refer to a culture’s belief in the malleability of the world around them. This is also supported by Schwartz’s (1999) view on the relationship between harmony and work centrality. According to him, emphasizing harmony values conflicts with “viewing work as central to life: work generally aims to modify the material and social environment” (pp. 40–41). Comparing national indices of work centrality with cultural profiles of the country (based on Schwartz, 1994), he found that nations that had the highest work-centrality scores (Japan, USA, & Germany) also had low scores on harmony (Schwartz, 1999).

Proposition 8. There will be greater managerial support for training transfer in cultures that value mastery compared to those that value harmony.

Support from Co-Workers Literature in TOT has generally assumed that support of supervisors, peers, and subordinates are all equally important for positive transfer (Baldwin & Ford, 1988; Rouiller & Goldstein, 1993). However studies using different European samples have found non-converging results, which hints at the existence of cross-cultural differences. In a sample of British executives, Huczynski and Lewis (1980) found that the support of peers and subordinates was of less importance than that of the super- visor. However, the authors point out that literature on the relative importance of managers, peers, and subordinates is quite mixed. While, Vandenput’s (1973) study on Belgian managers shows that the learner’s relationship with his/her supervisor is most crucial for positive transfer, other British studies have shown that peer support is more important than supervisory support for innovation adoption. To understand the relative importance of supervisors and co-workers, Huczynski and Lewis (1980) recommended looking at the role of the work environment among other factors to understand inhibitors and facilitators of TOT. Because the issue relates to the work roles of incumbents who have different levels of authority (i.e. power distance), there are likely to be cross-cultural differences (Hofstede, 1991). Cross-referencing the cultural value profile of Great Britain with that of Belgium on Hofstede’s (1991, p. 26) power distance (PD) index explains the differences in findings. On a scale of 0 (low PD) to 100 (high PD) the cultural focus on power distance is higher in Belgium (65) compared to Great Britain (35). Therefore, because of the greater authority that supervisors have in Belgium, their support is likely to be more important than that of the peers. On the other hand, in Great Britain, because of low PD and the interdependent nature of organizational work, support of co-workers may be more important for implementing new knowledge. Therefore, the relative importance of supervisors and co-workers on TOT is likely to vary by culture. The Role of National Culture in Enhancing Training Effectiveness 207

Proposition 9. The relationship between co-worker support and TOT will be stronger in egalitarian cultures compared to hierarchical cultures.

Advancement Opportunities In addition to social support, another aspect of the work environment linked to positive TOT is positive reinforcement. Advancement opportunities such as raises and promotions to reinforce attitudes and behaviors learned in training have been associated with increased and sustained implementation of training knowledge (Hand, Richards & Slocum, 1973). Implicit in the above relationship is a performance-reward contingency, a principle that has not been found to have universal appeal (Erez, 1994, 1997). Studies comparing reward allocation in different cultures have found that collectivist cultures (e.g. China, Japan, the Kibbutz community in Israel) prefer reward distribution based on principles of equality (to each equally). In individual- istic cultures (e.g. U.S., Great Britain) the preferences are stronger for distribution of rewards based on the principles of equity or extent of contribution to the work (Bond, Leung & Wan, 1982; Erez, 1997; Leung & Bond, 1984). In the context of TOT, selectively providing positive reinforcement to those who implement the newly learned knowledge and skills may have a negative influence on trainees in collectivist cultures. Such attempts may be seen as ways to discriminate among the members of the in-group. However, as demonstrated in the western literature, there may be a positive association between rewards and TOT in individualistic cultures. Proposition 10. The relationship between advancement opportunities and TOT will be stronger in hierarchical cultures compared to egalitarian cultures.

CONCLUSION

The goal of this chapter was to present a framework that considers the influence of culture on transfer of training. The growth in transnational business has created a need for training employees in a variety of cultural locations. However, the implication of cultural differences on training practices is not well understood. Using a multi-level perspective, this chapter outlines the factors that are likely to influence the use of newly learned knowledge and skills at work by considering the characteristics of the individual, the training, and the work environment of the employee. Previous studies by cross-cultural researchers have provided a taxonomy of culture values as well as indices of scores that show a country’s standing on the culture values. By combining information about a country’s culture 208 SHREYA SARKAR-BARNEY value scores and the relationships proposed by the framework, practitioners can make more informed decisions about ways of adapting their training systems to meet the needs of any particular culture in which training is to be conducted. Additionally, researchers can use this information to empirically test the proposed relationships and make refinements to the model.

NOTE

1. The “Big Five” dimensions of personality include Openness to experience, con- scientiousness, extraversion, agreeableness and emotional stability. See McCrae and Costa (1987).

ACKNOWLEDGMENTS

The author acknowledges the helpful comments of Jeffrey M. Stanton and Matthew F. Barney on earlier drafts of this material.

REFERENCES

Anderson, L. W., & Krathwohl, D. R. (2001). A taxonomy for learning and teaching and assessing. New York: Longman. Baldwin, T. T., & Ford, K. J. (1988). Transfer of training: A review and direction for future research. Personnel Psychology, 41, 63–105. Barrett, G. V., Polomsky, M. D., & McDaniel, M. A. (1999). Selection tests for firefighters: A comprehensive review and meta-analysis. Journal of Business & Psychology, 13(4), 507–514. Baumgartel, H., & Jeanpierre, F. (1972). Applying new knowledge in the back-home setting: A study of Indian managers’ adoptive efforts. Journal of Applied Behavioral Science, 8(6), 674–694. Baumgartel, H., Reynolds, M., & Pathan, R. (1984). How personality and organizational-climate variables moderate the effectiveness of management development programs: A review of some recent research findings. Management and Labour Studies, 9, 1–6. Berry, J. H., Poortinga, Y. H., Segall, M. H., & Dasen, P. R. (1992). Cross-cultural psychology: Research and application. New York: Cambridge University Press. Black, S. J., & Mendenhall, M. (1990). Cross-cultural training effectiveness: A review and a theoretical framework for future research. Academy of Management Review, 15(1), 113–136. Bond, M. H., Leung, K., & Wan, K. C. (1982). Hoe does cultural collectivism operate? The impact of task and maintenance contributions on reward allocation. Journal of Cross-Cultural Psychology, 13, 186–200. Briggs, G. E., & Naylor, J. C. (1962). The relative efficiency of several training methods as a function of transfer task complexity. Journal of Experimental Psychology, 64, 505–512. The Role of National Culture in Enhancing Training Effectiveness 209

Brinkerhoff, R. O., & Gill, S. J. (1994). The learning alliance. San Francisco: Jossey-Bass. Casio, W. F. (1989). Managing human resources: Productivity, quality of work life, profits. New York: McGraw-Hill. Cheung, F. M., Leung, K., Zhang, J., Sun, H., Gan, Y., Song, W., & Xie, D. (2001). Indigenous Chinese personality constructs: Is the five factors model complete? Journal of Cross-Cultural Psychology, 32(4), 407–433. Colquitt, J. A., LePine, J. A., & Noe, R. A. (2000). Toward an integrative theory of training motivation: A meta-analytic path analysis of 20 years of research. Journal of Applied Psychology, 85(5), 678–707. Davis, D. D. (1998). International performance measurement and management. In: J. W. Smither (Ed.), Performance Appraisal (pp. 95–131). San Francisco, CA: Jossey-Bass. DeShone, R. P., Smith, M. R., Chan, D., & Schmitt, N. (1998). Can racial differences in cognitive test performance be reduced by presenting problems in a social context. Journal of Applied Psychology, 83(3), 438–451. Diaz-Loving, R., & Graguns, J. G. (1999). Culture, meaning, and personality in Mexico and in the Unites States. In: Y.-T. Lee, C. R. McCauley & J. G. Draguns (Eds), Personality and Person Perception Across Cultures (pp. 103–126). New Jersey: Lawrence Erlbaum. Dipboye, R. L. (1997). Organizational barriers to implementing a rational model of training. In: M. A. Quiones & A. Ehrenstein (Eds), Training for a Rapidly Changing Workplace (pp. 31–60). Washington, DC: American Psychological Association. Eddy, W. B., Glad, D. D., & Wilkins, D. D. (1967). Organizational effects on training. Training and Development Journal, 22(2), 36–43. Erez, M. (1994). Toward a model of cross-cultural Industrial and Organizational Psychology. In: H. C. Triandis, M. D. Dunnette & L. M. Hough (Eds), Handbook of Industrial and Organizational Psychology (2nd ed., pp. 559–608). Palo Alto, CA: Consulting Psychologists Press. Erez, M. (1997). A culture-based model of work motivation. In: P. C. Earley & M. Erez (Eds), New Perspectives in Industrial/Organizational Psychology (pp. 193–242). San Francisco: New Lexington Press. Erez, M., & Earley, P. C. (1993). Culture, self-identity, and work (p. 4). New York: Oxford University Press. Fiske, A. P. (1995). The cultural dimensions of psychological research: Method effects imply cultural mediation. In: P. E. Shrout & S. T. Fiske (Eds), Personality Research, Methods, and Theory (pp. 271–294). Hillsdale, NJ: Lawrence Erlbaum. Ford, J. K., & Weissbein, D. A. (1997). Transfer of training: An updated review and analysis. Performance Improvement Quarterly, 10(2), 22–41. Ford, K. J., Quinones, M. A., Sego, D. J., & Sorra, J. S. (1992). Factors affecting the opportunity to perform trained tasks on the job. Personnel Psychology, 45, 511–527. Francis, J. L. (1995). Training across cultures. Human Resource Development Quarterly, 6(1), 101–107. Gehin, J.-P., & Jobert, A. (2001). International briefing 8 Training and development in France. International Journal of Training and Development, 5(1), 81–93. Gelfand, M. J., & Dyer, N. A. (2000). A cultural perspective on negotiation: Progress, pitfalls, and prospects. Applied Psychology: An International Review, 49(1), 62–99. Georgenson, D. L. (1982). The problem of transfer calls for partnership. Training and Development, 36(10), 75–78. Ghosal, S., & Bartlett, C. (1988). Creation, adoption, and diffusion of innovations by subsidiaries of multinational corporations. Journal of International Business Studies, 19, 365–388. Gidden, A. (1984). The constitution of society. Cambridge: Cambridge University Press. 210 SHREYA SARKAR-BARNEY

Goldstein, I., & Ford, K. I. (2001). Training in organizations (4th ed.). Wadsworth. Hand, H. H., Richards, M. D., & Slocum, J. M. (1973). Organizational training and the effectiveness of a human relations program. Academy of Management Journal, 16, 185–195. Haruki, T., Shigehisa, R., Nedate, K., Wajima, M., & Ogawa, R. (1989). Effects of alien reinforce- ment and its combined types on learning behavior and efficiency in relation to personality. International Journal of Psychology, 19, 527–545. Helms, J. E. (1992). Why there is no study of cultural equivalence in standardized cognitive ability testing? American Psychologist, 47, 1083–1101. Hofstede, G. (1980). Culture’s consequences: International differences in work related values. Beverly Hills: Sage. Hofstede, G. (1984). The cultural relativity of the quality of life concept. Academy of Management Review, 9, 389–398. Hofstede, G. (1986). Cultural differences in teaching and learning. International Journal of Intercultural Relations, 10, 301–320. Hofstede, G. (1991). Cultures and organizations: Software of the mind. London: McGraw-Hill Book Company. Holding, D. H. (1965). Principles of training. London: Pergamon Press. House, R. J. et al. (1998). Cultural Influences in leadership and organizations: Project GLOBE. In: W. H. Mobley, M. J. Gessner & V. Arnolds (Eds), Advances in Global Leadership (Vol. 1, pp. 171–233). Stanford, CT: JAI Press. Huczynski, A. A., & Lewis, J. W. (1980). An empirical study into the learning transfer process in management training. Journal of Management Studies, 17, 227–240. Huismans, S. (1994). The impact of differences in religion on the relation between religiosity and values. In: A.-M. Bouvy, F. Van de Vijver, P. Boski & P. Schmitz (Eds), Journeys into Cross-Cultural Psychology (pp. 255–267). Lisse, The Netherlands: Swets & Zeitlinger. Inkeles, A., & Levinson, D. J. (1969). National character: The study of modal personality and social systems. In: G. Lindsey & E. Aronson (Eds), The Handbook of Social Psychology (2nd ed., Vol. 4). Reading, MA: Addison-Wesley. Jensen, A. R. (1977). An examination of culture bias in the Wonderlic Personnel test. Intelligence, 1, 51–64. Jensen, A. R. (1985). The nature of Black-White differences on various psychometric tests: Spearman’s hypothesis. Behavioral and Brain Sciences, 8, 193–263. Kanfer, R. (1991). Motivation theory and industrial and organizational Psychology. In: M. D. Dunnette & L. M. Hough (Eds), Handbook of Industrial and Organizational Psychology (Vol. 1, pp. 75–170). Palo Alto, CA: Consulting Psychologist Press. Kanfer, R., & Ackerman, P. L. (1989). Motivation and cognitive abilities: An integrative/aptitude- treatment approach to skill acquisition. Journal of Applied Psychology, 74(4), 657–690. Kanungo, R. N. (1982). Measurement of job and work involvement. Journal of Applied Psychology, 67(3), 341–349. Kaplan, M. (1998, September). The cultural milieu of ergonomics. Paper presented at the Global Ergonomics Conference, Cape Town, South Africa. Kedia, B. L., & Bhagat, R. S. (1988). Cultural constraints on transfer of technology across nations: Implications for research in organizational behavior. Academy of Management Review, 13, 559–571. Kemper, C. L. (1998). Global training’s critical success factors. Training and Development, 52, 35–37. Kirkpatrick, D. L. (1994). Evaluating training programs: The four levels. San Francisco: Berret- Koehler. The Role of National Culture in Enhancing Training Effectiveness 211

Kostova, T. (1999). Transnational transfer of strategic organizational practices: A contextual perspective. Academy of Management Journal, 24(2), 308–324. Latham, G. P. (1988). Human resource training and development. Annual Review of Psychology, 39, 545–582. Lee, Y.-T., McCauley, C. R., & Draguns, J. G. (1999). Personality and person perception across cultures. New Jersey: Lawrence Erlbaum. Leung, K., & Bond, M. H. (1984). The impact of cultural relativism on reward allocation. Journal of Personality and Social Psychology, 47, 793–804. Lytle, A. L., Brett, J. M., Barsness, Z. I., Tinsley, C. H., & Janssens, M. (1995). A paradigm for confirmatory cross-cultural research in organizational behavior. Research in Organizational Behavior, 17, 167–214. Mathieu, J. E., Tannenbaum, S. I., & Salas, E. (1992). Influence of individual and situational character- istics on measures of training effectiveness. Academy of Management Journal, 35(4), 828–847. Mercer, J. R. (1984). What is a racially and culturally nondiscriminatory test? A sociological and pluralistic perspective. In: C. R. Reynolds & R. T. Brown (Eds), Perspectives on Bias in Mental Testing (pp. 293–356). New York: Plenum Press. McCrae, R. R., & Costa, P. T. (1987). Validation of the five-factor model of personality across instruments and observers. Journal of Personality & Social Psychology, 56, 589–595. McCrae, R. R., & Costa, P. T., Jr. (1997). Personality trait structure as a human universal. American Psychologist, 52, 509–516. McCrae, R. R., Costa, P. T., Jr., delPilar, G. H., Rolland, J. P., & Parker, W. D. (1998). Cross-cultural assessment of the five-factors model: The revised NEO personality inventory. Journal of Cross-Cultural Psychology, 29(1), 171–188. McGehee, W., & Thayer, P. W. (1961). Training in business and industry. New York: Wiley. Meindle, J. R., Hunt, R. G., & Lee, W. (1989). Individualism-collectivism and Work values: Data from the United States, China, Taiwan, Korea and Hong Kong. Research in Personnel and Human Resource Management, 1, 59–77. Naylor, J. C., & Briggs, G. E. (1963). The effect of task complexity and task organization on the relative efficiency of part and whole training methods. Journal of Experimental Psychology, 65, 214–217. Newstrom, J. W. (1986). Leveraging management development through the management of transfer. Journal of Management Development, 5(5), 33–45. Nisbett, R. E., Peng, K., Choi, I., & Norenzayan, A. (2001). Culture and systems of thought: Holistic and analytic cognition. Psychological Review, 108(2), 291–310. Noe, R. A. (1986). Trainees’ attributes and attitudes: Neglected influences on training effectiveness. Academy of Management Review, 11(4), 736–749. Noe, R. A., & Schmitt, N. (1986). The influence of trainee attitudes on training effectiveness: Test of a model. Personnel Psychology, 39, 497–523. Noe, R. A., & Wilk, S. L. (1993). Investigation of the factors that influence employee’s participation in development activities. Journal of Applied Psychology, 78(2), 291–302. Norenzayan, A., Nisbett, R. E., Smith, E. E., & Kim, B. J. (2000). Rules vs. similarity as a basis for reasoning and judgement in East and West. Unpublished manuscript, University of Illinois. Odenwald, S. B. (1993). Global training: How to design a program for the multinational corporation. Alexandria, VA: American Society for Training and Development. Ohbichi, K., Fukushima, O., & Tedeschi, J. T. (1999). Cultural values in conflict management Goal orientation, goal attainment, and tactical decisions. Journal of Cross-Cultural Psychology, 30(1), 51–71. 212 SHREYA SARKAR-BARNEY

Okeke, B. I., Graguns, J. G., & Sheku, B. (1999). Culture, self, and personality in Africa. In: Y.-T. Lee, C. R. McCauley & J. G. Draguns (Eds), Personality and Person Perception Across Cultures (pp. 139–162). New Jersey: Lawrence Erlbaum. Paunonen, S. V., & Ashton, M. C. (1998). The structured assessment of personality across cultures. Journal of Cross-Cultural Psychology, 29(1), 150–170. Petrini, C. M. (1995). The rising tide of global training. Training & Development, 49(2), 9–10. Ree, M. J., & Earles, J. A. (1991). Predicting training success: Not much more than g. Personnel Psychology, 44, 321–332. Rhinesmith, S. H. (1996). Training for global operations. In: C. L. Robert (Ed.), The ASTD Training and Development Handbook (4th ed.). New York: McGraw-Hill. Rokeach, M. (1973). The nature of human values. New York: Free Press. Ronen, S. (1994). An underlying structure of motivational need taxonomies: A cross-cultural confirmation. In: H. C. Triandis, M. D. Dunnette & L. M. Hough (Eds), Handbook of Industrial and Organizational Psychology (2nd ed., pp. 214–270). Palo Alto, CA: Consulting Psychologists Press. Rouiller, J. Z., & Goldstein, I. L. (1993). The relationship between organizational transfer climate and positive transfer of training. Human Resource Development Quarterly, 4(4), 377–390. Ryan, A. M., McFarland, L., Baron, H., & Page, R. (1999). An international look at selection practices: National and culture as explanations for variability in practices. Personnel Psychology, 52(2), 359–391. Ryman, D. H., & Biersner, R. J. (1975). Attitudes perspective of diving training success. Personnel Psychology, 28, 181–188. Sarkar-Barney, S. T. (2001). Extending a transfer of training framework to include the role of national culture. Unpublished doctoral dissertation. Bowling Green State University, Ohio. Schwartz, S. H. (1992). Universals in the structure and content of values: Theoretical advances and empirical tests in 20 countries. In: M. P. Zanna (Ed.), Awareness in Experimental Social Psychology (Vol. 25, pp. 1–65). Orlando, FL: Academic. Schwartz, S. (1994). Beyond individualism/collectivism: New cultural dimensions of values. In: U. Kim, H. C. Traindis, S.-C. Choi & G. Yoon (Eds), Individualism and Collectivism: Theory, Method, and Application (pp. 85–119). Thousand Oaks, CA: Sage. Schwartz, S. H. (1999). A theory of cultural values and some implications for work. Applied Psychology: An International Review, 48(1), 23–47. Schwartz, S. H., & Bilsky, W. (1987). Towards a psychological structure of human values. Journal of Personality and Social Psychology, 53, 550–562. Schwartz, S. H., & Bilsky, W. (1990). Toward a theory of the universal structure and content of values: Extensions and cross-cultural replication. Journal of Personality and Social Psychology, 58, 878–891. Schwartz, S., & Huismans, S. (1995). Value priorities and religiosity in four western religions. Social Psychology Quarterly, 58, 88–107. Smith, P. B., & Schwartz, S. (1997). Values. In: J. B. Berry, M. H. Segall & C. Kagitcibasi (Eds), Handbook of Cross-Cultural Psychology (Vol. 3, pp. 77–118). Boston, MA: Allyn & Bacon. Stevens, C. K., & Gist, M. E. (1997). Effects of self-efficacy and goal-orientation training on negotiation skill maintenance: What are the mechanisms? Personnel Psychology, 50, 955–978. Strohschneider, S., & Guss, D. (1999). The fate of the Moros: A cross-cultural exploration of strategies in complex and dynamic decision making. International Journal of Psychology, 34(4), 235–252. Tayeb, M. (1988). Organizations and national culture. London: Sage. The Role of National Culture in Enhancing Training Effectiveness 213

Teachout, M. S., Sego, D. J., & Ford, J. K. (1997). An integrated approach to summative evaluation for facilitating training course improvement. Training Research Journal, 3, 169–184. Tesluk, P. E., Farr, L. J., Mathieu, J. E., & Vance, R. J. (1995). Generalization of employee involvement training to the job setting: Individual and situational effects. Personnel Psychology, 48, 607–632. The Economist (1994). July 30, p. 57. Thorndike, E. L., & Woodworth, R. S. (1901). The influence of improvement in one mental function upon the efficiency of other functions. Psychological Review, 8, 247–261. Tracey, B. J., Tannenbaum, S. I., & Kavanagh, M. J. (1995). Applying trained skills on the job: The importance of the work environment. Journal of Applied Psychology, 80(2), 239–252. Triandis, H. C. (1994). Cross-cultural Industrial and Organizational Psychology. In: H. C. Triandis, M. D. Dunnette & L. M. Hough (Eds), Handbook of Industrial and Organizational Psychology (2nd ed.). Palo Alto, CA: Consulting Psychologists Press. Triandis, H. C., Leung, K., Villareal, M., & Clack, F. L. (1985). Allocentric vs. idiocentric tendencies: Convergent and discriminant validation. Journal of Research in Personality, 19, 395–415. Tubiana, J. H., & Ben-Shakhar, G. (1982). An objective group questionnaire as a substitute for a personal interview in the prediction of success in military training in Israel. Personnel Psychology, 35, 349–357. Vandenput, M. A. E. (1973). The transfer of training: Some organizational variables. Journal of European Training, 2(3), 251–262. Van De Vijver, F. (1997). Meta-analysis of cross-cultural comparison of cognitive test performance. Journal of Cross-Cultural Psychology, 28(6), 678–709. Weech, W. A. (2001). Training across cultures: What to expect. Training and Development, 55(1), 62–64. Wilkins, S. (2001). International briefing 9 training and development in the United Arab Emirates. International Journal of Training and Development, 5(2), 153–165. Williams, R. M., Jr. (1968). Values. In: E. Sills (Ed.), International Encyclopedia for the Social Sciences (pp. 283–287). New York: MacMillian. Williams, R. M., Jr. (1970). American society: A sociological interpretation (3rd ed.). New York:Knopf. Xie, J., & Wu, G. (2001). International briefing 10 training and development in the People’s Republic of China. International Journal of Training and Development, 5(3), 165–223. This Page Intentionally Left Blank 7. TOWARDS AN ANTHROPO- TECHNOLOGY. X. A NEW ACTIVITY FOR THE UNITED NATIONS IN THE SERVICE OF ECONOMIC DEVELOPMENT: SPECIFYING REQUIREMENTS FOR TECHNOLOGY TRANSFERS IN GIVEN GEOGRAPHICAL AND ANTHROPOLOGICAL LOCATIONS

[Translated from the French by Kendra Francisco. Southwestern University, Texas]

Alain Wisner

Laboratoire de Physiologie du Travail et d’Ergonomie Conservatoire National des Arts et Metiers, Paris, France

Cultural Ergonomics Advances in Human Performance and Cognitive Engineering Research, Volume 4, 215–221 Copyright © 2004 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1479-3601/doi:10.1016/S1479-3601(03)04007-4 215 216 ALAIN WISNER

EDITOR’S NOTE

With the recent tragic death of Monsieur le Professeur Alain Wisner, the peoples of industrially developing countries lost an ardent champion and the field of ergonomics lost a much loved friend, colleague, teacher, and researcher. Gentle, brilliant, caring and incisive in his thinking and analyses, he was the father of anthropotechnology, whose hallmark is its focus on comparative anthropological factors as they relate to work in industrial settings. He honed his thinking in this area while traveling and solving problems in many countries for Renault and other firms, and teaching students and conducting research at the Conservatoire des Arts et Metiers in Paris. In addition to many other scholarly writings, he brought together at the Conservatoire a collection of shorter pieces, working papers, and detailed memoranda under the theme Vers une Anthropotechnologie. One of these is this paper, which looks to the United Nations for international assistance to resolve issues affecting transfer of technology.

ABSTRACT

Seeking to equip industrially developing countries with functioning factories and machines, the author first used the concepts, knowledge, and methods of ergonomics. He then discovered that it is also necessary to develop, test, and use other concepts, other knowledge, and other methods, especially by bringing into consideration the human sciences (geography, demography, epidemiology, sociology, economics, linguistics, anthropology, history) for orienting such countries to means of actualizing technical devices. Reflecting this thinking, the present note proposes an anthropotechnological approach to improve international technology transfer.

It is fair to say that nowadays, anthropotechnology enriches and may be regarded as an integral part of thinking in cultural ergonomics. (1) For a number of reasons, of which the main point is demographic expansion, it is hardly possible to increase or even maintain the living standards for the inhabitants of numerous countries without industrialization or semi-industrialization of primary and secondary sectors. (2) The majority of countries on the path of industrialization do not possess sufficient human resources in order to assure the occurrence of activities necessary for this technical change. Hence, they call upon the transfer of technology. Towards an Anthropotechnology 217

(3) Multinational businesses have obtained brilliant economic and financial results in transferring their technology to diverse parts of the world. In general, this transfer brings not only the specified technology (hard) but also the method of serving it (soft), e.g. the organization of work and of the company, the particularities of recruitment and personnel training, and social activities related to the business (housing, health, school, etc.). Often, certain groups, technicians or even workers, coming from the parent company, continue to practice their job activities in the branches for years. In this manner, the local businesses in the countries where they are positioned, often appear as anthropotechnological islands, where we find, according to varying proportions, the advantages and inconveniences of the original ventures. These acts, along with many other politico-economical elements, are the source of ambiguous reactions or recorded negatives, often about businesses whose functioning sometimes appears satisfactory at first glance. Concern- ing the project proposed in this note, these considerations must, at least, have the effect of demonstrating the inconveniences of industrial transfer when it is pure and simple duplication of a technical and social stratagem in a foreign country. (4) The national companies of a country in economic development naturally find themselves in situations of extreme diversity. We will interest ourselves here exclusively in those who use, in a substantial way, an imported technology, and who benefit from loans coming from agencies of the United Nations, e.g. the World Bank, United Nations Development Program, International Organization of Work, etc. In the remainder of this text, we will call such companies Enterprises Benefiting from Technology Transfers Aided by the United Nations. (5) Whether one should challenge the use of technology transfer is a question of an activity as ancient as humanity. We can note, in this connection, that transfer was itself produced for a long time in the sense of East-West (towards Europe from India and China) and South-North (coming from the Islamic world) before its inversion in most recent centuries. This transfer took on an enormous expansion and has largely demonstrated its indispensable nature. Nonetheless, in numerous cases, the results do not arrive at the anticipated level. Several possibilities arise: the quality and quantity of products are such that the imported business can function only under low-satisfying financial conditions; the business closes after functioning for a relatively short time; or even that eventually, negative effects take over (for example, the considerable development of human and animal parasites in irrigated zones following the construction of a dam). 218 ALAIN WISNER

(6) Facing such failures, during the 1970s, we have applied solutions of the techno-economical type: servicing machinery after its sale; sale of coherent technological ensembles, sale of factories “everything included” (where everything is conceived within the details of the selling country from the machines to the buildings); sale of factories “product included” (where we present not only the “hard” – technology and buildings – but also the “soft” – organization of production and the policies regarding personnel). We were even up to the point of evoking the “market included” factory that can, in certain cases, lead to the creation of a company possessing a location neighboring that of a multinational branch. (7) It seems, following actual success limited by techno-economical solutions, that it is necessary to envision the problem of transferring technology in a slightly different way, and to admit that there is no successful transfer of technology if it is passive. However, for an active transfer of technology, it is necessary that there be an implementation of important national intellectual resources from the buying country, something that the latter is not always equipped to do. This is either because its own resources are too modest or because the paths linking the intellectual sphere and the economic and technical spheres are not sufficiently organized. (8) To clarify the envisioned course of action, it is not necessary to assess the “damage caused by progress” (that is to say, general effects of industrializing ancient civilizations, sociological perturbations linked to the increase of population and to very rapid urbanization). It is not even a question of showing the increased hazards for certain aspects of health (as with the parasites cited above). All of these aspects are of extreme importance, but are not, as much as others, the object of the proposed action. Nonetheless, their existence cannot be neglected, as they are eventual explanatory elements of techno-economical failures. By the same token, we will not directly evoke socio-political phenomena that can lead to a brutal rejection of certain forms of industrialization. (9) The precise purpose of the envisioned action is to understand the complex causes of the failure of a certain number of operations dealing with the transfer of technology and to formulate specific recommendations for each region or country and for each type of project. It is not necessary to stand by the decisions made by governments and companies for the choice of product types to be obtained or methods of production to be installed, but to specify the conditions in which such a specific project can succeed. An analogy exists between ergonomics (human factors in the United States), that which focuses on the adequacy of the machine for man, and the action of beginning the venture on an anthropotechnological basis, and claiming to improve Towards an Anthropotechnology 219

the adequacy of a device produced for the local population considering the geographical and cultural context. (10) In order to treat the anthropotechnological problems posed by a specific case of transfer of technology, it is necessary to proceed from relevant method- ology and categories of information. On the other hand, it is necessary to utilize knowledge pertaining to other cases of favorable and unfavorable transfers. It is not only a question of bibliographic documentation, but also of organized professional experience, since the purpose of this action is not only knowledge but also actualization. It is necessary, however, to make use of knowledge relative to the local situation. One of a number of difficulties in industrial development is insufficient and especially low operational knowl- edge of what each developing country imposes upon itself. It would be very interesting to favor in each country the composition of a center providing the geographic, demographic, sociological, and anthropological conditions, as well as technical and economic, that is to say, the entire ensemble of conditions necessary for economic development and, more particularly, for the success of technology transfer. One purpose of such establishment would be to generate contact between centers and study and research teams, who otherwise have little or none. It would also permit industry and research to better familiarize themselves with each other and to better collaborate in economic development. (11) It is evident that one such project can make the object of one of many actions of the International Program for Ameliorating Working Conditions an effort of the International Labor Office. Actually, this project is initially only a per- sonal unsolicited proposal by an expert from the International Program for Ameliorating Working Conditions and is based upon the experience acquired by this expert at the heart of the International Program for Ameliorating Work- ing Conditions. However, the specifically economic importance of the project would also give him his seat in the program adopted by the International La- bor Office, since the ventures created by transfer of technology that function poorly, or not at all, do not bring to the importing country neither the jobs that we would hope to create there nor the benefits necessary for establish- ing other ventures that create jobs. On the other hand, we know how much the budget of the International Labor Office and the International Program for Ameliorating Working Conditions in particular are limited in connection with the needs to which it is necessary to respond. That is the reason why such a project would be able to make its objective an effective collaboration with the World Bank and the United Nations Development Program and for that which the author suggests initially be applied to the case of the Enterprises Benefiting from Technology Transfers Aided by the United Nations. 220 ALAIN WISNER

(12) Some substantial sums are raised by the World Bank and the United Nations Development Program and then distributed, after careful studies of various projects pertaining to agricultural and industrial development, in the most diverse range of countries that have in common an urgent need to create wealth and jobs. These projects make the most frequent use of a transfer of technology. As mentioned previously, the preliminary studies are essentially of a technical, economic, and political nature, which is in the same way indispensable. A large majority of projects financed in this way arrive at satisfying results for the benefiting countries and translate as a reimbursement to the loaning bodies. The latter can reuse the reimbursed funds to finance new projects. However, we know that a certain number of projects do not arrive at the anticipated results and in such cases, the countries and companies benefiting from loans are not capable of reimbursement. (13) It is reasonable to think that in a number of substantial cases, the Enterprises Benefiting from Technology Transfers Aided by the United Nations would arrive at better technical, economic, and financial results, if study and recom- mendations of an anthropotechnolgical nature had accompanied the routine technical, economic, and political analyses, namely, an assessment of the following: the low rate of engaging the technical device, therefore the modest production volume; the unequal quality of products – or their mediocre reliability; the increased frequency of accidental destruction of materials as well as accidents affecting the transferred technical arrangements; and local characteristics from a geographical and an anthropological perspective. (14) The World Bank and the United Nations Development Program would be able to test the validity of such considerations by calling on specialists from various other agencies that the United Nations possesses: first, the International Labor Office (that has long and serious experience in this domain, revived recently in the heart of the International Program for Ameliorating Working Conditions), but also the Food and Agriculture Organization, the United Nations International Children’s Emergency Fund, the United Nations Educational, Scientific, and Cultural Organization, the World Health Organization, etc. ... (15) A preliminary study of feasibility could be accomplished under the direction of a limited group of experts from the World Bank, the United Nations Development Program, and the International Labor Office, working in two directions under a time limit of three years. In one respect, we could study a posteriori some Enterprises Benefiting from Technology Transfers Aided by the United Nations situated in a varied range of countries having benefited from analogous transfers of technology and having known varied Towards an Anthropotechnology 221

successes. On the other hand, we could join anthropotechnological experts and technical, economic, and political experts who would undertake the preliminary study and also assure the follow-up and evaluation of some projects actually envisioned by Enterprises Benefiting from Technology Transfers Aided by the United Nations. (16) A much more extensive project could be put into effect in the case where the preliminary study of feasibility would appear to have been conclusive. This project, on the one hand, could consist of (i) a World Center elaborating one or several methodologies and collecting the studies from the differing cases under the Enterprises Benefiting from Technology Transfers Aided by the United Nations, or, on the other, (ii) aid to several centers that are briefly described in point eleven. This Page Intentionally Left Blank 8. ERGONOMICS IN INDUSTRIALLY DEVELOPING COUNTRIES (IDCs): SOCIO-CULTURAL PERSPECTIVES

P. A. Scott and J. Charteris

ABSTRACT

This chapter asserts that, notwithstanding the forces of globalization, the gap between Industrially Advanced Country (IAC) “haves” and Industrially Developing Country (IDC) “have nots” is increasing. Poverty, deprivation, over-population, illiteracy, and sub-optimal working conditions attend this scenario. This widening chasm, among other things, feeds discontent which, justified or not, exacerbates tensions. To the extent that this view is valid, the challenge for ergonomics is to contribute to a narrowing of the gap by recognizing that the vast majority of the biomass of humanity is engaged in basic IDC, not advanced technical IAC issues. Their needs are best met by culture-compatible, participatory, small-scale, unsophisticated, low-cost improvements that can be easily assimilated. Strategies for assisting in redressing the imbalance between IDCs and IACs are discussed.

...social, scientific and technological changes have enhanced the multi-disciplinary character of ergonomics, and widened the scope of the work environment concept to the extent that the globe must now be considered to be a subject of ergonomics interest. (Kaplan, 1995, emphasis ours)

Cultural Ergonomics Advances in Human Performance and Cognitive Engineering Research, Volume 4, 223–248 © 2004 Published by Elsevier Ltd. ISSN: 1479-3601/doi:10.1016/S1479-3601(03)04008-6 223 224 P. A. SCOTT AND J. CHARTERIS

PROLEGOMENON

The “First World” – “Third World” dichotomy has, apparently, never sandwiched a “Second World” interface. When French demographer Alfred Sauvy conjured his “Third World” reminiscence of the 18th Century Revolutionary “Third Estate” of “have-nots,” his construct immediately suggested a “First World” (cf. “First Estate”) of “haves,” and a dichotomy was born, which in diverse guises has since been a central theme in international relations. Whether political correctness or the axe one is grinding at the time favours or rebuffs the term “Third World”; whether it suits one to argue against the utility of a Third World concept or, conversely, to see it as a useful term, synonymous with “deprivation dulling aspiration,” it is important for present purposes to accept the terms, “Third World Countries” and “Industrially Developing Countries (IDCs)” as synonymous. Like the Third Estate of French Revolutionary commoners, the IDCs are prevented, by diverse circumstances, from any hope of success in achiev- ing living conditions which the First Estate (cf. Industrially Advanced Countries, or IACs) take for granted and acquire more efficiently in more conducive environments. It should be self-evident that the widening disparity between IACs and IDCs (“haves” and “have nots”) should be stemmed. Ergonomics can (indeed has a responsibility to) play a major role in unifying a drive to enhance the quality of life, worldwide. There is an ominous foreboding in all of this, which discerning readers will already have perceived: when people long denied see disparity paraded before them, subliminal discontents are prone to manifest; the greater the disparity, the more bizarre and chilling the response.

INTRODUCTION: WORLDS APART

The theme of this volume, how cultural considerations relate to ergonomics issues, presupposes that there is agreement as to what an IDC, as distinct from an IAC, is. No such universal consensus in fact exists, partly because the nations involved are themselves in heterogeneous stages of development (Wisner, 1985), and partly be- cause the answer given will depend on the perspectives of the respondent. Figure 1 may serve to make the latter point more economically. Map (A) depicts the “Third World” of IDCs as an ergonomist from one of the world’s most affluent, technologically advanced nations might see it. Map (B) depicts the Third World as an ergonomist from the Horn of Africa might see it. Clearly, our backgrounds colour our perceptions, and no individual’s perception is in this sense more valid than any other’s. As Borg (1970) succinctly put it, Ergonomics in Industrially Developing Countries 225

Fig. 1. Socio-Cultural Relativity and Perception. Note: The IDC/IAC distinction is usually not clear-cut, except in the case of extremes, because whichever of a variety of criteria are employed to distinguish them, all are part of a continuum, with many straddling the divide inconclusively: (A) How an ergonomist from the USA, Sweden, or Australia might perceive the domain of IDCs (light shading). (B) How an ergonomist from Eritrea, Haiti or Bangladesh might perceive the domain of the IACs (dark shading). Those who would be critical of dark shading of regions they feel are unquestionably IDCs, are perhaps revealing an inability to see the world through, say, Eritrean eyes, and may have forgotten that, to a man with no feet, barefootedness is indeed luxury. 226 P. A. SCOTT AND J. CHARTERIS

“Man reacts to the world as he perceives it, not as it really is.” Therefore, we must somehow draw a useful dichotomy, for how else can ergonomists of Texas begin to understand, let alone empathise with, ergonomics issues in Eritrea? When perceptions that are grounded in a First World culture are extrapolated to inform views relative to a Third World milieu using cues that are fundamentally inappropriate, the result is, often justifiably, labelled as arrogance, and often exacerbates the problem rather than improving the situation. There is consensus on the broad characteristics shared by IDCs, viz low standards of living, low productivity, high rates of population growth, high unem- ployment, significant dependence on agriculture, and vulnerability in international relations. Clearly IDCs are countries which have dependent economies based on the cheap production of primary products destined for beneficiation by IACs and available in that form only at usurious rates to IDCs. Not surprisingly, IDCs are countries in debt, unable to get out of debt and plagued by endemic problems which their poverty keeps them from being able to solve. As the gap widens, the alienation and seething festers in a nurturance of ignorance, poverty, population explosion and disease, and is exacerbated by an increasingly accessible mass- media which sensationalises the disparity. Almost no-one on earth today does not know how the “haves” live: ostensibly oblivious “haves,” whose very affluence is an affront when one’s day-to-day existence is a nutritional roll of the dice. Apparently IDCs are constituted, or dissolved, in the minds of those considering them. A Somali ergonomist, not surprisingly, might well look incredulously at journal articles on macroergonomics: what possible relevance can corporate sick-system management styles have for an ergonomist preoccupied with the best way to harness two asses to a cart, or to manually draw water out of a deep well in a remote wadi? Why would a Swazi ergonomist read an article on cognitive ergonomics when his focus begins, and ends, with the problems of female head-load carriers who must carry bundles of firewood 1.7 times their own body-length, and 60 to 100% of their body mass, in 35 ◦C temperatures, barefooted, on rough terrain over 10 km distances? (See Fig. 2.) Is ergonomics just for the technophilic, computer dominated, sedentary-living masses in the less bio-prolific nations where the poorest are rich by comparison? A San Francisco bag-lady, pushing a cart and raiding the left-overs of average- to-affluent San Franciscans is shod, clothed, can eat bread, drink clean water, and even afford an occasional cheap bottle as solace. This is a far cry from having to pick up individual rice grains or corn kernels from the dirt to make a daily fistful to sustain life. The picture painted, while stark, covers the lot of untold millions. Academically distasteful it may be; emotional it unabashedly is. But it is a picture that needs to be painted, as a precursor to ensuring that the reader is sensitised to the disparate ergonomic needs of the planet: needs that many are oblivious to, or which Ergonomics in Industrially Developing Countries 227

Fig. 2. African Women Headloading Prodigious Weights: (A) in the Field; (B) Under Controlled Conditions in the Laboratory. are being dismissively swept aside by the allures of IAC orientations; ergonomics in space, ergonomics in a post-robotic world; ergonomics focusing primarily on a technology dominated world; ergonomics beyond a “macro” approach.

CRYSTALLISING THE CONCEPT OF AN IDC

Have we overstated the case that ergonomists are, by and large, focused more at the technophilic cutting edge than on the plight of the bulk of humanity at work? We think not, and the luminaries in the field of IDC ergonomics cited in this chapter have all said much the same in seminal works to a large extent paraphrased here. Wisner, for example, hinted in 1985 that First World ergonomists are largely unaware of the work being done in IDCs, much of which is original, closely linked to social demand, and bearing important theoretical implications. Shahnavaz, in his 1995 lecture to the Ergonomics Society, was clear that when, occasionally, IAC ergonomists have turned attention to IDCs, it is issues of technology transfer, rather than the more mundane (but for the people concerned, more meaningful) problems of excessive manual labour under heat-stressed conditions, that have tended to capture their interest. He notes that even when IDC issues began to surface in IAC circles in the mid sixties “interest in and attention to the subject was low among professional ergonomists in the USA” and virtually exhausted by 228 P. A. SCOTT AND J. CHARTERIS

1967. His breakdown of IEA World Congress publications, spanning more than a decade, showed that 93% of Congress papers emanated from the IACs and few were devoted to the burning IDC issues of inappropriate technology transfer, outmoded technology and indigenous practise (Shahnavaz, 1996). It is apparently a fact that, by and large, the interests of many established international ergonomists would be no different if IDCs did not exist. A survey of major ergonomics journals reveals that in this millennium, of 300 articles consulted only 5 were related to research in IDCs, and at the IEA Triennial Congress held in San Diego (2000), of nearly 2,000 papers presented the same percentage, approximately 2%, focused on work conducted in IDCs. It is a sad irony that in IDCs, where ergonomics is the most sorely needed, the need for it is least appreciated (see Table 1). Scott (1999a) has argued that ergonomics awareness in IDCs is inversely proportional to perceived necessity. Scott proposed that the key to successful implementation of ergonomics in IDCs was to eschew westernized approaches and search for innovative indigenous solutions via encouragement of local initiatives and commitment to improving working conditions. The objective should be to meet the needs of an indigenous work force from within its own working ambience. Indeed, many internationally recognised ergonomists have felt that they have failed in their duty to establish ergonomics in IDCs (Hendrick, 1998).

Table 1. Demographics of Affluence Relative to Organisational Structure of Ergonomics.

World Bank Categories of the World Federated Societies of IEA Classification Bank (2002)a (2002)b Number of Percentage of Number of Percentage of % of 41 States 208 States States 208 States in IEA

Low-income economies 66 32 2 <1.0 4.9 Lower-middle income 52 25 8 3.8 19.5 economies Sub totals 118 57 10 4.8 24.4 Upper-middle income 38 18 7 3.4 17.1 economies High-income economies 52 25 24 11.5 58.5 Sub totals 90 43 31 14.9 75.6 Totals 208 100 41 19.7 100 aWorld Bank Group Data and Statistics (2002). bRoster of the IEA (2002). Ergonomics in Industrially Developing Countries 229

Shahnavaz has characterised ours as a “humane profession.” He dreams of a globalised ergonomic endeavour (Shahnavaz, 1996), while admitting that this remains a figment of the imagination as long as the majority of the world’s ergonomists are engaged in further ameliorating the conditions of the world’s least abused, best organised, most affluent workers. His vision is for a narrowing, by whatever means available, of the gap between the two worlds. It is arguable that many IAC gap-closing advances made on one front are balanced by reversals on another, so that any hope of IDC ergonomics advance- ment seems to be left to the struggling initiatives taken by the IDCs themselves, which we contend, is not altogether a bad thing. Yet, facilitatory inputs from the IACs, which can afford them, would do much to accelerate a closing up of the present gap, a gap dramatically illustrated in Table 1. The IEA itself, noble as its aspiration is to globalise ergonomics, reflects the two-world split. Table 1 shows that 41 (19.7%) of the 208 countries listed in the World Bank’s Classification of Economies are represented by ergonomics societies Federated in the IEA. Over 75% of IEA Federated Member Countries enjoy high- or upper-middle income economies. Yet the 24% of IEA member countries with low- and lower-middle income economies (10 countries) account for over 45% of the world’s population! Also evident from Table 1 is that almost 92% of the world’s poorest economies (accounting for over 1.3 billion people) are not countries with membership on the IEA. Against this, barely 21 million people, mostly from small wealthy nations (like Liechtenstein, Luxembourg and Brunei) are citizens of IACs which are not affiliated with the IEA, while 840 million in the world’s 52 highest income economies are citizens of countries that enjoy Federated Member status in the IEA. The IEA, in turn, doubtless would welcome with open arms applications from Uganda, Tajikistan and Benin for example, but by and large, it presently reflects First World orientations and its major publications concentrate on issues that, to all intents and purposes, are “extra-terrestrial” for the over 3 billion people who toil, when they can, for less in a year than an average worker from a high-income economy country makes in a week. The picture, in our view, cannot be more eloquently painted than in the fol- lowing excerpt from P. K. Nag’s opening Presidential remarks at the International Congress in Humanizing Work and the Work Environment (Mumbai, 2001):

One of the notable features in the world we live in today is the huge difference between what is known and what is put into practice. There is a difference between the islands of excellence and the oceans of deprivation; the difference between the use of leading edge technologies and conventional ones; the difference between organized sectors of the economy that have achieved high productivity and quality, and unorganised sectors that are striving for emancipation from their economic backwardness. The difference is evident between the visible workforce of the 230 P. A. SCOTT AND J. CHARTERIS

formal economy and the invisible vulnerable workforce, including women and children in the informal sectors (Nag, 2001, emphases ours).

Shades of Grey

A caution needs to be sounded at this point. The dichotomy between IDCs and IACs can be stretched to a far-fetched extent. Geographic location does not itself constitute grounds for labelling a country an IDC. Third-World countries may choose to squander meagre resources on ultra-modern nuclear-weapons capabil- ities, for instance. Not inconceivably, the developed infrastructure for this may rival that of the most affluent and bellicose of the IACs. Conversely IACs may, for reasons of lop-sided prioritising, experience IDC conditions in some of their own isolated rural areas, or in respect of poorly-tolerated minority groups, sub-cultures within the broader society which, one way or another, may be denied access to resources. Whole countries, moreover, may straddle the divide, South Africa being an excellent example. Infra-structurally in many ways an IAC exemplar, South Africa none-the-less shares with typical IDCs a burgeoning population, large-scale poor literacy, poverty, an exploited economy based more on raw material provision than on beneficiation, with the effect of the high incidence of HIV-AIDS becoming more apparent in the workplace on a daily basis. One can expect, and will surely find in South Africa, every possible challenge for ergonomists of all persuasions. The corporate “hi-tech” mining and manufacturing sector sets challenges to the most modern exponents of the newer ergonomics fields of macro-ergonomics and cognitive ergonomics, while rural and urban poor environments are seed-beds for the most traditional ergonomics; the sort of designed interventions that might have been implemented when Jastrobowski coined the term ergonomics in nineteenth century Poland. The reality is that when governments are preoccupied with reticulation, elec- trification, housing, disease eradication, staple food provision, basic education and job-provision for the masses, they pay scant attention to the niceties of work-practice optimisation. Few ergonomists in the IACs would be so naive as to attribute their national affluence and industrial productivity solely to ergonomics, for correlation is not causation. But the onus would be on the gainsayers of ergonomics to explain why the correlations are strong and positive in the IACs, and so strongly negative in IDCs. Probably ergonomics is part of a more general mobilization ethos in countries which have gone through, or never had, the problems faced by IDCs. And probably the poor development or absence of Ergonomics in Industrially Developing Countries 231 ergonomics is part of a more general malaise in IDCs, which labour under, and cannot get free of, the yoke of tired apathy.

A MODEL FOR CULTURAL ERGONOMICS

However briefly made, a declaratory statement is necessary at this point, precisely because if, as proposed, the First and Third worlds are indeed worlds apart, then the likelihood exists that American or Western European ergonomists will be absorbed with issues which a Sudanese or South-East Asian ergonomist might never encounter in a lifetime of service. Our aim here is succinctly to propose a conceptual model which would be immediately recognisable to luddites and technophiles alike, and which goes to the core of the domain, no matter what level of technological sophistication is involved. Figure 3, then, all but speaks for itself. Despite burgeoning new areas of specialisation within the field, and culture- based differences of emphasis, a reasonably secure consensual position can be

Fig. 3. A Conceptual Model for Ergonomics. Note: The sum total of purposeful human endeavour is a cultural milieu of created things, our attitudes towards them and the effect of this endeavour on altered perceptions. Ergonomics aims at optimisation of this process by designing-out incompatible situations and designing-in humane methods of achieving the purpose of work. 232 P. A. SCOTT AND J. CHARTERIS taken that ergonomics aims to ameliorate human working conditions by taking cognisance of the ways in which humans acquire, process and use information as they act, under imposed workplace and natural environmental conditions, to acquire or produce materials perceived to enhance their lives. What our simple model lacks in specificity, it makes up for in universitality: the macro-ergonomist engaged in improving the operation of a systemically inefficient European Corporation, and the micro-ergonomist facilitating selection between two designs of pick handle for use by female subsistence farmers in Malawi, are in fact doing the same thing, optimising work practices set in specific contexts only some of which can be modified. Differences between these extremes are of degree, not of kind. Assuming broad consensus in respect of this simple model, its utility for present purposes lies in the fact that it accommodates the different emphases that ergonomists from IDCs and IACs would place: differences between them would centre, not on disparity of concept, but rather on differences in technological sophistication and perhaps, system complexity. How might the process modelled in Fig. 3 be influenced in an IDC setting? Consider the following: At [A]: Cultural perceptions of the endeavour itself are invoked. Information is processed in respect of the task, but interpreted in the light of a cultural milieu that might be antithetical to a perceived materialistic product-output orientation.

At [B1]: Consider a European factory, transplanted in toto into an equatorial setting. Are designed-in work spaces, which may suit morphological ranges of variability of largely Caucasian workers equitable for largely tropical African, Asian or South American workers, and if not, are indigenous ergonomic resources available to optimise mismatches?

At [B2]: Will machinery manufactured in the (largely temperate) IACs function optimally in the (largely tropical) IDCs; and if not, does the recipient country have the urgency ethic, the technical know-how, the financial resource base, and the spare-parts infrastructure to effect efficient repairs timeously? At [C]: The artifacts, attitudes and information generated in this process may exacerbate cultural tensions. For example, when workers (particularly females) are perceived to be “westernising,” is this seen as a threat to the mores of the local culture as enunciated by (particularly male) authoritarian traditional leaders? To the above work-process influences we need to add those of the society as a whole. Technology transfer cannot succeed: if governmental policy is not actively facilitative; if the need for such transfer is not driven from within the recipient IDC itself; if infra-structural (not least, financial) resources are not available to ensure Ergonomics in Industrially Developing Countries 233 success; or if the technology recipients themselves cannot perceive, and forestall, all likely problems associated with attitudes to work, to shift-work, to women at work, to health and safety, and the need for training. In short, if there is poor absorptive capacity, the transfer of technology might as well not occur, for the probability of failure is all but guaranteed. Endemic culture lies at the heart of all the points raised above, and in the long run, it is as obvious that the technology must be fitted to the culture, as it is that the task must be fitted to the human doing it, not the reverse.

Technology Transfer Issues

“...technology is not a good traveler unless it is culturally calibrated” (Maurino, 1993). Without care and consideration of the indigenous workforce, the im- portation of sophisticated technology is more likely to be perceived by local workers as a threat rather than a help. The phenomenal success of “anthro- potechnological islands of excellence” (Wisner, 1985) in some IDCs shows that managed technology transfer tends to obliterate notions of fundamental cognitive differences between peoples. Where people from one culture are introduced, in terms meaningful to them, to foreign technologies, and then properly trained and supported, their performance is on a par with that of people from the cultures that have spawned those technologies. These “islands of excellence” abound in IDCs when the multinational companies involved create the appropriate ambience to maintain the same product quality, the same level of efficiency and the same standards of safety wherever they are located. The claim is that these success stories are due to requisite infrastructure provision and successful eradication of cultural barriers to standardisation of training and attitude enhancement. If this is so, then the sources of low productivity, of defective infrastructure and of inadequate training and incentive provision are probable reasons for such phenomena as the cavalier attitudes encountered among flight and ground staff in the aircraft industry in some IDCs (Kaplan, 1995). People living in squalid circumstances, in which existence itself is based on a resignation that nothing works as it should, can hardly be expected to bring to the workplace an ethic of meticulousness in respect of concern for a rusty valve or a corroded pipe. If it goes unnoticed, it is not reported, and when it is, the records are kept poorly, in a language not the mother tongue of the record keeper. In retrospect, under the circumstances of its installation and lack of oversight and maintenance by the multinational concerned, Bhopal must be seen as an accident whose inevitability was all but guaranteed at the outset. 234 P. A. SCOTT AND J. CHARTERIS

Against such tragedies, the everyday technology-transfer issues routinely faced by IDC ergonomists, though they may daily cost lives, seem small. These include, for example, a complete disregard for chain-saw guards, which are seen as adding unnecessary weight and summarily discarded, and the relative absence in IDCs of safety barriers and/or alarms in front of metal presses or material cutters. Industrial monitoring agencies are themselves inefficient, often open to corruption, usually overstretched in terms of well-trained personnel, so that practices which would seldom escape notice in IACs go on as the norm in IDCs. The authors observed, in one of the world’s largest open-cast mines, an instance in which, despite intensive operator training in the use of massive ore-carriers, an accident costing tens of millions of rand was nearly caused because it had simply, and incorrectly, been assumed that operators would know how to adjust their own cabin drive-seats to suit their anthropometric needs. What are the special needs of IDCs in the present context? To consultants with widespread international experience, poverty underlies all other considerations. Poverty explains why it is that only small, cheap interventions which manifest immediate beneficial results, garner any interest in IDCs. It explains why, while a British Columbian sees work and family as quite separate entities, an Eritrean blurs all distinction between personal survival, family security and work. It explains why Governments characteristically fail to commit more than good intentions in circumstances where poor roads, defective education systems, collapsing economies, disease pandemics and disaster relief efforts sap their energies and meager resources. Small wonder then that standards of workplace safety, culled from IAC sources, abound, but are not adhered to at anything like the level achieved in IACs. Small wonder that industrial accident rates are over 10 times higher in IDCs than in IACs (Shahnavaz, 1996).

IDC ISSUES EXEMPLIFIED

O’Neill has succinctly encapsulated the IDC milieu as one in which 71% of the labour force is engaged in (largely subsistence) agriculture generating a per capita GNP around US$270, or less than 0.75¢ a day. In IACs, on the other hand, agricultural workers constitute about 7% of the labour force and generate a per capita GNP almost 50 times that of IDC counterparts (O’Neill, 2000). Almost two decades ago, Wisner pointed to heated debate as to whether rural development or rapid industrialisation was the panacea to IDC woes, or whether the latter would merely exacerbate the existing dire poverty of the rural masses because industrial projects tend to be financed from capital resources generated by agricultural activity. Ergonomics in Industrially Developing Countries 235

So we can expect that some will emphasise rural upliftment to eradicate poverty, while others will advocate rapid industrialisation by every possible means, and will cite isolated success stories in support. No strategy, however, will work unless whatever is done involves assimilation of the proposed strategy into the cultural milieu of the target society. This cannot occur when IAC ergonomists propose interventions that work in IAC cultural contexts, as solutions to problems rooted in IDC cultural contexts. This is not to say that only ergonomists from a particular cultural background are competent to solve ergonomics problems in that culture. The qualification is perceptiveness, not birthright. A few examples may suffice to illustrate the point that facile First World solutions to what might appear to an IAC ergonomist almost a “non-problem,” really are no solutions at all when the cultural context is disregarded. Recently Pradhan and Thakur (2002) have turned attention to the plight of employees in the public transport sector in Kolkata, a city of over five million people. Among problems faced was the fact that bus conductors complained of excessive back and leg discomfort caused by working unduly long shifts in standing positions and without rest breaks. To a culturally imperceptive ergonomist, the problem solves itself by provision of a conductor’s seat, from which, conceivably, ticket transactions could be effected. But, as anyone who has travelled on an Indian train or urban bus knows, a conductor’s seat in (or outside) the bus would invite usurpation by two, possibly three would-be passengers. The realities of socio-cultural context have to be factored into any potential solution to the incidence of back and leg fatigue of Indian bus conductors. A Nigerian anthropometric survey (Okunribido & Olajure, 1999) of hand dimensions of rural female workers revealed that Nigerian women have wider hands and shorter fingers than comparative samples from Europe and China. Two sources of hand tools were identified; those that emanate from IAC mass-production suppliers, and those that are hand-crafted in Nigeria following centuries-old tradition. The authors do not indicate which of these sources relates more to the anthropometric mismatches they identified, but it appears that tools requiring a power grip generally had shorter and thicker handles than are optimal for reducing palmar force concentrations in Nigerian women, while those requiring a claw grip were designed for users with longer fingers. Wisner’s (1985) recounting of Manuaba’s (1976) classic story in the annals of cultural ergonomics deserves retelling here, for the reader is unlikely to encounter a more dramatic example of failure to set solutions in context. When a new variety of rice, producing three bumper crops per year, was introduced in Bali, the immediate result was acute back strain among the harvesters. The wonder rice was shorter of stem, demanding excessive trunk flexion in its harvesting. Traditionally the longer stemmed, less prolific variety had been head-loaded out of the rice 236 P. A. SCOTT AND J. CHARTERIS paddies in 20 kg bales. Because the short stems could no longer be tied into bales the rice now had to be beaten and scooped into sacks in situ, adding several more steps of stooped physical labour to the thrice yearly harvesting process. These examples indicate that work, in context, demands endemic solutions.

Perceptions

Wyndham’s (1975) account of ethnically-based confusions when non-literate South African workers were confronted by unintelligible safety posters, conceived from the perspectives of an “alien culture,” still makes informative reading today. Over a quarter of a century later, the problem remains, notwithstanding increased literacy and the advent of, and increased access to television. Wyndham pointed out that it is essential to acknowledge the influence of culture on workers’ perceptions of information, particularly in the form of visual displays. Wisner (1985) also pointed to problems associated with translating instructions from the original into other languages; if the basic concept being communicated, by whatever method, is not fully understood by the interpreter, then the information risks distortion in various ways. A semiliterate worker may do the opposite to what is required simply because his perception of the instructions was the opposite of what was intended. One must query in such cases where the blame lies. Wisner’s caution was confined to cases when one other language is involved: it scratches the surface of misinterpretation in a country such as South Africa where there are 11 official languages. Scott (1985, 1986) addressed problems associated with the use of the Borg RPE scale and her attempts to get nuances such as “very, very light” or “somewhat hard” translated into four of the local languages. To overcome the loss of meaning in the translation, she followed the example of Hogan and Fleishman (1979) using pictoral representations of perceived exertional levels; but even then, to someone who has no concept of using a “jack hammer,” the image of one will be meaningless. The outcome of attempts over several years to establish a universally acceptable scale, was to revert to the original with its numerical ratings, using verbal anchors translated into the local language and supplementing these with basic diagrams depicting the common task of pushing a wheelbarrow up increasing gradients with increasingly demanding loads (see Fig. 4). Awareness of culture-based perception is so important, precisely because these perceptions, incorporating personal proclivities, social factors and educational backgrounds play a major role in the appropriateness of an individual’s response to any situation within a specific working environment. Wyndham (1975) pre- sented some interesting results of projects investigating the apparent differences with visual perceptions of rural African workers viewing safety posters. These Ergonomics in Industrially Developing Countries 237

Fig. 4. (A) Borg’s RPE Scale; (B) Scott’s Culturally-Adjusted RPE Scale. semi-literate workers failed to perceive dimensional views so that relative sizes and positions of people and objects were lost. A smaller figure in the background was not perceived as being further away, but rather as a child. Workers also perceived the contents of the pictures to be fully representative, so that if a picture only illustrated the upper body, the perception was that the subject had been severely injured, having lost both legs. Figure 5 is an example of a danger sign encountered in a South African factory. One also has to be sensitive to the use of colours and symbols, particularly those depicting animals, as these may carry very different connotations to people from different backgrounds. Red, for example, may mean love and tenderness (the heart), or blood and injuries (the heart). It could also mean stop (traffic lights) or danger (fire), therefore indicating that one should halt or run away. Red is also the colour of many trade union groups among whom it symbolises camaraderie and solidarity. This is a fascinating area for continued research, for although we glibly talk of a “global village,” the importance of understanding human variability and cultural influences is a challenge to all who strive for universal standards to 238 P. A. SCOTT AND J. CHARTERIS

Fig. 5. Failure of “Top-Down” Intentions. Note: This poster, quite apart from conveying nothing of workplace risk in terms comprehensible to the workers, caused a minor uproar when one group complained of discrimination because their tribal symbol, the lion, had not been displayed on the factory floor, while that of the crocodile, symbol of another tribe had. ensure a smooth transfer of goods and information from one area to another in the global market place. Wisner (1985) was at pains to show that assertions of cognitive differences between peoples, whether made consciously or inadver- tently, may reflect ethnocentricity rather than reality. He cites as evidence the “anthropotechnological islands of excellence” where multinational companies are able “...to maintain the same product quality and often the same productivity, no matter where their production units are located in the world.” So, when artists from one culture create safety posters unintelligible to a work-force from another, the issue is more likely one of culture-based misunderstanding than of cognitive difference. This apparently anomalous dichotomy, i.e. a widening First-World, Third- World gap on the one hand, and its obliteration in small pockets of success on the other, deserves specific treatment. In IDCs, ergonomics issues may appear Ergonomics in Industrially Developing Countries 239 on cursory view to be inherently unsophisticated: a hoe is redesigned to suit the anthropometry of a local user population; a previously unquestioned traditional work practice is modified to reduce energy cost. More often than not, however, these issues are not resolved by simplistic treatments. They may require deep appreciation of local mores as a prerequisite to any prospect of effective resolution. This appreciation may underlie the point of departure of a native ergonomist while a foreign consultant might be oblivious to the context and may, with far more sophisticated input, achieve far less. An energy-saving tool, for instance, might be eschewed for cultural reasons despite its mechanical advantages.

Cold IDC Realities

Why, some may ask, is a focus on developing countries so important. The facts presented in Fig. 1 should speak for themselves. Developed countries continue to exploit IDC resources, without regard to the local ill-effects caused. Thus, so long as coffee is supplied to IACs, the local agricultural ill-effects of the demand for more land under coffee production are seen as a local, rather than international responsibility. So long as West Africa supplies crude oil, for refinement overseas, the fact that West Africans cannot, in turn, afford the refined version of fuels they provided in the first place, is seen by IACs as a non-sequitor. Under this ethos, 75% of the world’s working population continues to work under horrendous conditions impacting negatively on every measure of what constitutes quality of life.

Excessive Rates and Loads

While Ergonomists in IACs have for decades recommended single-figure lifting rates per minute (NIOSH, 1981; Waters et al., 1993), in IDCs, the reality is that workers routinely exceed these recommendations two or threefold. Thus brick stackers in South Africa have been seen to exceed 23 lifts per minute while stacking 3 kg bricks, two at a time, in pallets of 500. At 17 pallets a day, each worker is moving nearly 26,000 kg in a shift under sometimes extreme climatic conditions (see Fig. 6). In some areas in South Africa, where temperatures approach 40 ◦C and humidity is as high as 98%, workers are allowed “flexi-work- time.” In consequence, they start at 4.00 a.m. and during the cooler hours work at excessive rates instead of pacing themselves over a standard 8-hour workday. By completing the day’s work in 4.5 hours they are able to access a second job to supplement their meagre wage from brick stacking. At the other extreme is the characteristic IDC phenomenon of excessive loads. One commonly finds a small undernourished worker moving huge masses unaided 240 P. A. SCOTT AND J. CHARTERIS

Fig. 6. Brick Stacking in a South African Brickyard. Fired Clay Bricks (3.04 kg) are Palletised, Two at a Time, at Rates Often Exceeding 20 lifts.min−1. and under poor environmental conditions. Figure 7, for example, illustrates a 62 kg worker maneuvering an 80 kg block of concrete in a sand dune township under construction in the summer months when ambient temperatures peak as high as 40 ◦C.

Fig. 7. A Gracile (62 kg) Manual Labourer Manoeuvring Concrete Blocks (80 kg) on a Sandy Base. Ergonomics in Industrially Developing Countries 241

Fig. 8. Laboratory Simulation of Crate Off-Loading Technique Observed in the Field.

Combinations of rate and load are seldom optimal, as the following example, encountered at a small family bottle sorting industry illustrates. In this scenario, stackers were required to unload a truck as quickly as possible. They had to lift crates of up to 15 kg mass from stacks as high as 1.8 m and then slide the crates over a distance of 3 m to the end of the truck (see Fig. 8). Heights, loads and slide distances varied, but what did not alter substantially was the pace of work, which averaged 15 lift-slide exertions per minute. The reason for the pace is simple: the greater the tonnage of crates collected and cleared, the better the remuneration. Investigating manual work in a soap factory in India, Samanta and Chatterjee (1981) reported that employees were moving 6–16 kg loads, at 8–16 lifts per minute. Working heart rates were as high as 170 bt.min−1. Unfortunately there was no report on the nutritional status of these workers. Lambert et al. (1994) reported on sugar cane stackers in South Africa, whose average mass-loss was 1.3 kg per shift, despite ingested energy of 5281 kJ while working. Typically, cane workers expended 14,127 kJ in meeting work requirements. Characteristic of IDCs are individual and small group (family) subsistence enterprises. While ergonomists in IACs tend to focus on large companies which 242 P. A. SCOTT AND J. CHARTERIS are clearly easier to work with, ergonomists in IDCs must deal with both the formal and informal sectors in their attempts to establish an ergonomic ethos in developing areas. Again, the importance of understanding traditional patterns of work is emphasised. It is well known that loads in many developing countries are moved by the people (not machines) and the ubiquitous method of carrying in Africa is “head loading.” Maloiy et al. (1986) proposed a “free ride” hypothesis: that light loads can be headloaded at no additional energy cost than that of unloaded walking. Charteris et al. (1989a, b) tested this hypothesis using local rural South African women (see Fig. 2). Their findings corroborated those of Maloiy and associates, confirming that head-loads of up to 25% body mass cause no significant change in energy cost. In the real world, however, head-loads exceeding 70% of body mass are common, and it is culturally accepted that women are responsible for most of the carrying of these loads and that it is the older women who carry the heaviest loads.

Specific Areas of Concern

Although ergonomists in affluent countries purport to show some interest in the needs of workers outside the workplace, in reality, the focus is on the working third of an employee’s day. On the other hand, if ergonomists in IDCs do not take detailed cognizance of the 24-hour lives of workers, they fail to come to terms with socio-economic and health conditions which substantially impact on work capacity. This is of particular significance, since the bulk of the work requirement in IDCs is of a physical nature. Genaidy et al. (1992) introduced an Ergonomics Stress Index (ESI). While they focused on IACs, Scott (1999b) emphasised IDCs. Both advocate that in any physically demanding task, the physical capabilities of the workers should be substantially greater than the physical demands of the task. But, as Scott points out, more often than not in IDCs, the reverse is true; workers are characteristically over-taxed by the manual operations required of them. Particular sensitivity is required when addressing the compounding factors which independently and collectively impact on worker selection, work capacity and worker’s attitudes and commitment to the job in IDCs. The majority of workers come from very poor backgrounds, live under the most squalid of conditions and endure deficient nutritional input. Associated with these appalling conditions are health problems, including HIV, malnutrition and tuberculosis. The outcome is lowered work capacity, exacerbated by limited education and neglect of skills training, forcing people to take on unskilled and physically demanding tasks. The result is that there is a complete imbalance between energy input and energy expenditure. Workers in this scenario are often perceived as lazy, Ergonomics in Industrially Developing Countries 243 whereas the fact is that they are in a chronic state of fatigue, highly susceptible to occupational diseases and musculoskeletal injuries. Ergonomics is about expert solutions rather than inexpert chronicling of prob- lems. Why then have we stressed so depressing a picture of corrupt officialdom, near hopeless medical ineptitude, and societal tolerance of the devaluation of human life, all markers of the antithesis of what is meant, in any culture, by the word “civilized”? The answer is because the disparity between protestation and practice, on any marker of “civilized society,” tends to be greater in IDCs than IACs, simply because the former do not have the resources to practice what they preach. Abuses and indignities that would never be allowed to spread in IACs, go unchecked in IDCs because survival is the consuming focus, not the niceties of human dignity. Poverty underlies the ethos in IDCs of tolerance of social indignities that would never be accepted in IACs, and it is within this context that ergonomics interventions in the Third World have to be seen. It takes a sensitive empathy for the total setting to effect small ameliorative changes in an IDC context. The IDC ergonomics approach is the antithesis of a Utopian programme: it is not about maximising safety, efficiency, operational reliability or comfort. It is, more realistically, about minimising the everyday human-system abuses which characterise IDCs but would never be tolerated in IACs. Minimising suffering as an ergonomics ethos produces an immediate perpetual stream of instances demanding urgent attention. It is an approach rooted in the Popperian tradition, and might be seen as the workplace equivalent of the macro-social order which Karl Popper called for in his epic post-war political philosophy (Popper, 1966). To those unfamiliar with it and unwilling to delve further, it is important to realise the implication for ergonomics, that a programme geared towards minimising worst abuses rather than towards maximising productivity and worker well-being, whether in an IAC context or in IDCs, involves ceaseless intervention, because at- tention is deflected from situations in less dire need which, if they were addressed, would simply serve to widen the gap between them and the worst of conditions. Thus, though we agree with O’Neill (2000) that in application, ergonomics is not in principle different between IACs and IDCs and is addressed only at “different starting points,” we stress that preoccupation with coddling issues in already reasonably viable workplace situations has no currency in IDCs, where the focus should be on unremitting intervention to reduce the worst of existing levels of abuse. In IACs, good ergonomics makes corporate sense because it ices an already rich multi-layered cake. By Third World standards, IAC ergonomics pampers the already economically viable and gainfully employed masses. In IDCs, good ergonomics puts dry bread on the table, with some prospect of sustainability. If it secures the most basic physical rights, viz., food, water, shelter 244 P. A. SCOTT AND J. CHARTERIS and energy to sustain life, it has succeeded. The slogan, “good ergonomics is good economics” applies universally, but it means very different things in the two (presently) disparate worlds. In that context, if dockside workers in an African port get some rest, some protective clothing, some say in how their work is done, some consideration of their safety, some small measure of job security and livelihood, then, so long as the process of improvement continues unremittingly, what matters more is how far they have come, not how far they still have to go. Moses Amweelo, Namibia’s Minister of Works, Transport and Communication, himself an ardent amateur ergonomist, has personally addressed problematic areas in cargo handling at the port of Walvis Bay, noting that in a labour cohort that is semi-literate and semi-skilled and for whom manual labour is a way of life, any machine-assisting innovation intended to reduce work demands and improve productivity will be strenuously resisted if it is perceived as technologically threatening and job-usurping (Amweelo, 2001). As the driving force for an ergonomics orientation at least in the service departments under his Ministry, Amweelo has inculcated an ethos steeped in the tradition which Hendrick (1998) and Imada (1991) in IACs, and Kogi (1985) and Scott (1997) in IDCs have argued for, namely intervention via an all-inclusive participatory process in which the involved people have most of the say. This tradition has long been advocated in Southern Africa, and is the means whereby Scott (1999c, 2001) seeks to promote “no cost – low cost” interventions in respect of the people whose work-practices are under proposed revision. Apparently the repeated admonitions of Kogi and co-workers (Kogi, 1985, 1997; Kogi & Sen, 1987) to ensure work-practice revisions of the people, by the people, for the people, have not gone unheeded and this luminary in the field of IDC ergonomics deserves credit as one of the foremost contributors to participatory ergonomics. Scott (1997) proposed an ethos of “cooperative co-responsibility” in order to establish a commitment from employees and employers alike, to work together to identify problems and improve working conditions.

LOOKING FORWARD

Kogi, O’Neill, Scott, Sen, Shahnavaz and Wisner have all repeatedly stressed the form in which IDC ergonomics works best: culture-compatible, participatory, small-scale, unsophisticated, cheap improvements are the ones that are best assimilated. What is often not added, but surely always assumed by IDC ergonomists, is that unless this small-step approach is relentlessly pursued, there Ergonomics in Industrially Developing Countries 245 is little prospect of the technological heterogeneity within any specific IDC being reduced, and none of the gap between IACs and IDCs being closed. It is the thesis of this chapter that the IDC problem is well understood, having been enunciated expertly by the workers in this field cited here. What is needed in the new millennium is practitioners, in increasing numbers, to do yeoman field work implementing culturally compatible participatory interventions. This can be established through collaborative work between global leaders in the field of ergonomics and concerned local ergonomists. A balance must be established between re-inventing the wheel, and modification of theoretical principles and practical applications to suit indigenous workers in local conditions. Input from ILO and WHO personnel dedicated to improving working conditions and occupational health already substantially alleviates IDC workplace problems. Ergonomists and health professionals, by acknowledging each-other’s areas of expertise, identifying links between poor working conditions and associated diseases and injuries would benefit if a “functional partnership” was created between the various disciplines which share a common concern and common objectives; the improvement of working conditions, in order to enhance the safety and well-being of workers while at the same time improving productivity. It is self-evident that a key component of the implementation of Ergonomics in IDCs is education. In order to help IDCs help themselves – they need to understand how to identify and address problems. To this end, the ILO and IEA collaborated in publishing “Ergonomic Checkpoints” in 1996. Every effort was made to ac- knowledge cultural differences and to present a point form “what-to-look-for” and “steps-to-take” checklist in a well illustrated basic text. There are also a growing number of “Ergonomic Guidelines” being developed in various areas which are basic and easy to implement in situ. While these checkpoints and guidelines play an important role in facilitating awareness of basic ergonomics, it is important to acknowledge that ergonomics is based on rigorous science which implies a need for tertiary education programmes in ergonomics. Fortunately there are an increasing number of such programmes being developed in IDCs themselves, together with distance learning programmes involving collaboration between universities in IACs and IDCs. Tertiary ergonomics education programs in IACs should expand and facilitate access to students from IDCs, while IAC governments should underwrite this process as a far more efficient type of foreign aid than shovelling cash into bottomless pits. Ergonomics is about people at work, and most of them are at work in IDCs. Undoubtedly, most certified ergonomists are in IACs. The need is to “level the playing fields” so that ergonomists from around the globe feel themselves to be de facto as well as de jure integral of the whole. In ergonomics terms, this translates 246 P. A. SCOTT AND J. CHARTERIS into taking concrete steps to reduce the widening gap between the technophilic “men-on-mars” ethos so seductive to ergonomists of the most industrially affluent societies and the “bent-handle-works-better-than-a-straight-handle” ethos of the struggling ergonomists in the world’s least modern nations. In attempts to assist IDCs, millions of dollars in cash and technology have been handed out by IACs, with the result that, as IDCs attempt to participate in global markets, they are not only coping with the numerous negative factors referred to throughout this chapter, but they are also going through incredible transitional processes, and ergonomists operating in this unstable and troubled ambience must try to manage the challenging mix of sophisticated “hi tech” (and limited education of workers) and basic manual labour (and poor nutrition and health status). There is no panacea to the problems facing IDCs, as they all differ in the level of advancement and the extent of the problems within the working (and daily living) conditions. Therefore, in order to assist IDCs to fulfil their potential, it is essential to understand the culture, needs, problems, strengths and limitations of the indigenous work force within their own micro and macro environment. At the recent “World Environment Summit” held in South Africa (2003), a common cry was that developing countries do not want hand-outs, they want fair and unrestricted trade. And global trade means they need to meet global standards in both the quality and quantity of their products. Ergonomics, in our view, has a responsibility to play a significant role in making IDCs viable and contributing members of the global village.

REFERENCES

Amweelo, M. (2001). Safety and accident prevention in cargo handling at the port of Walvis Bay. Ergonomics SA, 13(1), 37–49. Borg, G. A. V. (1970). Perceived exertion as an indication of somatic stress. Scandinavian Journal of Rehabilitation Medicine, 12, 92–98. Charteris, J., Nottrodt, J. W., & Scott, P. A. (1989a). The “free-ride” hypothesis: A second look at the efficiency of African woman headlaod carriers. South African Journal of Sciences, 85(1), 68–71. Charteris, J., Scott, P. A., & Nottrodt, J. W. (1989b). Metabolic and kinematic responses of African women headload carriers under controlled conditions of load and speed. Ergonomics, 32(12), 1539–1550. Genaidy, A. M., Karwowski, W., Guo, L., Hidalgo, J., & Garbutt, G. (1992). Physical training: A tool for increasing work tolerance of employees engaged in manual handling tasks. Ergonomics, 35(9), 1081–1102. Hendrick, H. W. (1998). Integrating macroergonomics with EWA and anthropotechnology: A systems approach for work system design. In: O. Brown, Jr. & H. W. Hendricks (Eds), Human Factors in Organizational Design and Management (Vol. V). Amsterdam: North-Holland. Ergonomics in Industrially Developing Countries 247

Hogan, J., & Fleishman, E. (1979). An index of the physical effort required in human task performance. Journal of Applied Psychology, 64, 197–204. Imada, A. S. (1991). The rationale and tools of participatory ergonomics. In: K. Noro & A. S. Imada (Eds), Participatory Ergonomics. London: Taylor & Francis. Kaplan, M. (1995). The culture at work: Cultural ergonomics. Ergonomics, 38(3), 606–615. Kogi, K. (1985). Participatory approaches in applying ergonomics for workplace improvements in developing countries. In: I. D. Brown et al. (Eds), Ergonomics International 85. London: Taylor & Francis. Kogi, K. (1997). Ergonomics and technology transfer into small and medium-sized enterprises. Ergonomics, 40(10), 1118–1129. Kogi, K., & Sen, R. N. (1987). Third world ergonomics. In: D. J. Oborne (Ed.), International Reviews of Ergonomics (Vol. 1). London: Taylor & Francis. Lambert, M. I., Cheevers, E. J., & Coopoo, Y. (1994). Relationship between energy expenditure and productivity of sugar cane cutters and stackers. Journal of Occupational Medicine, 44, 190–194. Maloiy, G. M. O., Heglund, N. C., Prager, L. M., Cavagna, G. A., & Taylor, C. R. (1986). Energy cost of varying loads: Have African women discovered an economic way? Nature, 319, 668–669. Manuaba, A. (1976). Problems in ergonomics in Bali, Indonesia. International Symposium on Impact of Industrialisation and Ergonomics in Asia. Tokyo. Maurino, D. (1993). See Kaplan, M. (1995). Nag, P. K. (2001). Presidential address and welcome: International Congress on Humanizing Work and the Work Environment (December 11–14), Bombay, India. NIOSH (1981). A Work Practices Guide for Manual Lifting. NIOSH Technical Report No. 81122, U.S. Department of Health and Human Services, National Institute for Occupational Safety and Health, Cincinnati, OH. O’Neill, D. H. (2000). Ergonomics in industrially developing countries: Does its application differ from that in industrially advanced countries? Applied Ergonomics, 31, 631–640. Okunribido, O. O., & Olajure, K. A. (1999). A survey of hand anthropometry of female rural workers in Ibadan, Western Nigeria. Ergonomics SA, 11(1), 2–6. Popper, K. (1966). The open society and its enemies (5th (rev.) ed.). Routledge & Kegan Paul. Pradhan, C. K., & Thakur, S. (2002). Subjective assessment of work stress in the automobile transport sector. Ergonomics SA, 14(1), 2–7. Roster of the IEA (2002). Samanta, A., & Chatterjee, B. B. (1981). Energy expenditure in manual load carriage. Industries Health, 19, 145–154. Scott, P. A. (1985). The rating of perceived exertion in a multi-ethnic society: A review of the problem and some preliminary solutions. Proceedings: First Annual Conference of the Ergonomics Society of Southern Africa. C.S.I.R. Pretoria (5–6 February), 16, 1–4. Scott, P. A. (1986). Ratings of perceived exertion using diagrammatic delta-intensity tasks. Proceed- ings: Second Annual Conference of the Ergonomics Society of Southern Africa (14–15 April), U.C.T. Cape Town. Scott, P. A. (1997). Macro-micro approaches to establishing ergonomics in industrially developing countries. Proceedings: I.E.A. 13th Triennial Congress (June 29 to July 4), Tampere, Finland. Scott, P. A. (1999a). Potential and problems associated with ergonomics in industrially developing countries. Proceedings: Hellenic Ergonomics Society Conference (31 August to 1 September), Santorini, Greece. Scott, P. A. (1999b). The effect of a work conditioning programme on manual labourers in South African industry. International Journal of Industrial Ergonomics, 24, 253–259. 248 P. A. SCOTT AND J. CHARTERIS

Scott, P.A. (1999c). Global guidelines for Ergonomics in Industrially Developing Countires. ABERGO ’99 Conference (3–5 November), Salvador, Bahia, South America. Scott, P. A. (2001). The key to humanizing the work environment and improving productivity in Industrially Developing Countries. Proceedings: Humanizing work and work environment, International Indian Society’s Ergonomics Conference (11–14 December), Mumbai, India. Shahnavaz, H. (1996). Making ergonomics a world-wide concept. Ergonomics, 39(12), 1391–1402. Waters, T. R., Putz-Anderson, V., Garg, A., & Fine, L. J. (1993). Revised NIOSH equation for the design and evaluation of manual lifting tasks. Ergonomics, 36(7), 749–776. Wisner, A. (1985). Ergonomics in industrially developing countries. Ergonomics, 28(8), 1213–1224. World Bank Group: Data and Statistics (2002). Wyndham, C. H. (1975). Ergonomics problems in the transition from peasant to industrial life in South Africa. In: A. Chapanis (Ed.), Ethnic Variables in Human Factors Engineering (pp. 115–135). Baltimore: Johns Hopkins University Press. 9. COGNITION IN NATURAL SETTINGS: THE CULTURAL LENS MODEL

Helen Altman Klein

ABSTRACT

Intercultural interactions, in domains such as civil aviation and international peacekeeping, expand awareness of national differences in cognition. At the same time, experience with national differences in natural settings provides a more complex picture of cognition. The Cultural Lens Model captures the nature and origin of the cognitive differences. This paper reviews cognitive dimensions that vary over national groups. It uses the Cultural Lens Model to describe the implications of these cognitive differences for five intercul- tural challenges: problem definition, planning, coordination, prediction, and training. Finally, the paper suggests mechanisms for increasing international effectiveness in the face of cognitive differences.

INTRODUCTION

International enterprises are becoming more common. Work environments are now likely to include professionals from Western nations – the United States, Western Europe, and the English-speaking world – as well as from other regions of the world. Equipment developed in one nation may be exported to distant regions. Business leaders work with their counterparts around the world. A typical

Cultural Ergonomics Advances in Human Performance and Cognitive Engineering Research, Volume 4, 249–280 Copyright © 2004 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1479-3601/doi:10.1016/S1479-3601(03)04009-8 249 250 HELEN ALTMAN KLEIN

Western hospital, research institution, or technical facility includes multinational professionals with varied training and backgrounds. Scientists can coordinate complex research efforts with colleagues they know only through the Internet. By assembling multinational groups they can tap a wider range of expertise. We now have unprecedented opportunities to share resources, extend perspectives, and expand markets. We can benefit from globalization only as we can work with people from dif- ferent nations. Human factors specialists and ergonomists see national differences in settings as varied as the space station, military and peacekeeping operations, commercial aviation, and global business. Professionals in these domains must detect problems, make sense out of complex and confusing information, plan, make decisions, and coordinate with others. At the same time they must manage the stress and uncertainty often endemic in these domains. In order to accomplish task demands, professionals need to understand how people from other nations perceive events, how they think about issues, how to anticipate their actions, how to influence their beliefs, and how to negotiate agreements. The shift towards multinational operations, together with a growing complexity of work environments, presents a problem: Intelligent and thoughtful people from different national groups sometimes identify different problems, make different plans, negotiate and coordinate differently, and make different decisions during complex cognitive tasks. Cutting edge technology and procedures, carefully and competently developed in one nation, may be incompatible with the equipment, procedures, and professional practices of other nations. Training packages that are effective in one nation can receive harsh criticism in others. Professionals often struggle during interactions with competent, well-meaning peers from other nations. These problems can compromise productivity and the quality of work. Here are two examples:

A pilot reported that a pump was not working. The maintenance crew is supposed to pinpoint the exact cause of the problem. But they didn’t. They replaced three separate parts as they had done before. Had they gone to the manual, they would have seen that two of the parts were unnecessary. The shotgun approach is very costly in time and materials (Klein, H. A., Klein, G., & Mumaw, 2001, p. 17).

Many citizens in Bosnia-Herzegovina have guns and explosive devices. Operation Harvest is a program to collect weapons in order to lower the risk of aggression. Some soldiers will act only when they have complete information, even if it means fewer collections. They want to meticulously script each home visit and specify many variations. Officers from other countries are critical of this caution. They want a general plan and the flexibility to respond to deviations along the way. This attitude troubles the first group of officers: “When we sit down to plan, they drive us nuts! They want to keep everything open. We have to make decisions and we should do it when we have the time to think!” (Klein Associates, 2002, p. 16). Cognition in Natural Settings 251

These examples illustrate differences seen in natural settings between people from different nations. These differences alter how people collaborate in work settings and follow procedures. Participants in multinational interchanges often struggle with the seemingly strange, counterproductive, and/or dangerous actions of others. These can make it difficult to establish and sustain common ground during collaboration. There are two common explanations for such differences: Explanation 1. People from some other nations are stupid, lazy, amoral, and/or obstructionist. This explanation is usually couched in polite terms: “With the education they get, this is all we can expect.” “They are just not used to working hard – look at their history.” “They really don’t seem to care about safety. Life is not valued as much in their country.” “As soon as you take your eyes off them, they stop working. They haven’t been taught responsibility.” From this perspective, there are two solutions. You can try to change the people who are different from you. Alternately, you can use work-arounds to avoid giving critical tasks to personnel from “suspect” nations. Both solutions are found in aviation and multinational military operations. The commercial aviation industry strives to provide the safest possible transportation regardless of the nationality of personnel or the carrier. Safety is emphasized through procedures and guidelines. This includes standardization of training and certification for pilots and maintenance technicians. Increased automation is, in part, a response to concerns about the variability in performance. It has not yet been possible, however, to make enough rules and specify enough procedures to protect this complex work setting. In multinational military operations, Western officers are often placed in com- mand of non-Westerners in order to avoid their “mistakes” and “misjudgments.” Particularly important tasks may be diverted to Western teams. This solution can sometimes provide effective but costly stopgaps for routine operations. It does not, however, address the underlying problem of national differences. This problem can lead to performance breakdowns when people from different nations need to coordinate difficult tasks, particularly under time pressure. At the very least, it makes it difficult to fully use the strengths and expertise of some other nationals. There is a second explanation that has guided my research on cognitive differences in natural settings: Explanation 2. People from some other nations differ in cognition in ways that result in different perceptions, judgments, and decision making. It is easy to assume others think as you do; that the cognition of others mirrors your own. People are easily insensitive to cognitive differences because cognitive 252 HELEN ALTMAN KLEIN processes develop outside awareness. Most cognition research comes from West- ern scientists using Western paradigms with Western subjects and it supports this universal assumption. There is, however, a growing body of research challenging the universal nature of cognition. Explanation 2 demands a shift away from control tactics and work-arounds. To function effectively in multinational settings, participants must understand national differences in cognition. Knowledge from cross-cultural research can help shape the skills needed to adapt to national differences. Human factors and ergonomics professionals need to consider the demands of natural settings. These demands often contrast with more traditional work in cognition (Ross, McHugh, Moon, Klein, Armstrong & Rall, 2002). Traditional research advanced theoretical understanding of human cognition by focusing on a core set of operations including short-term memory, attention span, category judgments, spatial representation, and puzzle solving as commonly described in cognitive psychology textbooks. Cognitive psychologists typically looked at the performance of college students in controlled laboratory settings with relatively brief time scales. This cognition in isolation differs from the cognition often needed in natural settings. Natural settings may involve interactions between multiple processes, contexts, and players. This makes cognition in natural settings particularly vulnerable to national differences in cognition. This is similar to how each medication can have different effects depending on other medications present in the system. In the last decade, cognitive psychology has pushed the boundaries by considering natural settings. This has required different methods and has identified new mechanisms for cognition. The field of Naturalistic Decision Making has emerged and has extended the science of human cognition (Zsambok & Klein, 1997). As researchers move toward natural settings, they need to extend their methods and approaches beyond those of traditional research for three reasons:

Traditional research task demands may not map directly onto those of natural settings. Laboratory studies may minimize extraneous pressures while actual practitioners may experience information overload, ill-defined problems and goals, time pressure, and high stakes in natural settings (Orasanu & Connolly, 1993). Some cognitive functions needed in natural settings may be emergent properties not open to laboratory control. Even a sophisticated flight simulator cannot include the life-and-death reality of flight.
Traditional paradigms usually initially use subjects naive to the task so that experience does not introduce variability. Laboratory tasks can eliminate the confounding effects of experience but experience is often important for performance in natural settings. Cognition in Natural Settings 253

Traditional research assumes that cognitive operations are universal. The complex cognitive processes needed in natural settings, however, exhibit variation across national groups. Natural settings are increasingly likely to include people from different national groups.

Natural and laboratory settings provide complementary rather than competing accounts of cognition. Laboratory investigations typically seek to identify and delineate basic processes, whereas investigations in natural settings attempt to describe complex relationships. When cognition is embedded in the world, it is difficult to separate the “purely” cognitive elements from the behavioral and social contexts of cognition. The purpose of this chapter is to explore cognition in natural settings and to present the Cultural Lens Model, a framework for understanding national dif- ferences in cognition. First I describe the Cultural Lens Model. Then I review cultural dimensions that affect performance in natural settings. Finally, I examine implications of the model for understanding vulnerabilities introduced by national differences.

NATIONAL CULTURE AND THE CULTURAL LENS MODEL

A national culture provides a functional blueprint for group member’s behavior, social roles, and cognitive processes. Culture provides rules about food safety, cleanliness, and health care. It provides the basis for verbal and nonverbal communication, and guidelines for acceptable social behavior and emotional expression. Culture also provides cognitive tools for making sense out of the world. National cultures are rooted in the physical and social ecology of the national group. If the national culture is responsive to the physical and social ecology, it confers a survival advantage. Cultures that lacked effective guidelines for survival in their particular physical and social setting became extinct. Because national cultures emerge from a particular setting, the shared experiences of the setting shape a common vision of the world. Cultures are dynamic systems. They emerge from a particular setting and they change as that setting evolves over time. Modernization, for example, requires adaptation. Cultures are composed of integrated components rather than a haphaz- ard collection of interchangeable parts. One change has repercussions throughout the system. A culture cannot adapt to the cognitive demands of industrialization without also altering social and educational patterns. Industrialization generally brings changes in reasoning and education along with changes in family structure. 254 HELEN ALTMAN KLEIN

The integrated nature of cultural components means that some cultural elements regularly occur together. These clusters characterize types of cultures. Industrial nations, for example, are likely to show social and cognitive similarities. This def- inition of culture is consistent with Berry’s Ecocultural framework (Berry, 1986; Segall, Dasen, Berry & Poortinga, 1990; Triandis, 1994).

The Cultural Lens Model

The Cultural Lens Model provides a framework for understanding the concept and origins of national culture. Figure 1 shows the Cultural Lens Model. The model assumes that members of a national group, growing up in similar ecological and social contexts, have shared experiences. It also assumes that members of a national group have experienced child rearing based on contextual commonalities. Taken together these similarities, through learning and modeling, generate common behavioral, social, and cognitive patterns. The model captures the dimensions that typify national group differences. The dimensions provide a lens through which each member of a national group “sees” the world. The lens filters and organizes incoming information, focuses sensemaking, structures planning and adaptation activities, and frames interactions and communication. Because members of a national group tend to share patterns of origin, they also share the way they see the world. This provides common ground during complex cognition in natural settings.

Fig. 1. The Cultural Lens Model. Cognition in Natural Settings 255

When people differ in origins and hence in cognition and the behavioral and social context of cognition, their views of the world will not match. Mismatched views of the world can create dissonance and conflict during international interchanges in natural settings. This is seen in multinational peacekeeping efforts where allies struggle to work together on common goals. It is a problem in commercial aviation where standard Western procedures present surprising challenges to non-Western users. While mismatches are a problem, they can also provide clues for working successfully with others in a multinational setting. When mismatches are identified, differences can be accommodated. The application of a cultural lens is designed to prepare people to work in international interchanges.

Origins Humans, like other species, show phenotypic plasticity: the same genetic endow- ment can lead to different phenotypic expressions depending on external pressures (Agrawal, 2002). Although all people are born with an essentially identical range of endowments and potential, each group develops with different resources and different physical and social pressures. Common ecological and social pressures lead to common adaptive outcomes (Diamond, 1999). Adaptations include cognition, but also behavioral patterns such as approaches to subsistence and to defense against social or natural threats. Members of a group share the experiences associated with their resources and pressures. These experiences provide common ground and support for group members in their physical and social ecology. Child-rearing practices are a mechanism for translating resources and pressures into behavior, social roles, and cognition (Berry, 1986; Segall et al., 1990). They shape and maintain the skills, expectations, and goals needed for individual survival and group continuance. Parents use rewards and incentives as well as punishments to establish necessary behavior, social roles, and cognitive patterns (Rogoff & Lave, 1984). Role modeling is another mechanism for shaping adult patterns (Mishra & Tripathi, 1996). From early handling and weaning, to discipline and schooling, the child’s plasticity is shaped to the patterns that are successful in the ecological and physical contexts. Because culture is responsive to ecology, similarities in ecology lead to similar practices while differences lead to different practices. Traditional agriculture depends on soil condition, precipitation, sunlight, and length of growing season. It requires a fixed domicile and adherence to seasonal markers. Group members follow past practices and the guidance of knowledgeable elders. Farmers cannot gamble on planting early or harvesting late. Agricultural groups generally punish risk taking and games of chance. Rewards and punishments are geared to developing obedience and conformity. Hunting/gathering groups, in contrast, tend to live in unstable environments where innovative reasoning confers an 256 HELEN ALTMAN KLEIN advantage. Groups that have competed for resources with hostile neighbors, tend to encourage aggression. As nations move towards industrialization, families become more nuclear and the reasoning more flexible and innovative.

Vision: Seeing the World Experience within a national group provides the basis of behavior and social cus- toms. It also sets expectations of others. People assume that everyone behaves and thinks as they do. Standards of modesty, food taboos, and religious observances all seem universal. Social patterns of prescribed roles and relationships seem obvious. For most Westerners, it is generally inappropriate for women to expose their breasts in public and also excessive for women to wear chadors in public. The way Westerners define modesty seems just about perfect. It is difficult to see modesty in dress or food taboos as culturally defined until encountering a different culture. When we remain in our own national culture, everything seems natural and correct. Behavior and social differences make multinational interchanges an enduring challenge. We have only to skim the travel guides to see differences in behav- ior and customs. Expect a hardy breakfast of meats and cheeses in Sweden. Arrive promptly for meetings in Germany. Anticipate extended pre-negotiation formalities in Japan. Cultural rules for national groups govern terms of respect, acceptable physical distance, polite conversation, and appropriate dress. Just as travelers struggle with language differences, behavioral differences present formidable difficulties. In the same way that early experience sets expectations for behavior and social roles, it also directs how people see the world. Because those from the same group “see” the world in similar ways, they interpret events and make decisions similarly. They share a “lens” for making sense of the world. The lens provides common values, beliefs, and reasoning scripts that group members use to interpret and react to the environment. Like behavior and customs, logic and reasoning processes seem obvious and optimal. National differences influence how people make judgments, reason, and make decisions. Peacekeepers from similar nations are likely to select similar strategies. Pilots from similar nations are likely to share judgments about safety compliance. Westerners are generally happier with Western medicine. They do not see them- selves as making choices but only as doing the obvious – what should be done. Teamwork requires anticipation of how others will define the problems, respond to unexpected events, and revise ongoing plans. For high-stakes, time-pressured decisions this is difficult even when team members see the world similarly. When team members see the world differently, it can seem impossible. Cognition in Natural Settings 257

There have been several efforts to describe differences in cognition and in the behavioral and social context of cognition important in natural settings. These dimensions are culturally dependent and thus highlight barriers to common vision and fault lines along which international interchanges falter. The dimensions emerge from four research traditions:
Kluckhohn and Strodtbeck (1961) used anthropological methods to identify differences among pre-industrial groups. They noticed that some groups planned for the weeks ahead while others looked to the long-term needs of their grand- children. The term Time Horizon describes this difference. They also noticed that some groups valued work and achievement while others valued people and relationships. Achievement vs. Relationship captures this dimension. Finally, some groups accommodated to events in the world while others appeared driven to master them. Mastery vs. Fatalism describes this difference.
Hofstede (1980) used the methods of social psychology to identify differences among employees of a multinational corporation. Some groups were comfort- able with uncertainty while others worked to provide definition. Tolerance for Uncertainty describes this dimension. Some groups respected and conformed to hierarchical structure while others showed an egalitarian structure. Power Distance describes this dimension.
Markus and Kitayama (1991) looked at reasoning across national groups. While most people are capable of a range of reasoning, some groups prefer reasoning grounded in concrete reality while others favor more speculative, hypothesis-based reasoning. The distinction is the basis for Hypothetical vs. Concrete Reasoning.
Nisbett and his colleagues (Hong, Morris, Chiu, & Benet-Martinez, 2000; Ji, Peng, & Nisbett, 2000; Morris & Peng, 1994) noted that some national groups attributed cause to individual characteristics while others looked to situational factors. Attribution describes this distinction. They also noted that some groups make decisions by seeking distinctions and choosing between options while others seek commonality. Differentiation vs. Dialectical Reasoning describes this dimension (Peng & Nisbett, 1999). These eight dimensions – Time Horizon, Achievement vs. Relationship, Mastery vs. Fatalism, Tolerance for Uncertainty, Power Distance, Hypothetical vs. Concrete Reasoning, Attribution, and Differentiation vs. Dialectical Reasoning – will be reviewed later.

Mismatches People can only “see” their culture when they encounter a mismatch between their lifelong patterns of thinking and that of other people. Americans may feel 258 HELEN ALTMAN KLEIN disconcerted when team members will not critique ideas during a planning session. They may feel irritation when others seemingly “give up” rather than seeking a way around a problem. Different natural settings have particular vulnerabilities for potentially troublesome and disruptive mismatches. Consider the conflicts created when people who are comfortable with and even prefer to keep procedures and rules flexible work with people who expect procedures and rules that are well defined and provide firm structure. Their ways of “seeing” the planning process will diverge and create disagreement. How specific should plans be? When should the plan be discarded? Is debate and conflict healthy or anxiety producing? It is hard to really discuss differences as each is based on a different but hidden assumption. Similarly, there are conflicts between people who attribute problems to the dis- positions of individuals – perhaps training or work style – and those who look to situational complexity, perhaps organizational policy. Based on these different attributions, each would select a different remedy. Do you provide training to the person you see as responsible or do you review the potentially faulty organizational policy to identify a systems solution? Again, the mismatch can create disagreement. A mismatch that plagues peacekeeping operations occurs when personnel from Western nations try to help those from emerging nations. In their well-intentioned attempts to support development, they may push for the rapid acceptance of innovative technology. They want to master the difficulties facing the host nation and provide a better life. If the people of the host nation differ in their approach to change, Westerners may view them as lazy and uncaring. The Westerners may be viewed as dominating and disrespectful by hosts who are more fatalistic and relationship oriented. Finally, when Western designers and engineers develop equipment, they incorporate their own vision of the world. They tacitly assume that all users will share their own reasoning. The equipment will function effectively with Western personnel. Performance, however, may be weak for some non-Western personnel. Westerners may view poor performance as resistant and uncooperative. This is easier than viewing the equipment as inconsistent with the cognition of its users. These mismatches are typical of the problems encountered in natural settings where nations come together to work and solve problems. In each, participants assume their own vision of the world is correct. Common ground is impossible and coordination suffers. I now turn to the accommodation of mismatches that is a part of multinational interchanges. Adjusting Mismatches: A Mechanism for Common Ground. When people differ in the cognition and the behavior/social dimensions important for a particular natural setting and task, there is a potential source of conflict and failure. People cannot adjust mismatches by altering their underlying cognitive processes, i.e. how Cognition in Natural Settings 259 they think about the world. Differences on dimensions cannot be changed at will because they reflect the demands of earlier experience. Even when the dimensions cause conflict in a new ecological and social context, they tend to persist. Although people can learn new content, it is difficult to acquire new reasoning forms. It may be possible to adjust mismatches by tapping the mechanisms initially used to create the dimensions in the first place. The Cultural Lens Model postulates a dynamic system in which the outcome of one action provides feedback for future actions and a mechanism for long-term awareness. Drawing on this dynamic, we can seek clues about how to sensitize people to use alternate thinking styles. Children rely on experience and modeling during development (Bandura, 1986). They watch for clues of reactions and events so that they can predict behavior and respond appropriately. Children learn to predict the actions of others by taking their perspective in the form of “make believe.” Perspective taking and modeling provides models for adult training. Experiential learning builds on perspective taking. When people can “pretend” alternate patterns, they can see how others see the same situation. As adults become aware of cognitive differences, they are more sensitive to signs of these differences in others. This sensitivity provides an extra “lens” to function more effectively with those from different nations. Cultural lens-based training can enable practitioners to see the world through the eyes of someone from a different nation. It can facilitate effective interaction by allowing a person to take the perspective of another person. For example, an American may be surprised when a foreign business associate shows distress at a change of plans. With an understanding of different reactions to uncertainty, the American would be able to select adaptive tactics to facilitate continued coordina- tion. This perspective is also useful for formulating procedures, preparing training material, and designing equipment.

Conclusions and Caveats

The Cultural Lens Model differs from other available models in that it is directed at intercultural mismatches in complex natural settings. Unlike most cognition and decision-making research, it can accommodate the cognitive variations introduced by national differences. Unlike cultural and cognitive research grounded in labora- tory studies with undergraduates, it can address the complexity of natural settings and the domain expertise of real practitioners. The Cultural Lens Model provides a framework for understanding the origins of national culture in their physical and social contexts. The ecology is the basis for child-rearing practices. Context and shared experience contribute to national commonality of behavior, social structure, and cognition. Together, collective 260 HELEN ALTMAN KLEIN experience provides a functional blueprint for a dynamic system of integrated behavior, social roles and personality, and cognition. The model captures the way experiences in a national group are translated into a view of the world. It describes the mismatches created when people differ and how knowledge of these mismatches can reduce barriers to multinational interactions in complex work settings. There are five caveats that should be considered when applying the model:

National culture is only one contributor to individual differences. Also important are differences from social class, genetic endowment, birth order, and the like. Understanding culture can increase accuracy in predicting group trends but cannot predict individual patterns.
Nations vary in homogeneity and hence in culture-linked characteristics. Southern and northern China differ. Rural Spain is different from urban Spain. Although the model could potentially incorporate within-nation variability, there is little data to specify within-national differences.
Nations can show ecological and ideological/religious commonality with neighboring nations and also with distant nations. The Cultural Lens Model could incorporate such commonality to extend predictions to unstudied nations. This would require a data base of national commonalities.
The Cultural Lens dimensions are likely to cluster. Identifying overlapping dimensions would identify clusters of distinct cognitive patterns affecting multinational interactions. Clusters extend the usefulness of the Cultural Lens Models for multinational interactions.
Nations change over time and over exposure to new ideas. Continual reexam- ination would be required to maintain the timeliness of dimension placements.

Cultural Dimensions in Natural Settings

Before I describe the implications of the Cultural Lens Model, I present the cog- nitive and social dimensions. These dimensions underlie the model and are the building blocks for applications in natural settings. Time Horizon. Time Horizon describes how far ahead people set goals and look to justify their actions. It influences how people assess information, plan, and make decisions. Time Horizon contributes to priorities about resources and actions. It influences customary responses to delay and impediments to actions. Present- horizon groups seek short-term goals, even at the expense of long-term ones (Adler, 1991). The distant goals and system building are downplayed because it is believed that nobody can see the future (Lane & DiStefano, 1992). In con- trast, future-horizon groups will sacrifice immediate payoffs for the expectation Cognition in Natural Settings 261 of long-term gain. They are less concerned with limits imposed by immediate circumstances (Adler, 1991; Lane & DiStefano, 1992).

The directors of a Japanese firm and a Canadian firm met in 1984 in Vancouver to negotiate the sale of coal shipments from British Columbia to Japan. The companies reached a stalemate over the length of the contract. The Japanese wanted the Canadians to sign a ten-year contract. The Canadians did not wish to commit a lengthy agreement as a more lucrative opportunity might appear. Whereas the Japanese wanted to plan for the long term, the Canadians were willing to leave the future open for the potential benefits of a more profitable agreement (Adler, 1991).

In aviation, routine maintenance takes equipment out of service, reducing immedi- ate productivity to achieve long-term productivity and safety. In some nations, the emphasis is on immediate over long-term productivity. Time Horizon is important in the balance between performance and safety. A present horizon may give on-time departures and low cancellation rates the highest priority. Maintenance may be ignored when it hampers productivity. This saves money immediately in exchange for low frequency, high-cost problems later on. Safety concerns lead to canceling flights in threatening weather conditions and when defects in equipment are suspected. They lead to expensive and time-consuming error-management approaches. The decisions of pilots and maintenance technicians as well as the reg- ulatory policies of managements reflect Time Horizon (Klein, H. A. et al., 2001). Training related to aviation safety typically carries an implicit assumption of a future horizon. For this reason it may be incomprehensible to those with a present horizon. Training must bridge the gap between the trainee’s Time Horizon and the designer’s intent. When a present horizon is characteristic of a national group, it becomes even more important to work with management to expand their under- standing of long-term considerations. In many other work environments, Time Horizon can be important. How are the various tasks prioritized? When do you reassign resources from short-term goals to a longer-term goal? Do you build costly infrastructure or speed production? Time Horizon directs the priorities for emergency management. When participants share Time Horizon, the process will be easy. Mastery vs. Fatalism. A mastery orientation is based on the belief that people are dominant over nature and can control their environment (Kluckhohn & Strodtbeck, 1961). This means that with enough time, money, and thought, almost anything can be achieved. This personal efficacy leads to seeking solutions to problems and working to implement change. Setbacks are seen as signals to try other routes or other methods. People with a mastery orientation are less likely to accept events as being beyond their control. Those who hold a fatalistic orientation respect the external factors that control their lives (Kluckhohn & Strodtbeck, 1961; Lane & DiStefano, 1992). To the extent that events are beyond one’s control, change 262 HELEN ALTMAN KLEIN is ineffectual or even inappropriate. Fatalistic thinkers strive for acceptance and adaptation rather than problem solving. A sense of mastery leads people to try to overcome seemingly impossible difficulties. North Americans and most Western Europeans show mastery (Lane, DiStefano, & Maznevski, 1996). They respect a “can do” attitude. Our folklore is filled with rags-to-riches stories, tales about people who let nothing take them from their quest. Asian cultures are more fatalistic (Wright & Phillips, 1980). In the Sioux City landing of an inoperative aircraft, the crew would not stop trying to master a set of fatal flaws. In a Guilin, China crash, a crew appeared to accept as inevitable a potentially solvable irregularity.

In 1989, the crew of United Airlines Flight 232 experienced an unusual confluence of failures. Three critical systems were lost when an engine failed and exploded, destroying the hydraulic systems. They had no ailerons to bank the airplane, no rudder to turn it, no elevators to control pitch, no leading edge flaps or slats to slow the airplane down, no trailing edge flaps for landing, no spoilers on the wing to help get down. On the ground, they had no steering, nose wheel, or tail, and no brakes. The Flight Manual had no contingency plan for recovery. Flight crewmembers were experts but none of them had any experience in this situation. They certainly understood the seeming hopelessness of the situation. This crew, nevertheless, managed to land in Sioux City. They found new uses of existing devices; they coordinated innovative procedures and brought down a fatally flawed aircraft (Klein, H. A. et al., 2001, pp. 8–9).

The 1992 crash of a 737 occurred at Guilin, China after one of the two thrust levers got stuck. As the other lever moved forward, it led to split throttles. The position of the throttles is, in itself, not a strongly salient feature. The airplane, however, started to lose control. This performance change should have been the tip-off. The pilots appeared as if they did not respond to the performance change and scan for the simple malfunction. Proper detection would have allowed the crew to move the thrust lever that was stuck – a correction that would have regained control of the aircraft. They didn’t do it (Klein, H. A. et al., 2001, p. 15).

Medical decision making can also depend on the balance between Mastery vs. Fatalism. Although death is inevitable, medical science has made great strides in the mastery of disease. Many medical conditions have no standard procedure and multiple courses. Western medicine tends to prolong treatment even when the treat- ment is painful, expensive, and untested. We are more likely to adopt aggressive means to a cure and talk about “the good fight.” Eastern medicine is faster to provide palliative care for patients with life-threatening conditions. This means comfort and compassion over heroics. Easterners are more apt to accept and adapt to fatal illness. Mastery vs. Fatalism influences how groups respond to threat and the cognitive resources they devote to change. The accommodation of fatalism may seem like a lack of concern for personal safety. The safety procedures and training of mastery may seem like an arrogant disregard of the inevitable. Cognition in Natural Settings 263

Achievement vs. Relationship. People differ in their emphasis on Achievement or Relationship. This distinction corresponds to Kluckhohn and Strodtbeck’s (1961) Doing or Being. It affects the way people approach life, work, and relationships. For achievement groups, work related activities are a central focus and accomplishment a defining goal. Western thinkers are generally achievement oriented. They look for task demands and how best to accomplish them. In relationship groups, cultures, interpersonal dynamics, and nurturing relationships are central focus. Differences in orientation can be a barrier during multinational interchanges. This arises when achievement people, who opt for task-centered interactions, work with relationship people who prefer nurturing, interpersonal interchanges. Those high in achievement believe that they can hasten change when plans are outlined, target dates set, and frequent reports made (Adler, 1991). Relationship-oriented people believe in allowing change to happen at its own pace without rushing things. Speeding up change is considered unwise. Differences in orientation interfere with ongoing operations when they are not recognized and managed. The conflicts and misunderstandings are seen in this interview record.

For American aviation personnel, keeping aircraft safely in the air and on schedule is a high priority. For them, maintenance personnel must be ready to support this goal. In China, workmen will stop to socialize or have lunch instead of doing a needed repair. An American Field Service Representative interpreted this behavior as showing that the workers do not understand the big picture of what the task implies. But for those workers, maintenance can wait but relationships with people cannot be postponed (Klein, H. A. et al., 2001).

Here is another example:

Army Major: “I was doing an exercise with the Italians. Their job was to provide security for the camp. To Americans, this means watching 24 hours a day, 360 degrees. The Italians would come in and do a pretty good job. Then mealtime would come. They would all leave to eat and drink and socialize. This was alien to Americans. At mealtime, we didn’t stop. We rotated through and the job got done” (Klein Associates, 2002, p. 9).

Achievement vs. Relationship influences work behaviors by setting priorities and framing decisions. Are team members critical or protective of the suggestions and assumptions of others? What are expectations for time use during work hours? How much attention is to be given to the needs of individual over the needs of the task? When participants share orientation, the process will be easier. Power Distance. Power distance is the extent to which members of a group expect the uneven distribution of power (Hofstede, 1980). The differences in interpersonal power and influence between superior and subordinate team members reflect differences in this dimension. This includes the acceptance of unequal distribution of power by institutions (Dorfman & Howell, 1988). Low Power Distance was associated with egalitarian working patterns and team 264 HELEN ALTMAN KLEIN interchanges. Those with low Power Distance expect to listen to others based on the merit of their ideas not on their rank. They expect that their own ideas will be evaluated based on merit. In contrast, those with high Power Distance expect that those with power will provide leadership and make decisions. Those who hold power maintain their rank in decision making. Power Distance has received considerable attention in commercial aviation.

A Boeing representative reported that on the airport employees’ bus in Saudi Arabia, the Captain always sat in the front row, and the First Officer in the second row. The cabin crew never talked to the Captain. He contrasted the situation with what he would find in the U.S.: “Because the Captain’s most important role is as manager, a lot is lost here. An American pilot will talk with the crew, which facilitates communication in flight. The Saudi Arabian style is not compatible with Crew Resource Management (CRM). The Saudi Arabian pilots do not appreciate this” (Klein, H. A. et al., 2001, p. 12).

The cockpit voice recorder told the story. During a flight in China, it became clear that the Captain had a bad attitude director indicator. The First Officer had a good attitude director indicator. Nevertheless, the Captain persisted in using his indicator and crashed the airplane. It would have been unseemly for the Captain to have to depend on the First Officer (Klein, H. A. et al., 2001, p. 12).

Chinese pilots take off in a storm if the air traffic controller says, “take off.” If an American pilot had to take off, they would order additional fuel in case they are diverted. The Chinese would never request additional fuel. It is not their job to decide to take off nor is it their job to decide on needing extra fuel. Chinese do not request to divert an airplane, even in an emergency. If they are told they can land, they attempt a landing (Klein, H. A. et al., 2001, p. 10).

In multinational collaborations, allies may have limited experience working together and they may need to work in different locations. The structure and the lines of command for decision making and for implementation may cross national boundaries. Several leaders may need to coordinate actions to maximize produc- tivity. This works best when everyone adheres to and respects the same command structure regarding responsibility for decisions. If participants differ in Power Dis- tance, they may struggle to define a working relationship rather than accomplishing goals. When operations are complex, it is sometimes lower ranking technical staff that has the expertise to make the best decision. Discrepancies in Power Distance can interfere with the use of expertise and can delay or compromise action. Tolerance for Uncertainty. Complex natural environments can include consid- erable uncertainty. Tolerance for Uncertainty describes how people function in the face of uncertainty. Those with low Tolerance for Uncertainty experience un- certainty as stressful and work to avoid it. They seek stability by adhering to formal rules and ritualistic behaviors (Lane et al., 1996). Those who are low in Tolerance for Uncertainty prefer detailed plans and abhor incomplete information. They resist changes because it is disconcerting and they feel unsettled until there Cognition in Natural Settings 265 is a final decision (Helmreich & Merritt, 1998). Disagreement among team mem- bers is stressful because it can generate questions and uncertainty. People low in Tolerance for Uncertainty value consensus (Lane & DiStefano, 1992). In contrast, those who are high in Tolerance for Uncertainty are comfortable with ambiguity and incomplete information (Dorfman & Howell, 1988; Hall & Hall, 1990; Hofstede, 1980). They adapt readily to change and act with limited information (Hall & Hall, 1990; Helmreich & Merritt, 1998; Lane & DiStefano, 1992). Those high in Tolerance for Uncertainty may ignore rules and rituals or treat them flexibly because they view them as ineffectual (Helmreich & Merritt, 1998). Those high in Tolerance for Uncertainty accept dissent and are not threatened by deviant ideas. This dimension was illustrated in the Operation Harvest example provided earlier. In this example, those with low Tolerance for Uncertainty want to have meticulously structured operations where all details are specified. The motto was “a dull day is a good day.” In contrast, those who are high in Tolerance for Un- certainty are comfortable with the uncertainty of the general plans. They are more likely to “play it by ear.” They are also more likely to encourage team members to identify and express problems with planning along the way. “Over planning” is viewed as counterproductive because you can never anticipate everything. The national differences in flight check rides illustrate this dimension. When the U.S. Federal Aviation Administration assesses a pilot’s skill, each pilot must perform well during structured situations and surprises. The assessment requires flexible responses and reflects high Tolerance for Uncertainty. Pilots are expected to make good decisions quickly in the face of uncertainty. During a Japanese check ride, pilots must perform precision responses during specified test situations. Low Tolerance for Uncertainty personnel see skill assessment as a well-defined task with unambiguous rules. High Tolerance for Uncertainty people are comfortable with general instructions and some flexibility to alter procedures as they go along. Tolerance for Uncertainty influences many aspects of work in natural settings. How much information is needed before you are willing to identify a problem? How much detail should a plan have? At what point are you willing to adapt ongoing plans? Hypothetical vs. Concrete Reasoning. Hypothetical thinkers use mental repre- sentations of future events to consider alternate outcomes (Markus & Kitayama, 1991; Tetlock, 1998). They examine situations by going beyond the actual occurrence. They ask “what if” questions. External events are analyzed in the abstract and do not have to be grounded in reality. It is a mental playing out of alternative strategies to imagine different outcomes (Markus & Kitayama, 1991; Roese, 1999; Tetlock, 1998). Most Westerners use hypothetical thinking to make plans and to examine their implications. They separate reasoning from 266 HELEN ALTMAN KLEIN reality to consider options in an abstract, hypothesis-driven manner (Markus & Kitayama, 1991). Concrete reasoning is a different approach to the same goal (Markus & Kitayama, 1991). Those people respect the constraints imposed by context and carefully integrate those constraints into their thinking. Reasoning is grounded in past personal and national experience not in mental simulation activities. Concrete thinkers work to improve future performance in similar contexts. Rather than ab- stract speculation, concrete thinkers review past events and their context in order to improve future performance. Concrete reasoners view hypothetical thinking as distorted because it is not grounded in reality. The differences in this dimension can be seen in the patterns of terrorists. Intelligence officers know that all terrorist groups are not the same. Some have a well-known modus operandi – always kidnapping or perhaps using explosives. Each attack is meticulously planned and information officers can trace the evolution of technique over time. If the group makes a mistake, they will correct it in their next attack. Other groups go for the most available target or approach at the time. A single approach cannot be identified. Their work, though flexible, is not precision but opportunistic and unexpected (Klein Associates, 2001). These two approaches are both potentially successful. The first pattern exemplifies concrete reasoning and builds on past experiences to generate precision plans. The second provides flexibility to take advantage of opportunities. Deterring these two patterns of planning requires different strategies. One trainer told us:

In an emergency, U.S. pilots use Plan A and keep trying to adjust it and make it work. But they are ready to switch if that first plan fails. I’ve had a Chinese student in the simulator where the situation is very different. There was an engine fire but the engine was not failed, and the indications did not suggest a failure. The student responded by rote as if there was an engine failure and crashed. He didn’t use visual cues. When something goes wrong, the Chinese pilots have trouble thinking hypothetically. They learn flying as procedural – Step 1, Step 2, and Step 3 and don’t try to figure out what is happening. If they start with Plan A, they finish with Plan A. In an emergency, U.S. pilots always have Plans A, B, and C in their head (Klein et al., 2001, p. 14).

U.S. pilots use hypothetical reasoning to achieve flexibility. During times of rapid and unanticipated change, hypothetical thinking will be more successful. One cost comes when different crewmembers have different backup plans and switch at different times. This flexibility also comes at the expense of precision. Concrete reasoning leads to increases in the precision and synchronization of thinking. When times are stable, context-based, concrete reasoning tends to be more effective and reinforced. The precision is achieved at the expense of flexibility. Attribution (Root Cause vs. Systems Approach): Faced with complex pressures or opportunities, people assign probable cause or describe dynamics. Attribution Cognition in Natural Settings 267 focuses attention and narrows the selection criteria for approaches or reme- dies. National groups differ in their Attribution of causality (Choi, Nisbett & Norenzayan, 1999; Ji et al., 2000; Norenzayan & Nisbett, 2000; Peng & Nisbett, 1999, 2000). Some people who use root-cause Attribution attend to the unique characteristics of the person or object. They locate responsibility primarily in the individual (Choi et al., 1999). People with root-cause Attribution consider retraining and counseling to be an appropriate remedy. They expect employee selection and promotion to be based on skills and knowledge. Those with a systems approach to Attribution adopt context-dependent and occasion-bound thinking (Schweder & Bourne, 1992; Wegner, 1987). When a problem is detected, the cause is attributed to the broader context and holistic so- lutions are implemented. Those with systems Attribution are uncomfortable with retraining that targets specific individuals. They favor efforts to modify organiza- tions and procedures while placing less weight on selection standards. Everyone may undergo extra training or the company may seek organizational changes. Attribution influences interactions in organizations. Those with a root-cause Attribution view themselves and others as composed of separate characteristics. People may see a colleague as an effective technician but weak in organizational skills, for example. They do not view a negative statement about one capacity as demeaning to the whole self. For those with root-cause Attribution who compartmentalize their traits and abilities, loss of face is an uncommon emotion. They work to avoid occasional failures, but not because it is a personal threat. In contrast, those with a systems Attribution view a negative statement about their work as a threat to their integrity. Loss of face occurs because of a criticism. Such reasoning means that people will work to avoid losing face. This may mean covering up problems and failings. They are careful in their words to others because they assume others see the world in this way. In team environments, differences in comfort with negative statements can be a damaging barrier. Those with root-cause Attribution view feedback as a contri- bution to improvement. Team members may ask others how they are doing and sometimes even elicit negative reports for their own benefit. This ordinary process does not translate well for those with Systems Attribution. It may be construed as an indictment of the person not an appraisal of specific capacity. This can disturb teamwork in multinational groups. Differentiation vs. Dialectical Reasoning. Natural settings are often complex with multiple and sometimes incompatible goals. This leaves practitioners with tough choices. There are national differences in reasoning about such contradictions (Peng & Nisbett, 1999). Consistent with the Greek and Roman tra- dition, differentiation reasoners work to understand contradictions by separating, 268 HELEN ALTMAN KLEIN analyzing, and evaluating distinct qualities. They sharpen distinctions by high- lighting the strengths and weaknesses of each view. Polarization of contradictory perspectives is viewed as exposing the root cause or best option. There may be secondary causes or options, but the focus is on the most important. Dialectical reasoners evaluate ideas by seeking their connections rather than sharpening distinctions. They avoid conflict (Chu, Spires, & Sueyoshi, 1999) and believe that different perspectives may contain truth. Consistent with Eastern philosophic tradition, dialectical reasoners seek harmonious, intermediate positions, deny dichotomous descriptions, and retain elements of different perspectives (Peng & Nisbett, 1999, 2000). They see differentiation as closing out options. Maintenance personnel feel pressure to keep equipment in the air at the same time that they feel pressure to ensure the safety of every aircraft. A pilot strives to arrive on time at the same time that he or she is committed to arriving safely. Why choose between being safe and being on time? Do you weigh alternatives and choose or seek an integration of all goals? This dimension can hinder team interactions during ongoing activities. Differ- entiation reasoners view each person as potentially having strong and weak areas of performance. They isolate one characteristic from the person as a whole and so do not view criticism on one characteristic as criticism of the person. They value frank and even critical analysis of individual performance. Dialectical reasoners consider the person as a whole. For a dialectical person, critical comments about individual characteristics are considered to be intimidating and demeaning to the whole person. They avoid direct criticism and may, when working with Westerners, even hide errors to avoid it.

IMPLICATIONS OF THE CULTURAL LENS MODEL: FIVE INTERCULTURAL CHALLENGES

Practitioners cannot simply apply research findings gleaned from Western research and expect help in multinational environments (Klein, in press). The complex cognition required in natural settings can require that practitioners adapt equip- ment, procedures, and instruction to the cognitive characteristics of multinational participants. Further, practitioners cannot simply borrow research findings from controlled laboratory settings and expect help in natural settings. Five cognitive challenges – problem definition, planning, prediction, coordination, and training – reflect the complexity that distinguishes natural settings from traditional labo- ratory paradigms. I now review these intercultural challenges and suggest their vulnerabilities to the national differences outlined above. While I review these five sequentially, they are often undertaken iteratively or in parallel. Cognition in Natural Settings 269

Problem Definition

Problem definition includes both problem detection and sensemaking. In labo- ratory paradigms, problem detection is usually unnecessary. The experimenter specifies the problem. In natural domains, such as medicine, weather forecasting, and intelligence analysis, the practitioner encounters an unusual pattern. A patient mentions occasional abdominal pain, dizziness, and thirst; blood pressure is elevated but blood sugar average; the physical exam reveals unusual neurological responses. There may be a lot of information but no clear pattern. Making sense out of an array of anomalies requires the active integration of elements. The practitioner must organize the information into a meaningful story. During difficult incidents, people need expertise to appreciate the cues and patterns (Klein, Pliske, Crandall & Woods, 1999). They need to continually reframe their interpretation (Klein, Wolf, Militello & Zsambok, 1995). The physician, for example, may order additional tests. This is an active view of problem detection rather than an accumulation of discrepancies until a threshold is passed. Weick (1995) described how individuals and organizations come to understand confusing events as a process of sensemaking. Practitioners must construct an ex- planation for anomalies and use that construction to define what counts as relevant data. This is not a mechanically generated interpretation. It requires the use of inferences to achieve interpretation. This active view of problem definition means that it is vulnerable to differences in cognition. This vulnerability is described for several dimensions below. Tolerance for Uncertainty influences the threshold for initially reacting to anomalies. Problem detection is most likely to occur when people are prepared to reframe their understanding of the situation (Klein et al., 1999). Those who are high in Tolerance for Uncertainty change more easily and select a new ex- planation with less data. A person with a low Tolerance for Uncertainty is more reluctant to change and more likely to wait for more information. Each pattern carries its own strength and weakness. People who are high in Tolerance for Un- certainty are comfortable reframing situational understanding but slow to settle on a final interpretation. People with low Tolerance for Uncertainty are more complete in their review of information but tend to stick with an interpretation once accepted. Those with a sense of mastery assume that they can make sense of an ambiguous situation and that there is a solution to detected problems. They initiate an active process of discovery and aggressively construct explanations. In contrast, a fatal- istic orientation leads to a passive process of problem detection and sensemaking. Those with hypothetical reasoning typically use this as a mechanism for sensemaking. Mental simulations allow them to evaluate their efforts to organize 270 HELEN ALTMAN KLEIN confusing events. Concrete reasoners, in contrast, would seek comparable cases for sensemaking. This approach would generate a different outcome. Attribution is expected to contribute to the problem detected. For the person with root cause Attribution, individual contributions are very salient. For a person with systems Attribution, situation contribution would be expected to have a low threshold for detection while individual contributions may be outside the radar. Differences in Attribution lead to the identification of different problems. This is important because problem identification directs the search for solutions. Differentiation and dialectical reasoners are expected to differ in their sens- making. Differential reasoners polarize potential explanations and seek the best explanation. People who show dialectical reasoning do not try to fit options into categories. Imposing abstract category structures is seen as distortion. They work to fit all data into a coherent picture. Differentiation vs. Dialectical Reasoning influences the explanations considered and the criteria for accepting them. Whereas, differentiation reasoners think of problems as anomalies suddenly arising and easily solved, dialectical reasoners see problems as long present, but only recently noticed. They do not seek a “quick fix” but rather the creation of conditions needed for a return to balance.

Planning

I use “planning” to include initial planning, ongoing adaptive replanning, and the decisions these entail. Traditional laboratory research often assumed that people first generate a comprehensive set of options. Consistent with a multi-attribute utility analysis, they then detail and evaluate the options using a common set of criteria. In natural settings, experienced planners must often go beyond the mech- anisms studied in laboratories (Lipshitz, 1993; Schmitt & Klein, 1996). They may lack the time and resources needed to generate a set of options and then compare each to a criterion. Experienced planners may use past experience to generate a plan as they make sense of the situation (Klein, 1998). Alternately, they may use a constructive process to recognize and synthesize potential leverage points. This strategy uses mental simulation to test and revise the plan. Both strategies meet the stressful planning demands of some natural settings. As planning demands increase, so does vulnerability to national differences. Four dimensions illustrate this vulnerability. The nature of plans is dependent on the Time Horizon of the planner. Those with a future horizon make plans that consider longer term consequences and goals more than those with a present horizon. Their mental simulation strategy extends further in time. Cognition in Natural Settings 271

Tolerance for Uncertainty influences planning. Planning often presupposes an assessment of risk related both to the situation at hand and the proposed remedy. Both judgments depend on the Tolerance of Uncertainty. Those who are low in Tolerance for Uncertainty want more information before formulating plans. They generate detailed, fixed plans and may be reluctant to engage in improvisation and adaptation. They may see people with high Tolerance for Uncertainty as impulsive. People with high Tolerance for Uncertainty are comfortable initiating plans with incomplete information and adapting their plans as additional information becomes available. People with low Tolerance for Uncertainty are uncomfortable with this seemingly casual planning. Differences in this dimension are even seen among Westerners as is clear in an interview with a U.S. military officer.

The British plan and plan and plan ...Let’s say you need to move troops quickly. The British would have all these maps and all this stuff before they’d do anything. Americans would start packing. We’ll plan the best we can and we’ll kinda shoot from the hip, because it’s time sensitive. The British would rather be late. They want every kind of weird contingency laid out, like what happens if we have real severe weather? In a military operation you’ve got to leave some stuff to chance, and the Brits don’t like to leave anything to chance (Klein Associates, 2002, p. 12).

The clash between high and low Tolerance for Uncertainty is common in multi- national operations. There is conflict between members of national groups who want everything clearly and firmly pinned down and those who grow impatient with what they see as “micromanagement.” Planning is vulnerable to differences in Hypothetical vs. Concrete reasoning. Concrete reasoners develop plans by identifying the precedent case that provides the best match. The result would be a time-tested plan rather than an innovative plan. Hypothetical reasoners see planning as a constructive activity of recognizing and synthesizing leverage points (Klein, 1998). Hypothesis testing rather than con- crete comparisons is required and mental simulation forms the basis of evaluation. Their plans would be innovative at the expense of safety. Attribution can contribute to differences in planning. Those with root-cause Attribution seek causal factors in individual characteristics. Thus plans must address individual characteristics. Training is seen as correcting individual limita- tions. Those with system Attribution address plans to the broad environment con- sidering organizational and systems changes. While those with systems attribution may see the focused plans of root-cause thinkers as shallow, root-cause thinkers may view the plans of the systems approach thinkers as unfocused and illogical. These differences create contention and mistrust during collaboration. The differences, however, have the potential for enriching the planning process by incorporating the strength of divergent cognitive approaches. 272 HELEN ALTMAN KLEIN

Coordination

While individuals are typically the unit of analysis in laboratory research, the complex tasks found in natural settings can require more people and a broader range of skills and competencies. During surgery, the physician focuses on a particular procedure. At the same time, the anesthesiologist monitors the patient to detect and manage problems in life functions. A nurse ensures that equipment is available when and where it is needed. Each professional needs different information and has different responsibilities. Coordination is essential because of the high-pressure, interdependent demands of surgery. Team members must monitor each other, along with the overall team status, to make needed adaptations. Even space is limited and team members need to anticipate the physical location of others. These dynamics are not readily captured in the laboratory. National differences emerge as a source of friction in multinational teams (Thomas, 1999). Teams depend on a shared vision of the world. Team members need to have common expectation for how information will be shared and how de- cisions will be made. Research with multinational teams suggests the importance of culture for teamwork effectiveness (Adler, 1991, 1997; Granrose & Oskamp, 1997; Helmreich & Merritt, 1998; Lane & DiStefano, 1992). Fatalism vs. Mastery influences how team members react in the face of threats. Those with a sense of mastery will actively identify and modify barriers. In contrast, fatalism may lead to seeking adaptations in the face of perceived limitations. Those with a sense of mastery may view fatalism as defeatist and ineffectual. Those with a sense of fatalism may view mastery as unrealistic and fanciful. Team members who rely on hypothetical reasoning use divergent thinking and option generation at the expense of concrete, context-sensitive solutions. They expect others on their team to do the same. In contrast, those with concrete rea- soning might show a preponderance of precedent-based solutions at the expense of flexible solutions. They might find the hypothetical thinking of their peers to be groundless. This difference interferes with a coordinated team planning. Differences in Attribution can contribute to conflict. Faced with a problem to solve, a person with root-cause attribution will look for the responsible person and characteristic. Those with a systems approach Attribution will look to the broader context of the problem. A team will struggle to solve a problem when they lack a common understanding of the nature of the problem. Differences in the use of criticism can also hinder coordination. Those with root-cause Attribution encourage critical analysis of ideas at the same time that those with a systems Attribution find this criticism demeaning. This is not an easy difference to bridge. Variations in Differentiation vs. Dialectical Reasoning present a parallel obstacle. Differentiation reasoners approach cognitive tasks by identifying Cognition in Natural Settings 273 and prioritizing options. They would expect others to make discriminations and carefully contrast options. The dialectical reasoners would try to integrate information into a comprehensive picture. Differentiation reasoners might accuse their teammates of being indecisive, “Why can’t they settle on the problem so we can solve it?” Their dialectical teammates would accuse them of being narrow and limited, “Why this rush to judgment? Real problems are never that simple.” Power Distance is a last example of the pervasive importance of national differ- ences during coordination. Teams in low Power Distance nations are egalitarian with members free to contribute their own ideas and critique the ideas of oth- ers. High power distance is associated with a firm command structure with team members expecting and wanting direction. Multinational military operations are plagued by differences in Power Distance. A U.S. officer reported, “They won’t do anything unless I tell them exactly what to do and how to do it.” A man under his command sees it differently, “He is my officer. He must make the decisions. I can’t do that.” Teams whose members differ on any of the dimensions may waste time and resources struggling with national differences. These barriers to coordination de- tract from their real tasks and compromise team effectiveness. When people have a shared vision, this supports teamwork. The shared national vision may also lead to a preferred pattern of teamwork. This analysis of within-nation similarities and between-nation differences questions the universality of current teamwork mod- els. It suggests that the optimal functioning of a team will depend on national differences.

Prediction

Predicting the plans and actions of team members is vital for coordination. It is also important to predict the decision and actions of allies and partners, international users of technology, and adversaries. When people come from the same national group, they are better able to predict judgments, decisions, and limitations. This mutual knowledge, or common ground, allows the detection and repair of commu- nication breakdowns (Clark & Brennan, 1991). In the Sioux City landing incident, the pilots were American and so likely to share values on the dimensions. This allowed them to anticipate actions, frame effective communication, develop and adapt plans, and provide useful feedback. This incident would have been even more difficult with a multinational flight crew. The dimensions that provide common ground within a national group can introduce mismatches between national groups. I review three natural settings to illustrate the role of the cognitive dimensions. 274 HELEN ALTMAN KLEIN

Air traffic controllers are most effective when they can adapt their directions to the cognitive characteristics of a flight crew. Crews from a high Tolerance for Uncertainty nation, for example, may be slow to acknowledge problems. Air traffic controllers may need to be more attuned to other signs of trouble and more assertive in their directives. During emergencies, crews from high Power Distance nations will respond to directives differently than those from low Power Distance nations. High Power Distance crews may respond best to authoritarian directions while low Power Distance crews will work better with egalitarian interactions. Air traffic controllers need to anticipate that pilots from fatalistic nations may be less aggressive in seeking solutions than would a crew from a mastery nation. Technological innovations intended for international markets should provide the design features preferred by people in those markets. When the cognitive assumptions of the equipment and instruction match the cognitive characteristics of users, demand for an innovation is maximized. Early in the design process, these cognitive features must be predicted. An understanding of national charac- teristics can support predictions. Nations with future Time Horizon, for example, select technology with an eye to long-term integration. They are concerned with the continued availability components and technical support. Tolerance for Uncertainty can help predict the desired timing and specificity of feedback provided by the equipment. Errors in predicting needs and potential mismatches can undermine technology dissemination no matter how well the technology is designed for the originating nation. National security depends on anticipating the actions of adversaries. Military leaders need to predict the decision making of hostile armies as well as that of paramilitary and terrorist groups. Adversaries high in Power Distance are expected to act quickly and decisively in emergencies but fall apart when their central command or leadership is disrupted. Low Power Distance nations or groups may be initially slower to react to surprises but suffer less when communication is disrupted. An adversary with hypothetical reasoning can adapt more readily to unexpected events but will lack the precision of an adversary more dependent on concrete reasoning. Diplomats and military leaders need to predict the impact of their messages. An assertion intended as hypothetical and conciliatory may be interpreted as concrete and provocative. An adversary with a mastery orientation is expected to be more vulnerable to different persuasion information than one with a fatalistic orientation. A pamphlet warning about the massive force might firm the resolve of a mastery adversary but destroy the will of a fatalistic nation. A statement meant as describing a long-term goal might be interpreted as an immediate threat. Increased knowledge of cognition dimensions may improve efforts to predict actions, disrupting factors, and vulnerabilities. Cognition in Natural Settings 275

Training for Multinational Effectiveness

Training is important in two distinct ways. First, ongoing training is often critical for competent performance in natural settings. Training is needed to accommodate the rapidly changing technology of software, medical devices, aviation, commu- nication, and weaponry. When technology is used internationally, instructional methods and materials are best if they reflect the cognitive characteristics of intended users. Several dimensions illustrate potential interference with existing training approaches. Those groups low in Tolerance for Uncertainty are uncomfortable with self- paced and flexibly structured training. It is a poor match with their need for structure and definition. Their feeling may be, “If this instructor knows what he’s doing, why am I supposed to decide how to complete this course?” Instructors need to bridge this mismatch if they intend to use self-paced material. Alternately, knowledge of national differences can direct the selection of a more appropriate learning tool. High Power Distance groups are most comfortable with formal teaching styles. An instructor reported, “I’d always had students who questioned and pushed me. I can say anything to these guys and they just write it down.” Some instructors have developed pre-training exercises designed to encourage more interactive learning. In the same way, low Power Distance professionals find it demeaning to have instruction delivered in a way they view as authoritarian. They expect to have their ideas heard. Good training may be lost when it does not match the user. Much scenario-based training depends on working through simulated problems. The trainee may be encouraged to think through alternative solutions and speculate on likely outcomes. One goal is to foster flexible thinking. While this is effective with hypothetical reasoners, it may be ineffectual and baffling for concrete reasoners. They typically learn by examining past cases to understand actual outcomes. Scenario-based training may seem like “ungrounded mind-games.” The second importance of training is as a tool for overcoming barriers seen during international collaboration and teamwork. The Cultural Lens Model can provide the perspectives of other groups. It can provide a “lens” to adjust for mismatches in cognition and in relevant behavioral and social dimensions. This may help multinational collaborators and team members identify, appreciate, and manage national variations. It might also convey an understanding of the strengths and limitations of allies, potential product users, and adversaries. Team members need to identify and accommodate mismatches between their own cognition and that of others on their team. The Cultural Lens Model has guided the development, implementation, and evaluation of a scenario-based 276 HELEN ALTMAN KLEIN training program (Klein & Steele-Johnson, 2002). Tolerance for Uncertainty and Analytic vs. Holistic reasoning were targeted for training. Analytic vs. Holistic Reasoning includes Attribution as well as Differentiation vs. Dialectical Reasoning. Training for each dimension lasted less than two hours and included definitions, examples from international settings, and role-playing exercises. There were increases in knowledge, positive attitude, and accuracy of model-based predictions. Based on this initial success, the training concept was then used to develop materials multinational peacekeepers to be used prior to deployment. Initial trials at the Stabilization Force (SFOR) Headquarters in Bosnia supported the efficacy of the approach (Hahn, Harris, & Klein, 2002). The training focused on personnel from Western nations. An additional validation study also demonstrated the impact of using a Cultural Lens Model based training program to increase functional understanding (Sutton, 2003). Enthusiasm and demand for multinational training far surpasses current training technology. Practitioners in domains as varied as the military, commercial aviation, medicine, and business want to function more effectively. Much of the available training, however, focuses on behavior and customs. The Cultural Lens Model offers an approach that is grounded in the nature of national differences and is directed to the cognitive demands of natural settings.

Cognitive Psychology Through a Cultural Lens

Practitioners now have the benefit of decades of laboratory research on human cognition. The research has advanced theoretical understanding of human thought processes and functioning. More recently, considerations of cognition in natural settings have extended the science of human cognition far beyond the laboratory. The inclusion of national differences in cognition is an emerging opportunity for cognitive psychology. The majority of research upon which cognitive psychology rests has used subjects from the United States, Western European, and English- speaking nations. Yet national groups show qualitative differences in cognition. As cognitive psychologists better understand these differences, they see the limitations of Western cognitive psychology. It can no longer be assumed that all people think, make decisions, plan, and make sense, in the same way. Human factors and ergonomics professionals must go beyond their own projections about how people from other nations cooperate in work situations, use Western technology, develop new skills, and negotiate. Commercial aviation and other areas of technology transfer are increasingly international efforts. A small number of manufacturers, primarily from the West, Cognition in Natural Settings 277 design and produce commercial aviation equipment for the international mar- ket. The same organizations develop procedures and training programs for both maintenance and operational staff. This concentration has supported industrial growth, safety, and international standardization. It has also meant that cognitive differences across national groups are hidden or easily ignored. As human factors specialists learn to incorporate national differences into the design of equipment and training, the industry will be better able to improve productivity and safety worldwide. Peacekeeping and humanitarian efforts have become increasingly international allowing the pooling of resources and expertise. At the same time, the demands for multinational coordination increase the difficulty of problem definition, planning, and training. Practitioners cannot assume others will share our vision. Multina- tional teams perform best when participants understand how others on their team “see” the situation. This demands common ground and shared understanding – a real challenge for multinational teams. Human factors professionals cannot ignore national differences in cognition. We also cannot ignore the strength that multinational interchanges provide. We will need knowledge to capitalize on what such interchanges can offer. We need tools to avoid the barriers and problems introduced by national differences in cognition. We can better serve the needs of business, aviation, and the military, as well as other natural settings, as we learn to tap the skills and resources of people around the world. The study of national differences in natural settings has opened up a new and exciting window on cognitive processes as they play out in practical, complex, and often demanding domains. As more work domains cross national borders, there will be a greater need for tools to support common ground and shared vision. The Cultural Lens Model can serve as a framework toward this goal.

ACKNOWLEDGMENTS

This chapter was prepared under funding to Klein Associates Inc. from the Defense Advanced Research Projects Agency (Contract DAAH01–00-C-R094 from the U.S. Army Aviation & Missile Command). Some of the examples cited were taken from complementary research efforts for the Army Research Laboratory, Army Research Institute, and The Boeing Company. Support for this research was also provided by Wright State University. I would also like to thank Dr. Gary Klein, Mei-Hua Lin, Amy Meininger, Dr. Randall J. Mumaw, and Dr. Jeffrey Sanchez Burks for their contributions to this manuscript. 278 HELEN ALTMAN KLEIN

REFERENCES

Adler, N. (1991). International dimensions of organizational behavior (2nd ed.). Boston: Kent. Adler, N. (1997). International dimensions of organizational behavior (3rd ed.). Cincinnati, OH: Southwestern College Publishing. Agrawal, A. A. (2002). Phenotypic plasticity in the interaction and evolution of species. Science, 249, 321–326. Bandura, A. (1986). Special foundations of thought and action: A social-cognitive theory. Upper Saddle River, NJ: Prentice-Hall. Berry, J. W. (1986). The comparative study of cognitive abilities: A summary. In: S. E. Newstead, S. H. Irvine & P. L. Dann (Eds), Human Assessment: Cognition and Motivation. Dordrecht, The Netherlands: Martinus Nijhohh. Choi, I., Nisbett, R., & Norenzayan, A. (1999). Causal attribution across cultures: Variation and universality. Psychological Bulletin, 125(1), 47–63. Chu, P., Spires, E., & Sueyoshi, T. (1999). Cross-cultural differences in choice behavior and use of decision aids: A comparison of Japan and the United States. Paper presented at the 2000 Command and Control Research and Technology Symposium, June 26–28, 2000, Naval Postgraduate School, Monterey, CA. Clark, H. H., & Brennan, S. E. (1991). Grounding in communication. In: L. B. Resnick, J. M. Levine & S. D. Teasley (Eds), Perspectives on Socially Shared Cognition. Washington, DC: American Psychological Association. Diamond, J. (1999). Guns, germs, and steel: The fates of human societies. New York: W. W. Norton. Dorfman, P. W., & Howell, J. P. (1988). Dimensions of national culture and effective leadership patterns: Hofstede revisited. Advances in International Comparative Management, 3, 127–150. Granrose, C., & Oskamp, S. (1997). Cross-cultural work groups. Thousand Oaks: Sage. Hahn, B. B., Harris, S., & Klein, H. A. (2002). Exploring the impact of cultural differences on multi- national operations (Final Report under Prime Contract #DAAD19–01-C-0065, Subcontract No. 8005.004.02 for U.S. Army Research Laboratory). Fairborn, OH: Klein Associates. Hall, E., & Hall, J. (1990). Understanding cultural differences. Yarmouth, ME: Intercultural Press. Helmreich, R. L., & Merritt, A. C. (1998). Culture at work in aviation and medicine: National, organizational, and professional influences. Aldershot, ENG; Brookfield, VT: Ashgate. Hofstede, G. (1980). Culture’s consequences: International differences in work-related values. Newbury Park, CA: Sage. Hong, Y., Morris, M., Chiu, C., & Benet-Martinez, V. (2000). Multicultural minds: A dynamic constructivist approach to culture and cognition. American Psychologist, 55(7), 709–720. Ji, L. J., Peng, K., & Nisbett, R. E. (2000). Culture, control, and perception of relationships in the environment. Journal of Personality and Social Psychology, 78(5), 943–955. Klein Associates (2001). Finding answers amidst uncertainty: Cognitive challenges of counter-terrorist analysts (Semi-Annual Report sponsored by Defense Advanced Research Projects Agency is- sued by U.S. Army Aviation and Missile Command under Contract act #DAAH01–00-C-R094). Fairborn, OH: Klein Associates. Klein Associates (2002). Cross-cultural story/incident repository. Internal working document,. Fairborn, OH: Klein Associates. Klein, G. (1998). Sources of power: How people make decisions. Cambridge, MA: MIT Press. Klein, G., Pliske, R. M., Crandall, B., & Woods, D. (1999). Features of problem detection. Proceedings of the Human Factors and Ergonomics Society 43rd Annual Meeting, 1, 133–137. Cognition in Natural Settings 279

Klein, G., Wolf, S., Militello, L., & Zsambok, C. (1995). Characteristics of skilled option generation in chess. Organizational Behavior and Human Decision Processes, 62(1), 63–69. Klein, H. A. (in press). Cultural differences in cognition: Barriers in multinational collaborations. In: B. Brehmer, R. Lipshitz & H. Montgomery (Eds), How do Professionals Make Decisions. Mahwah, NJ: Lawrence Erlbaum. Klein, H. A., Klein, G., & Mumaw, R. J. (2001). Culturally sensitive aviation demands (Tech- nical Report Prepared for the Boeing Company under General Consultant Agreement 6–1111–10A-0112). Fairborn, OH: Wright State University. Klein, H. A., & Steele-Johnson, D. (2002). Training cultural decentering (Final Report for Order No. 84334, Department of the Interior/Minerals Management Service). Fairborn, OH: Klein Associates. Kluckhohn, F., & Strodtbeck, F. L. (1961). Variations in value orientations. Evanston, IL: Row, Peterson. Lane, H., & DiStefano, J. (1992). International management behavior: From policy to practice. Boston: PWS-Kent. Lane, H., DiStefano, J., & Maznevski, M. (1996). International management behavior (3rd ed.). Cambridge, MA: Blackwell. Lipshitz, R. (1993). Converging themes in the study of decision making in realistic settings. In: G. A. Klein, J. Orasanu, R. Calderwood & C. E. Zsambok (Eds), Decision Making in Action: Models and Methods (pp. 103–137). Norwood, NJ: Ablex. Markus, H., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98(2), 224–253. Mishra, R. C., & Tripathi, N. (1996). Reproduction of patterns in relation to children’s weaving experiences. In: J. Pandey, D. Sinha & D. P. S. Bhawuk (Eds), Asian Contributions to Cross-Cultural Psychology (pp. 138–150). New Delhi: Sage. Morris, M., & Peng, K. (1994). Culture and cause: American and Chinese attributions for social and physical events. Journal of Personality & Social Psychology, 67(6), 949–971. Norenzayan, A., & Nisbett, R. (2000). Culture and causal cognition. Current Directions in Psychological Science, 9(4), 132–135. Orasanu, J., & Connolly, T. (1993). The reinvention of decision making. In: G. A. Klein & J. Orasanu, R. Calderwood & C. E. Zsambok (Eds), Decision Making in Action: Models and Methods (pp. 3–20). Norwood, NJ: Ablex. Peng, K., & Nisbett, R. (1999). Culture, dialectics, and reasoning about contradiction. American Psychologist, 54, 741–754. Peng, K., & Nisbett, R. (2000). Dialectical responses to questions about dialectical thinking. American Psychologist, 55(9), 1065–1067. Roese, N. (1999). Counterfactual thinking and decision making. Psychonomic Bulletin and Review, 6(4), 570–578. Rogoff, B., & Lave, J. (1984). Everyday cognition: Its development in social context. Cambridge, MA: Harvard University Press. Ross, K. G., McHugh, A., Moon, B. M., Klein, G., Armstrong, A. A., & Rall, E. (2002). High-level cognitive processes in field research (Year One Final Report under Contract 02TA2-SP1-RT1 for U.S. Army Research Laboratory under Cooperative Agreement DAAD19–01–2–0009). Fairborn, OH: Klein Associates. Schmitt, J. F., & Klein, G. (1996). Fighting in the fog: Dealing with battlefield uncertainty. Marine Corps Gazette, 80, 62–69. 280 HELEN ALTMAN KLEIN

Schweder, R., & Bourne, E. J. (1992). Does the concept of the person vary cross-culturally? In: A. J. Marsella & G. M. White (Eds), Cultural Conceptions of Mental Health and Therapy (pp. 97–137). Dordrecht: Reidel. Segall, M. H., Dasen, P. R., Berry, J. W., & Poortinga, Y. H. (1990). Human behavior on global perspective. New York: Pergamon Press. Sutton, J. L. (2003). Validation of cultural awareness training concept. Unpublished manuscript, Army Research Laboratory. Tetlock, P. (1998). Close-call counterfactuals and belief-system defenses: I was not almost wrong but I was almost right. Journal of Personality and Social Psychology, 75(3), 639–652. Thomas, D. (1999). Cultural diversity and work group effectiveness: An experimental study. Journal of Cross-Cultural Psychology, 30, 242–263. Triandis, H. C. (1994). Culture and social behavior. New York: McGraw-Hill. Wegner, D. (1987). Transactive memory: A contemporary analysis of group mind. In: B. Mullen & G. R. Goethals (Eds), Theories of Group Behavior (pp. 185–208). New York: Springer-Verlag. Weick, K. (1995). Sensemaking in organizations. Thousand Oaks, CA: Sage. Wright, G., & Phillips, L. (1980). Cultural variation in probabilistic thinking: Alternative ways of dealing with uncertainty. International Journal of Psychology, 15, 230–257. Zsambok, C. E., & Klein, G. (Eds) (1997). Naturalistic decision making. Mahwah, NJ: Lawrence Erlbaum. 10. DESIGNING THE USER EXPERIENCE FOR INTERNATIONAL WEB USERS

Monica Chong

ABSTRACT

How do you efficiently design a global yet local user experience for Web sites? Arguably, the user-centered design approach has been one of the best methods in designing a successful user experience for Web services in the initial market, but why isn’t this process applied to international markets? This chapter makes a case for applying a user-centered design process to the international expansion of Web sites and discusses issues impacting the creation of a successful user experience for local audiences. Although this chapter primarily focuses on designing large scale Web services, many of the principles can be applied to any sites that undergo internationalization.

INTRODUCTION

Designing for the Internet presents a great challenge for product designers: how to efficiently create a Web service that attracts and motivates a large global audience. The Internet allows users from around the world to interact outside the traditional boundaries of time and place by making information and applications available on Web sites. This global quality appeals not only to those who want to stay connected wherever they go, but also to those who once depended on slower

Cultural Ergonomics Advances in Human Performance and Cognitive Engineering Research, Volume 4, 281–316 © 2004 Published by Elsevier Ltd. ISSN: 1479-3601/doi:10.1016/S1479-3601(03)04010-4 281 282 MONICA CHONG and costlier forms of communication such as wire, mail and telephone services. At its best, the Internet offers a cosmopolitan forum for social exchange, making new forms of interaction possible, and bringing long-distance communication within the reach of those who otherwise would have done without. But there is another equally important aspect to the Internet’s appeal: through Web sites, it offers the ability to access information and interact in ways that are highly local in form, function and style. This quality is grounded both in the “right-here, right-now” logic of consumer convenience and in a preference for a look, feel and “flavor” that are specific to user groups in particular locales. Well-executed Web services offer us the chance to avoid the line at our bank, find out what’s playing at our local movie house, or discover a new restaurant before everyone else does. Whether seen from a global or local perspective, the appeal of online services is related to the way they fit – or fail to fit – into their users’ lives. There are millions of Web sites to choose from, so how do users choose one? The design of Web services that are of real use to users and that can be understood and used by those for whom they are intended are certainly more likely to be chosen than those that don’t. Those that go beyond these minimum requirements to offer a user experience that is engaging, satisfying or fun will have even better chances for success. The ability to offer a simultaneously global and local experience on the Internet is at the heart of one of the most important business challenges facing Web designers: how to efficiently expand a locally successful Web service across boundaries of nation, language and culture. Although such an undertaking may seem simple, I will argue that successfully creating products and services for a global market is not for the faint of heart. Building and maintaining multi-market Web services can become very complex, time consuming and expensive. Errors in the initial phases can prove costly later on. If not created properly, a Web page may require duplicate development effort every time it is translated into a new language. Extrapolating this effort to the broader set of tasks required for serving multi-nation markets, it is not difficult to imagine how mistakes made early on in the international expansion of a product can grow exponentially with each additional version. This might well lead to the early demise of the product or even a financial disaster for the company that created it. Here, I will argue that user-centered design provides a way to overcome the challenges of designing Web services for a global market. The key to the successful development of a user experience for multi-national markets lies in the careful management of the creative process from the very beginning, avoiding the need to adapt or retrofit a Web product or service for export to a new country. I will also advocate the use of user centered design and development methods throughout this development process. Designing the User Experience for International Web Users 283

This may strike readers as a bit of a truism. After all, the efficacy of user-centered design techniques seems abundantly clear from the recent business history of the Internet. Countless ventures floundered precisely because they failed to consider the usability or the usefulness of their service in the lives of intended users. But how can this method be applied to the development of a product or service intended for use in multiple regions separated by boundaries of language, politics and culture? This is a timely question; we know that some companies are losing millions of dollars on their international Web services. For organizations that have committed to an international presence on the Web, the secret ingredient to managing the development of a service is the intention that they be international from the start.

WHY FOCUS ON USER EXPERIENCE?

A good user experience is the key to a popular and successful Web service. Before discussing how this concept applies to online services for international markets, it is worth defining what user experience is, and describing its relationship to business operations. User experience is an imprecise but nonetheless valuable concept. In this chapter, my use of this term is informed by my experience as a practitioner designing Web-based user experience for Yahoo!’s global services in twenty-three markets. In the broadest and simplest of terms, user experience refers to matching users’ needs and expectations to key elements and features of product design. At a minimum, the service must be useful, usable and desirable to offer a positive user experience. The usefulness of a service offering may be measured by how relevant it is to what the user wants to do. For example, if a user wants to learn about investing in a particular company on a Web site about personal finance, a useful user experience would provide enough relevant information about the company for the user to make an informed decision. Information such as current share price, company his- tory, background and professional ratings may be available. Relevant information is not only useful, but of value to the user. An even better experience might allow the user to choose from a range of stock exchanges in multiple countries, allow her to track the value of investments linked to multiple currencies and provide stock information in other languages she knows. An excellent user experience surpasses expectations and provides motivation such that a user might want to bookmark the site as one of the best sources of investing information on the Web. User experience comprises many interrelated factors. One of these is the usability of a Web service, which is important in helping the user feel comfortable and confident. If the user on the investment Web site described above is able to find a current share price, but isn’t able to quickly discern the currency in which 284 MONICA CHONG its value is listed, she might lose confidence in her ability to make sense of the information offered. Ultimately, she might abandon the service for one offering fewer features presented with greater clarity. Problems like this one are far from hypothetical. A report by a major research firm published in 2001 suggested that Web retailers were overlooking an easy way of increasing revenue from their Web sites: fixing usability issues that cause users to abandon the check-out payment process. The same report found that one sports retailer shaved 20% off of its customer support budget simply by correcting major usability problems on its Web site.1 Investing in the usability and usefulness of commercial Web sites clearly impacts the bottom line profitability of any such Internet venture. Although less easily quantifiable, the desirability of a Web site is an indicator of how closely the user identifies with the “personality” or branding of that site. “Design look” factors such as layout, color palette, graphics and the overall look and feel of a site impact the bottom line of Web-based businesses. These factors help determine not just brand perception, but also credibility, an intangible quality directly related to consumer acceptance. In a recent study of factors influencing the credibility of Web sites, 46.1% of respondents cited one or more of these “design look” factors in their assessments. As the authors stated: “[the idea that] looking good is being good also seems to hold true for evaluating the credibility of Web sites, especially since design look is so noticeable.”2 While the desirability of a Web service is determined by many factors, one of the most business-critical is its cred- ibility, the perception among its users that the company behind it is competent and capable of offering a given service in a manner that makes it competitive with others in its class. Branding is another important factor in determining the overall user experience. The Web site’s brand image affects how users choose to engage with the online service. For example, in the offline world, financial firms try to project a particular brand image by highlighting past successes and showing a carefully cultivated image of trustworthiness. These rules of engagement apply in the online world as well, and similar marketing tactics need to be communicated through the interface. Color, style, editorial voice, advertising message, and reputation all impact how users perceive the brand image of an online service. A financial service firm might choose a muted, conservative color palette that doesn’t call too much attention to itself for its Web site. In contrast, an online travel service selling holiday packages to Hawaii might want to have a bright, colorful Web site with appealing animation to motivate users to go on vacation and to help engage users in the fantasy of a luxury resort holiday. This same design look would not communicate the serious, trustworthy image sought by a financial services firm. Similarly, specific aspects of the look, feel and behavior directly contribute to brand experience, determining how users are likely to feel and think of and engage with a given online service. Designing the User Experience for International Web Users 285

More successful services don’t stop at offering a user experience that is useful, usable and desirable; they also motivate users by fostering a positive emotional response. In this instance, user experience refers to the broader set of feelings and thoughts that adhere to the use of a service over time, to a user’s willingness to describe or recommend it to others, and to the ability to imagine novel uses for it. For instance, a single person may not necessarily be motivated to look for a potential date by simply going to a search engine and typing in “date”; positive feelings are unlikely to develop for someone chosen at random. Most people would need more motivation and assurance to go on a series of blind dates. They might need to get over the social stigma of meeting strangers online. In this case, a marketing message that positions the Web service to be fun, socially mainstream, and successful in finding a potential partner needs to be communicated in order to motivate the user. A positive user experience provides a powerful marketing benefit to an online service. The Internet is perhaps the medium best suited to word-of-mouth referrals. When users have a good experience on a Web site, they frequently share that positive feeling with others. Word-of-mouth referrals for Web services can come from a variety of sources: professional and non-professional reviews, links from another Web site or the traditional friends-and-family networks. Virtually unknown just four years ago, Google gained its popularity almost entirely through word-of-mouth from satisfied users. The message a potential user hears in such a situation is a testament to the site’s value and meaningfulness. A good user experience can also generate invaluable marketing buzz. In 2001, BMW launched a marketing Web site that featured the manufacturer’s automobiles as the “star” in a series of short films along side A-list celebrities. These short films were directed by top filmmakers who were given budgets of $1 million and had to make a specific BMW car “star” in the film. BMW films were so popular and generated so much buzz in the press that Claudia Mueller, International Manager for Planning and Communication for BMW, estimates they generated $26 million dollars worth of marketing press value in 2001. She attributes the success to the “deep brand experience”: users watched the short film an average of four times, and almost all had recommended it to friends.3 It should be noted that word of mouth could also work the other way around when users tell each other about their frustration or negative perceptions of a Web service. Word of mouth is a powerful, low cost vehicle for communicating the value of a product to users, one that is particularly well suited to the Internet as a medium. From the very beginning, designers with the marketing department should set clear targets for the kind of impression they intend to make on their users and ensure that these objectives are compatible not just with their company’s business objectives, but with the nature of the medium. 286 MONICA CHONG

A good user experience depends on having a Web site that is useful, usable and desirable. When successful in combining these elements, the Web sites bring higher customer satisfaction, enhance the reputation and credibility of a company brand, as well as directly and indirectly improve the bottom line by facilitating marketing efforts and reducing customer service expenses. Achieving a good user experience is a challenge in a single market where issues of multiple languages, legal systems and conventions for representing things, such as currency, addresses and measures are not present. In developing online services for global markets, these challenges are compounded by varying perceptions of a company’s image, implying for location specific brand strategies to communicate local consumers’ concerns about Web services.

MEETING USER EXPECTATIONS WITH A USER-CENTERED DESIGN APPROACH

User-centered design methods help ensure that the overall user experience is useful, usable and desirable. These methods base product requirements around users’ behavior and environments and needs rather than on business objectives or the availability of specific technologies. This does not mean that business objectives or technological questions should be ignored or considered as factors of secondary importance. A careful balancing of these factors is required to ensure that users’ needs drive the development of a product or service that makes business sense and is technologically sound. User-centered design methods involve users throughout the development pro- cess. In an iterative design process, feedback from users is obtained in each major step of the design cycle, ensuring that a service does not lose its focus on the end user, even in the development of the business model. Defining business objectives, conducting research to establish user requirements, analyzing competitor services, and creating an initial site architecture are among the first steps in this process pre- ceding the production of early versions of the service. Mapping the user tasks (activities a user will do), prototyping the interface, usability testing and assessing the overall performance of the service, result in a second set of tasks in which users will interact with and provide feedback for versions of the online service. Specific steps of the design cycle may be repeated as necessary to enhance the overall performance of the product or service. In fact, iterative design presumes multiple rounds of design and testing in all phases of the product development process. User-centered methods are commonly employed in the development of Web sites intended for a user base in a single country or region. But how might this process work in the development of services intended to reach audiences Designing the User Experience for International Web Users 287 in multiple countries? Before describing how this process might work for multi-national projects, I’ll review some of the basic terms used to describe international product expansion.

BASICS OF INTERNATIONAL WEB SERVICE EXPANSION

When a service expands across national borders and is made available to new audiences defined by country or language, it undergoes a process of interna- tionalization. This process of adaptation can be difficult and complex because it generally involves the retrofitting of a Web site to meet requirements or needs that were not present in the original market. This retrofitting or international “izing” of a Web service is problematic not just because it involves adaptations to unforeseen user and business requirements outside its original context, but also because many of its most common challenges can be anticipated and avoided if the original design process is approached carefully. Internationalization is often a very resource-inefficient process, either because user needs or requirements vary significantly from one context to another, or because technological issues demand that the design be significantly modified prior to re-deployment in the new context. The time, expense and complexity of adapting a design for use in a new market should not be underestimated; unanticipated differences between markets may mean that the adaptation may be as or even more costly than the original development process. This additional time and expense is not always necessary, and would appear to negate many of the benefits of choosing the World Wide Web as a medium for offering a product or service globally. After all, why choose a global medium to deploy a design that is limited to one locale? There is a better way. The design of a product or service intended for use in markets outside the country in which it is initially developed can be approached through the development of a global design program. This requires that inter- national user needs be taken into consideration from the start. While it may be difficult or even impossible to know in advance what the exact needs of local users might be, this approach anticipates country-specific adaptations by dividing the design cycle into two distinct steps: globalization and localization. The term globalization has many connotations, depending on economic, cultural and political context. For the purposes of a global design program, globalization refers to the process of designing a service to be successfully deployed in multiple markets in such a way that fundamental changes to the design do not have to be made. The goal is to achieve a universal design that can be adapted through relatively small modifications in each market. The core of the design has to be based upon user needs and requirements that are essentially similar across 288 MONICA CHONG

Fig. 1. Ovens that Illustrate a Global vs. Local Design Intent. multiple countries, markets, languages and cultures. Locale-specific adaptations can be allowed for through flexibility in the visual design and by employing the concept of modularity in the design of interaction flows that vary from one place to the next. This concept is familiar to many manufacturers who make products for export. In Fig. 1, the two oven controls illustrate the difference that anticipating local adaptations in advance – through the establishment of a global design program – can make on a finished product. Both ovens have the same core function of cooking food, but the one on the top has controls that were clearly designed for use in multiple markets. The function of the controls is communicated through universal symbols rather than the oven on the bottom that has explanatory text. The bottom oven controls require far greater effort to be adapted for export to specific Designing the User Experience for International Web Users 289 markets that vary from the original based on language or the use of metric vs. Fahrenheit units to describe cooking temperature. These suggestions are especially problematic from the perspective of a global design program. The oven on the bottom, in addition to using text and units of temperature not used in most countries, presumes this oven will be used for cooking meat. Two out of three of these dials provide suggested settings for how to cook pork, veal and beef. Imagine how a person who doesn’t eat meat or whose culture considers pigs as an un-edible animal, or cows sacred, will be motivated to choose this oven model as a cooking method. As it applies to the user-experience design, globalization refers to an approach that, in the long run, will reduce production costs and improve overall efficiency. But this approach may not be enough to appeal to the needs or desires of local users, or to communicate as effectively in the new market as in the original. With language complexities and differences in local user requirements, design modifications will likely be necessary to improve the overall performance of the product. The difference between this approach and the traditional process of internationalization is that, in a global design program, such modifications are an- ticipated from the beginning: executing design changes is then much simpler and resource efficient. Tailoring a product to appeal to the needs and desires of local users, and to communicate with these users effectively, is called localization. The term locale is used because localization may refer to different market segments within a single country, although here it refers to adaptations made at the country level. In a sense, the oven controls on the bottom in Fig. 1 may be well localized for a specific market, even though they are poorly adapted for global consumption.

Translation Does Not Equal Localization

One of the first factors to consider when setting up a global design program is language, which may require translation based on usage and custom (e.g. Spanish in Spain vs. Mexico or Argentina). But there are other factors that may create deeper differences related to cultural identity. When designing a user experience for audiences in multiple countries, it may not be enough to simply translate the text that comes along with the service. How much of the local culture needs to be included to make a Web service successful? Nancy Hoft uses an iceberg metaphor (Fig. 2) to illustrate degrees of localness.4 Deeper levels of localization move beyond translation to include cultural and social aspects of the locale where the online service will be launched. When a service is being transformed for multiple markets, there are visible – or surface – issues for localization, such as text that requires translation or unit 290 MONICA CHONG

Fig. 2. The Ice Berg Model. Suggestions of Globalization and Localization Efforts Mapped onto Hoft’s The Iceberg Model (Hoft: International Technical Communication). measurements that must be converted. Once this is accomplished, the product has been generally localized. But this may not be enough if the service still doesn’t fit into local users’ expectations or understandings about what it should or might do, and how the service might carry out the steps necessary to accomplish the objective. A more adequate localization would enable users to feel that the service was intended for them, even though they are very distant (culturally or geographically) from the place where the product was originally developed. To achieve full local acceptance and motivate users, the service may need to be modified so that local users feel that it was made uniquely or especially for them. In this instance, the user experience may need to be radically localized to incorporate unspoken or even subconscious characteristics or rules that are specific to a country or market segment (see at the bottom half of the Iceberg in Fig. 2). It takes into consideration how users from a given locale think, feel and act in their contextual environment. For example, designing a user experience for a Web site for matching potential life partners demands high levels of sensitivity to local social behavior and cultural rules. In comparing Hindi Matrimony, a popular online Indian matchmaking service and Udate, a popular U.S.-based online dating site, it is evident that the sites serve the needs of two very different user groups, Designing the User Experience for International Web Users 291 which vary not only in the way they use the services, but what the users may expect for a positive outcome of that use. In India, where traditional matchmaking – including arranged marriages – prevails over Western style dating, those posting the ads are not necessarily the ones for whom the match is intended; ad posters may be siblings or a parent or other relative. In the profiles of potential matches, educational qualifications, technical skills, jobs, family heritage, or whether the match prefers “traditional” or “modern” values is carefully detailed. One description posted in mid-2003 read like a job resume, mentioning proficiency working with Windows 2000 and Office 98 as potential “selling” points for the potential match. But this site is not about user’s software skills or areas of study in any simple way. By describing the potential matches in this way, the ad posters are describing a position within complex matrices of social class and values, signaling to those with similar backgrounds and compatible families that the person in ques- tion might be a good fit. Although references to traditional systems for classifying an individual’s position within hierarchies of social class and religion are notably absent, ads on this site clearly try to describe the potential match’s social position and values through a vocabulary reminiscent of job resumes and family heritage. On the U.S.-based Udate site, overt classification of the ad poster’s social class and values is also avoided in favor of indirect – but still clear – descriptions of education, hobbies, taste in consumer goods and occupation, all of which have class connotations. The romantic ideal of finding one’s “soul mate” discourages users from overt – and distinctly unromantic – discussion of social position or values. Udate.com uses a different way to encourage users to describe their social class and values that is more appealing and motivating to Western audiences, for whom matchmaking services are often seen as a “last resort” and are potentially stigmatizing. Users can enter details to describe their leisure activities (such as preferred publications, sports, hobbies, favorite cuisines), personal traits (such as habits), appearances (such as sense of humor, will power, fashion sense), and physical descriptions, with this information meant to convey essential information about the person posting the ad. Details about salary and job position are brief and optional. In addition, Udate also offers the possibility for those posting ads to say how sexually adventurous they are. This design feature is absent in Hindi Matrimony, because it may offend users’ sense of public decency, and because sharing personal affinities – including sexual affinities – may not necessarily be seen as preconditions for a successful relationship. Such information fits modern Western ideas that condition romantic relationships – including marriage – on notions of personal affinity and intimacy rather than on social similarity. While both sites use English language interfaces, the vocabulary used to describe criteria for matching is clearly very different. These differences in vo- cabulary, visual design and even in the identities of primary and secondary users 292 MONICA CHONG map onto much deeper differences in the nature of long-term relationships such as marriage. In the West, the success of romantic relationships depends upon the establishment of intimacies that, in turn, can only be developed through personal affinities like hobbies, aesthetic styles, consumption habits and sex. Unlike users of Udate, ad posters on Hindi Matrimony may expect that a successful marriage is predicated on notions of social relatedness, and that love might be a product of such a union rather than the condition upon which it is based. For online matchmaking services, the ability to describe oneself (or a relative in the case Hindi Matrimony) in a flattering manner and to find relevant postings from others who meet specific social class and value criteria are key factors in determining the quality of the overall user experience. When stripped down to the technical bare bones, both services are basically Web sites that search and match a database of users’ profile information; it is the visual and editorial surface treatment that makes them radically different and highly local experiences. In the scenario of one company with services in both India and the U.S., the company could extend their service to international markets efficiently by re-using the same technology to search and match databases of user information. But ultimately, if the users aren’t encouraged to submit details about themselves that others will find interesting, it won’t be a positive user experience for anyone. It is the role of the user-experience design team to understand and reflect what users in each culture group expect and need from the service. In this case, radical localization of the presentation and preferences would be necessary for business success. Collectively, these differences between countries or markets (which may be bounded by culture, language or geography) can be thought of as international variables. Broadly, there may be surface differences, such as text or unit measures, or relatively more complex issues of style, custom or history. These variables might be sorted into two types: those whose impact should shape the global design of the user experience, and those whose impact can be accounted for at the local level. Local knowledge of user requirements and expectations is the only reliable way to document and plan for these international variables. As in any other application of user-centered design methods, user research is integral to the development and implementation of a global design program.

BUSINESS ISSUES CONCERNING THE USER CENTERED DESIGN PROCESS FOR INTERNATIONAL MARKETS

If user-centered design techniques are recognized as effective in building successful Web services for single country markets, why isn’t this approach extended to international markets? Ultimately, the business objective behind any Designing the User Experience for International Web Users 293 development process is to be as successful in the most resource-efficient manner possible. Initially, business constraints such as time to market and production cost pressures may dictate the path of international expansion. While these initial constraints tend to take center stage early on in a company’s expansion, it’s important to remember that local user requirements will have to be addressed and the overall user experience will have to be properly designed if the Web service is to be popular and financially successful. User-centered design and international expansion can be time consuming. Time-to-market pressures may contribute to a sense that a surface-level or general localization of the product will have to suffice. Between 1999 and 2000, Yahoo! rapidly expanded its operations outside the United States, bringing 15 separate product lines to over 20 different markets. At the time, there weren’t many established competitors outside of the U.S., so time-to-market pressures were prioritized in order to establish local brands. In the given time frame for launching Web services, it was impossible to fully understand all international user requirements. A few years later, some of the services are still successful with only a general localization treatment. However, as local markets matured and stronger competitors appeared, the need for more highly localized products also emerged. In December of 2001, Yahoo! acquired Kimo, the leading Taiwanese portal at the time. Focus groups indicated that some users preferred Kimo to Yahoo! Users perceived Kimo as a friendly, local service, whereas they saw Yahoo! as a strong but distant and foreign competitor. The integration of the two sites produced a new, radically localized service that combines the local qualities of Kimo with the global user experience that Yahoo! offers. While acquiring a local company to offer a local feeling site may not be an option, many companies, working with locally based teams or contacts and applying a user-centered design process, will help to overcome social and cultural problems in the design of products by identifying the international user requirements. Later in the chapter, this process will be discussed.

COSTS AND RESOURCES

The perceived cost of incorporating user-centered design methods into product development is often a barrier to getting international feedback. While cutting these costs may provide some short term advantages, in the long run, the Web ser- vice will become vulnerable to local competitors who know their users better and can offer a better user experience for that market. International business ventures can be expensive and risky, and the increased vulnerability to local competition is in itself a risk that should be managed. Another reason that user-centered design is often not used in international development is that it adds complexity to the overall 294 MONICA CHONG design process. The iterative design methods is complex and can be difficult to manage even for a single market; incorporating feedback from users at each phase of the design process makes initial production slower than it would otherwise be. Managing this process in multiple countries means allowing additional time for the complexity of working in multiple locations. In evaluating the relative importance of these factors, recall the overall goal: the production of a universal design system that will, in the long run, be more resource efficient than the alternative process of retrofitting and adaptation. If local user expectations are met, customer support and other related expenses can be reduced, and overall profitability increased. Furthermore, it is significantly more expensive in the long run to retrofit a Web service design and backend technology than collecting feedback internationally up front and anticipating and planning appropriately for a Global Program. Time and cost saving tips for collecting international user feedback are mentioned later in this chapter. Feedback from international users adds significant value to the business venture in the mid-to long term, and contributes to user perception of the product and the brand behind it. As Nadine Kano, of the Globalization Team at Microsoft explains:

[International feedback helps us] maximize the quality of all language editions of its product while minimizing the amount of time and money invested. The only way to achieve this is to have written specifications – before any coding work begins – that take internationalization issues into consideration. If you treat international concerns as just an afterthought, your final product will reflect that and your target market will definitely notice.

While the idea of adapting or retrofitting existing designs may appeal for short-term business reasons, it is important to understand the implications of such a decision. This will cause the overall product quality to suffer in the process of deployment in new markets and may cause complications – such as deterioration in brand perception – if not planned for up front.

COMMITTING TO AN INTERNATIONAL BUSINESS

Once the financial and time commitment to an international Web service offering has been made, the design process must be set up and managed in a way that reflects this commitment. The staff at eBay changed the design process to reflect their international business objectives and their company’s commitment to meeting them. Justin Miller (2003), Director of User Experience at eBay, explains how the popular online global auction network made the business commitment, and then altered the user experience design process: Designing the User Experience for International Web Users 295

EBay’s goal has always been to provide a consistent and easy to use experience to all of our users, while allowing for local differences. The methodology has evolved over the past several years as our international business has grown. Two years ago, the primary focus for design was the U.S. market since a vast majority of our users used eBay.com. Local adaptations were implemented after the fact. As our international business began to take off, it became clear that we should design for all of our local sites from the beginning of every project. Today, whenever we design a new feature or redesign an existing page, localization issues are taken into account in the very earliest stages.

Once a business commitment to expand internationally is made, planning for the design of a global program can start. First, an infrastructure must be established to support the development and maintenance of the Web site over time. This involves planning to ensure that people with the right kinds of expertise are present in the appropriate locations. Second, a plan must be developed for managing the relatively more complex design process that internationalization involves. Finally, all team members must be aware of and in agreement about the principles guiding the design and development of a product or service for multiple markets. Only when these requirements have been met can a user-centric process of design for international markets begin.

STAFFING AND RESOURCES FOR A GLOBAL DESIGN PROGRAM

Creating a global program based on a use-entered design process is a multinational and multi-disciplinary effort that requires good communication channels and clear processes to work efficiently. Ideally, it is a collaborative effort between the core design team located in the primary design center and group of counterparts in the region or country where the products are to be offered. The core team is responsible for the overall design of the global product experience while counterparts are responsible for the local user experience. The multi-disciplinary teams consist of: global project management, multi- cultural expertise, interaction, visual design, user research, and web development. The global project manager oversees the global schedule, mediates and manages the relationship between the core and local design teams. User researchers use qualitative and quantitative methods to collect user requirements and ensure the usability of international products. International user research is conducted in the location of the target users to understand their behavior and contextual environment. Once requirements are collected, interaction designers translate them into user tasks or descriptions of what activities a user will be able to do on the site. Then, in the form of design specifications, they are then mapped 296 MONICA CHONG out onto the overall global program site architecture and page templates. The interaction designers collaborate with visual designers to create the overall look and feel that is both usable and enhances the brand personality. The international design specialist advises on general international issues and potential cultural problems throughout this process. Web developers create rapid prototypes from the interaction and visual designers, and can help to prototype designs quickly in multiple international versions to be able to evaluate international problems for technical implementation. Depending on the complexity of the localization issues and organizational structure, multi-disciplinary local teams act as counterparts for gathering and defining user requirements, and designing specific features. This takes advantage of their familiarity with local language and culture. While having local design teams is not always feasible, especially for small companies, a local project contact is valuable in creating a more localized experience by either contributing design solutions or raising local cultural concerns.

DEVELOPMENT TEAM STRUCTURES

Once the local and domestic teams are in place, a team structure needs to be determined to define the roles and responsibilities. The user experience design team structure can be closely tied to the global software system structure. Erran Carmel (1999) describes four different structures and corresponding information system approaches. The structure of the user experience design team (Fig. 3) should be similar to that of the development teams. Depending on the globalization and localization requirements, as well as resource limitations, an appropriate structure should be chosen. There are advantages and disadvantages to each. A centralized structure (in Fig. 3) is efficient in rapidly deploying products and ideal for situations where locally based expertise is scarce or non-existent. It is also the least expensive in terms of resources. Centralization of the design team works best when the product will be generally localized and fairly consistent for all international versions. To ensure a globally consistent user experience, eBay chose a centralized structure, consulting local teams to ensure that localization issues are adequately anticipated. Justin Miller (2003) explains why eBay chose a centralized structure:

Because we want to provide a consistent experience to all of our users regardless of which eBay site they use, we centrally manage design from our San Jose [California] headquarters. To make sure the designs fit local needs, in-country teams are consulted for localization issues. The mix may depend on the project, but there’s a constant dialogue between the two teams. Designing the User Experience for International Web Users 297

Fig. 3. Global Corporate Structures and Corresponding Information Systems Approaches. (From Global Software Teams by Erran Carmel, 1999.)

In contrast, a decentralized (in Fig. 3) development structure produces a highly local feeling to the user experience because local teams have greater ownership and autonomy in the design process. This strategy is sometimes chosen for rapid deployment of services through third party licensing and partnering agreements. In other cases, business objectives may vary so significantly between markets that a single, unified design becomes impossible. The main disadvantage of this model is that design work and other resources are not necessarily being re-used. Also they require experienced duplicate design team skill sets in each development center. Decentralization also poses a relatively greater challenge from a management perspective, since the overall consistency of the product and the documentation process become more difficult to oversee. As local teams gain experience, a coordinated federation or transnational design team structures become possible. In many cases, these are the most useful of management structures because they offer the best possible combination of global efficiency and local expertise. Local teams benefit from the efficiency and consistency of a centralized staff who, in turn, benefit from the local cultural knowledge of decentralized teams. In addition to greater employee satisfaction for local teams, the quality of user experience is improved through the multiple checks that these decentralized models provide. To ensure that good ideas and innovation move freely across national borders, good communication and trust must be established. These are the hallmarks of a mature global team structure. 298 MONICA CHONG

Global team structures take time to learn and work efficiently, but they may change over time. Trust between the team members and clarity about roles and responsibilities will help the members of such teams grow professionally. A newly formed global team may choose to start with a centralized structure to maximize the technology of the primary development center. Over time, if user-centered design groups in local development centers are large enough and sufficiently experienced, the design of some product features may be led by teams outside the primary or original market. For example, Yahoo! Mail, a Web based e-mail program, was primarily developed in the U.S. headquarters, where the design team was accustomed to leading the development of new features. Yahoo! needed to add a feature enabling users to send text messages to mobile phones from their computer desktops. Since text messaging is currently more popular in Europe than in the U.S. – from both a business and a user perspective – this feature was designed and tested by the European User Experience Design team within the documented style guides set by the core design team in the U.S. This is an example of how the local expertise and experience of a multi-centered design (transnational) team can contribute to a global program. This distributed devel- opment model, in which a local team led the design of a modular feature to add functionality to a core service, is only possible where local design teams have had the time and the opportunity to mature as part of a multi-nationally based group.

PRINCIPLES FOR DEVELOPING A GLOBAL PROGRAM

Three general principles will help international organizations run more easily and efficiently: a scalable infrastructure, cultural awareness and sensitivity for local markets, and establishing clear communication channels.

Scalability

An infrastructure that scales globally sets a solid foundation for international ex- pansion of a service while keeping costs down. A scalable design solution is one that plans and calculates for growth, expansion and maintenance. Often, when the design of a user experience is developed around requirements for one market, difficulties occur with the addition of features for international versions, which cannot be simply or easily adapted from existing versions. For example, the icons for Yahoo!’s international Web mail product (Fig. 4) illustrate what happens when internationalization is not taken into account as one Designing the User Experience for International Web Users 299

Fig. 4. Yahoo! Mail Icons International Designs (1997–2002). Reproduced with permis- sion of Yahoo Inc. ∗2004 by Yahoo! Inc. YAHOO! and the YAHOO! logo are trademarks of Yahoo Inc. of the requirements for a global product. When Yahoo! Mail was designed in 1997, the original U.S. mailbox icon (top) had a strong visual affinity with the American suburban mailbox. With the launch of the first international versions, localized mailboxes were designed for the U.K., French, and German markets. Although it seemed like it was a culturally appropriate thing to do at the time, this local solution was very time consuming and not scalable for a global product. As more interna- tional versions became available, a more universal icon was designed for worldwide versions (right), facilitating the re-use of work done by the global design team. Greater standardization facilitates communication and design work while reducing costs by reusing design and technology. For example, the online advertising industry is trying to standardize ad formats sizes amongst different online publishers. Because these formats are currently unregulated, there are hundreds of ad sizes, and ad agencies have to redo their creative work to fit many publishers’ sites. Standardization is a scalable approach, because it promotes reuse and avoids a lot of unnecessary rework.

Cultural Awareness and Sensitivity

Promoting cultural awareness and sensitivity within an organization will help an- ticipate international user requirements, facilitate work with colleagues, clients and partners abroad, as well as improve internal communication. Understanding cul- tural patterns will help foresee behaviors, attitudes, and beliefs in places other than those with which we are personally familiar. Cultural awareness and sensitivity will also help avoid inaccurate stereotyping. There are many helpful resources to help build cultural awareness and sensitivity; one is Kiss, Bow, or Shake Hands,an overview of social behaviors and business conventions for 60 countries (Morrison, Conaway & Border, 1994). 300 MONICA CHONG

Clear Communication Channels

Clear communication channels and procedures between global teams are crucial to the success of any international business venture. In globally coordinated projects, teams rely heavily on remote communication methods such as email and telephone conference calls; most contact and almost all “meetings” will occur through the use of the phone and/or a computer. However, remote communication still leaves a lot of room for misunderstandings and does not always help build good rapport within the teams. It is recommended that the global teams meet at least once in person at the beginning of a project, and ideally periodically as work progresses, to build strong lateral team relationships. During these face-to-face meetings, they need to agree on a corporate team structure, workflow process, responsibilities, and communication channels, such as weekly updates. Good communication channels will pave the way for a much smoother project execution and help to avoid cross-cultural conflict.

APPLYING THE USER CENTERED DESIGN PROCESS TO A GLOBAL PROGRAM

Feedback from users at all phases of the development cycle helps ensure that the product maintains a focus on real users’ needs throughout the design process. It is important that this process meet the primary business objectives of international expansion. To do so, users from the markets in which the service is to be deployed must be involved in the design process. Understanding the contextual usage of the local audience, identifying user requirements and usability problems will certainly have an impact on the acceptance of the service in local regions. This section describes how the international users can be involved during the design cycle.

STAKEHOLDERS’ MEETING: ESTABLISHING BUSINESS REQUIREMENTS AND GOALS

Before work begins, the design team needs to understand the global business goals and research about user behavior as it relates to the product or service. The project usually begins with the design team meeting with global and local stakeholders, those responsible for the business and product development decisions and overall performance, to agree on the business goals for the final product. It is important Designing the User Experience for International Web Users 301 that the core design team understand how business goals and strategies vary in each market. Different regions have specific business and economic conditions as well as competitive climates that may affect design decisions and lead to new strategies or a new take on the service. During this phase, it’s best for the stakeholders to define measurable perfor- mance goals for the user experience design team. By establishing criteria up front, the design teams will be able to later gage success of the international versions, and will be better able to see where and how variations in the performance of the design occurred. Lastly, unless all markets are to be treated as equal, there should be up front agreement as to which markets are a business priority. This will help prioritize design work and user research. The target markets can be categorized into primary, secondary and tertiary priorities based on the business opportunity (the conditions, size of market, demographics, business partnerships, revenue projections, etc.), and the order of international releases (different language versions and releases may be staggered). This may affect the priori- tization of product design and allow more time to conduct research in different countries.

USER RESEARCH

After the business goals are defined, user research can be conducted to gather information needed to design the online service. Generally, the first phase of re- search concentrates on gathering and determining user requirements: what makes a site useful, usable and desirable for people in a given locale. Once initial product specifications have been determined, user research should also be conducted iteratively throughout the design cycle to ensure a design that is easy to use and that the user requirements are met. User research methods may include focus groups, surveys, ethnographic research, usability tests and the gathering of statistics about user patterns. This research can be executed internally or outsourced to a third party if resources permit or require. Some methods are easier to implement internationally than others. For instance, surveys and phone interviews can be done remotely, while user observation for focus groups and ethnographic research is best conducted in the cultural context of the user. If the product design teams par- ticipate and witness first hand user behavior of users in their local context, it makes more of an impact on those responsible for the design and better use of resources is achieved. If there are many international versions to be launched, it is unlikely that complete user research will be able to be conducted in every market all at once 302 MONICA CHONG in the first release; focusing on specific priority markets and issues is crucial. Language, travel, and lack of familiarity with the local environment extend the time frame for conducting research. The sheer volume of data that multi-country research generates will add complexity to the analysis, management and imple- mentation of findings. Despite these challenges, there are alternative strategies that can be incorporated before, during, and after the design development process to significantly reduce costs and time as well as the problems of complexity and distance. In addition to focusing user research on priority markets from a business per- spective, choosing markets with extreme cultural behaviors, attitudes and values, will produce findings as to how the user experience may need to be designed and be used under very different circumstances. In 1999, user researcher, Pia Harold, used this approach when she conducted research for Siemans comparing how users in the Germans and Chinese markets learn to use mobile phones. Her hypothesis, based on published cultural theory, was that these two countries had two distinct learning traditions: rote learning through imitation common in Asian cultures vs. learning through understanding in German schools.5 By conducting research in these two extreme circumstances, she was able to gather knowledge that could be potentially applied to international versions with similar learning traditions. International research is a continuous process; in some cases, research should be conducted in select markets between product releases. During July 2002, the Yahoo! Europe design team was required to design a multi-market service and run three European country user studies in a short time frame. To meet the project requirements throughout the design cycle, the design was tested and analyzed in each city. The findings of user problems in the study were quickly analyzed and implemented into the prototype, then reiterated in time for testing in the next city. While the pace was a little hectic, the method was successful in not only finding usability issues, but also in discovering cultural and local user requirements at the same time. While ideall, user research is conducted in every locale, this may not be a reality for most companies with time, budget, and resource limitations. These savings can be made by focusing user research in priority markets, and by choosing locales that may have extreme cultural differences, the design can be “stressed tested” under different circumstances. In addition, the time frame for an iterative user centered design process can be shorted with a mobile design team. For international markets that do not fall under the category of business priority or cultural extremes, less expensive and time consuming remote user research methods may be used (if appropriate), such as, phone interviews and surveys, to at least gage local user satisfaction. Designing the User Experience for International Web Users 303

Fig. 5. The Beijing Edition of Sina.com News Page (2002).

COMPETITIVE ANALYSIS OF USER INTERFACES

After gathering user requirements and business goals, a competitive analysis of the user interface should be done to benchmark the level of user experience in the market. Doing this for international Web products is a good way to understand how a local user may use a product or interact on the Web, since often, Web services for single markets are based on a primary market user’s needs and contextual environment. The product design needs to be examined to see if interaction conventions are the same in target markets. Introducing a foreign convention to another culture may confuse local users and or increase frustration. For example, in Fig. 5, in Mainland China, studies have indicated that users prefer to have separate windows that open from the original page. The pages are very dense with content and take a long time to load for users who want to quickly read information while waiting for other pages to download. Interaction elements as well as visual elements should be reviewed. In this way, the visual design of local competitors will help set the benchmark for style, colors, and brand identity for the Web products.

CONCEPTUAL DESIGN: DEFINING USER TASKS AND SITE ARCHITECTURE

After gathering user requirements from the target markets, interaction designers turn them into user task flow or site architecture for the overall global program. The user requirements are translated into user tasks that describe the necessary activities a user will need to be able to do to meet their needs and expectations. The tasks then need to be separated into core tasks to be included in all international 304 MONICA CHONG versions, and tasks that are international variables for specific local editions. These tasks will later become the basis for usability tests. The user tasks are then structured into the overall site architecture for the global design. Good site architecture is scalable, flexible, and modular so it can support the core functionalities and the international variables and be easily managed. For example, a large scale Web site with hundreds of individually designed Web pages (this happens when sites grow organically over time), would be virtually impossible to internationalize as is. A design approach where every page has to be designed is difficult to improve and manage. Instead, it is much easier to design a template sys- tem in which specific functions correspond to page types, which can then be reused. Even though Web page templates may be specified by type and converted into templates, some content and functions may have to be different from one market to another. It is best if the tasks, pages, and content are designed modularly. In Fig. 6, the global program has a core set of functions, and modular international variables that may be included or excluded in different market versions. For example, market 1 may have the core features and functions a, b, d, f; while market 2 may only have the core functions set and features a and f. It is important to identify these variables and design the platform to include this flexibility. To take a practical example, for an online auction site like eBay, core tasks may include browsing and posting auction items. Registration for this service may have to vary from one market to another and would require specific international versions. To establish and maintain credibility, an auction site like eBay must be able to verify the identity of its users; this helps establish trust between buyer and

Fig. 6. Modular Design Approach. Designing the User Experience for International Web Users 305

Fig. 7. The German and Spanish Versions of Yahoo! Games (2003). Take in same permission as on p. 297 seller. The conventions for doing so will vary from one country to the next. In the US, credit cards are a very common means for identifying individuals. However, in countries where credit cards are less prevalent, such as Taiwan and some European countries, a different site flow process is required. Registration may ask for a user’s national identification number, mobile or landline phone number. In some markets, users may be asked to complete registration by post. From the perspective of a global design program, the best way to go about managing these differences is to design the overall task flow into modules, which can then be switched out in accordance with the needs of the local market. The modularity of the design then ensures the most resource-efficient means of localizing a global product. In addition to taking a modular approach to the design, the global site architec- ture should take advantage of the communication potential of the Internet. Since having a global presence can be an advantage over a single country competitor, the site architecture might address how users interact, transact, and communicate with each other. Once built, global Web sites have a unique platform through which users of similar interests may communicate, or a platform for communicating informa- tion from the site to a broader, global audience. For example in Yahoo! Games in Fig. 7, a German user is playing checkers with a Spanish user (right). The users can interact while having the comfort of interfaces in their respective language.

UNIT CONVENTIONS

Part of making a usable site is presenting information in formats users can understand. Like language translation, unit conventions must be a part of 306 MONICA CHONG every localization plan. Although recent economic and trade globalization has encouraged standardization, there are still significant variations. For example, the United Kingdom officially uses the metric system to conform to European Union standards, but people commonly describe their height in feet and their weight in stones. There are many unit conversions to be accounted for, such as units of measurement, dates, calendars, time zones, currencies, number displays, name and address forms, etc. They need to be double-checked; international unit conventions, conversions and standards are published in various ISO publications. Lack of clarity in using unit conventions may cause user confusion. For example, NASA lost a $125 million Mars orbiter because a Lockheed Martin engineering team used old English Imperial units of measurement while the agency’s team used the more conventional metric system for a key spacecraft operation”6 While most usability problems with unit conventions are not as dramatic as the loss of a NASA orbiter, appropriate standard unit conventions and norms should appear properly on each version of an international version. It’s not just a matter of switching the format in the display, the design should also consider how the unit conventions may aid the user, however. In the example in Fig. 8, the eBay Australia design considers the perspective of all international users of their global design. Because this item is available on different eBay international English versions, they have built-in unit conversions to aid the local user. The currency values have been converted, and the time the auction will close is clearly listed in the user’s local time zone. This way, the specific configuration of unit display (currency,

Fig. 8. Example of a Foreign Item Being Auctioned on eBay Australia (2003). Designing the User Experience for International Web Users 307 time) reduces user confusion when international buyers and sellers interact on the eBay platform. By the end of the conceptual design phase, detailed product design specifica- tions that include all international user requirements are mapped onto a global site architecture, which describes how the user will use the site page by page. Once all the stakeholders agree on the product specifications, then the prototyping phase can begin.

Prototyping

Prototyping is the translation of user requirements into a functional set of interface design elements. Bringing the visual design and interaction flow together through connected Web pages, prototypes model how the site will look and behave before resources are committed to building in full functionality. At this point, users can evaluate the overall design, and changes to the visual design and interaction flow can be made relatively easily, without costly re-engineering of the technical back end. Although there are different prototyping methods, one of the best involves developing multiple designs (i.e. alternatives that meet the requirements) and then testing these in more than one international version. This way, errors can be corrected before coding begins and before the design system becomes too complete or elaborate to permit quick and easy changes. For example, testing multiple international versions will help designers see that an alphabetical index won’t work in Chinese (a language without an alphabet), obliging the design team to offer a specific localization strategy or change the overall design. This testing will also allow designers to see whether content will fit in international sites; compared to English language versions, French, Spanish and German may require additional space to accommodate syntax or word length. In addition to these basic functionality tests, experts or users from the target audience in the country where it will be launched should review prototypes. The visual design of a Web service should reflect its brand objectives while allowing for the possibility that local users may find very different elements appeal- ing. If developed without due consideration to local users’ preferences, the design may communicate the wrong message about the service. To manage the visual de- sign for international versions efficiently, a design system should include modular elements that can be easily swapped for different versions. For example, instead of choosing a fixed color scheme and a set of icons that can only work together, vary- ing “skins” or “themes” might be developed. These will allow for surface changes in the design system without impacting the interaction flow or the technical back 308 MONICA CHONG end. In general, visual design choices should be universally appropriate while still communicating as clearly as possible. The prototyping phase is an excellent opportunity to avoid many common and potentially costly pitfalls related to visual design. Redoing images for each market – translating text, managing quality, reviews and corrections – is time consuming and expensive. Plan on using imagery in a way that is scalable while meeting design goals. This will help reduce the repetition of tasks. Since managing different looks for each market is labor intensive, it is important to have user feedback – including on business objectives – before alterations are made for each market. Once this has been done, new visual styles can be created and tested against specific brand attributes and user goals. While there is no substitute for in-market user and expert feedback, knowledge about cultural differ- ences can help develop a visual design system that will be culturally appropriate. For example, a visual design should try to avoid un-scaleable imagery like the individual mailbox icons for the Yahoo! Mail in Fig. 4. If possible, separate text from images and culturally biased symbols such as flags, geographic landmarks, and historical references that only natives of a given country would understand. In addition, attention needs to be paid to country-specific representations of ethnicity and local graphic trends and styles. When a Web site prototype is complete, the visual and interaction design elements come together. The design prototype models how the product will look and behave when finished; at this point it is ready to be further evaluated for international issues and undergo a final user test before implementation.

EVALUATION

Whether a product is being internationalized or a global program is being designed from scratch, it is important to re-evaluate the prototypes for international issues prior to implementation. Ideally, domestic and global teams will do this collabora- tively with the help of an international design specialist. Remember that just being a native of the country or a speaker of the language in which the product is to be deployed does not necessarily qualify someone to articulate local or global require- ments. Evaluating products requires knowledge of specific cultural and technical issues. Although local contacts may be of great assistance, a multicultural expert will ensure a smoother deployment. Critical areas for review include visual design, site architecture, editorial voice, content, unit conventions, and technical limita- tions. Other factors outside the responsibility of user experience teams also need to be evaluated at this time: e.g. translation and legal issues. When writing text for international versions of a web site, translations need to be handled carefully in the Designing the User Experience for International Web Users 309 context of the service. Attention to the translation of tone, idiomatic expressions, colloquialisms, spelling conventions, grammar and punctuation rules, register (e.g. familiar vs. formal), and humor, contribute to localizing the user experience. An incorrect translation can confuse users, or convey the wrong brand image. Legal issues are especially important since they vary from one country to another and users may or may not be aware of them; even so, companies are obliged to comply with laws and regulations whether or not they appear in user feedback; doing otherwise may result in unwanted legal or public relations exposure. Legal issues include taxation, consent and consideration for a contract, accessibility, customs, tolerance to specific kinds of content, and user privacy. Local legal counselors familiar with local laws should review all product design and features prior to launch.

Implementation and Documentation

After the design has been tested and has a satisfactory level of usability in local context, the results of the global design are connected to the backend technology. The details of the design are communicated through clear visual and interaction guidelines and through the prototype. Hopefully, the technical teams will have worked with the design teams to accommodate a modular design approach for creating a deeper local user experience. Frequently, the design is implemented in one language first, and this should be taken as an opportunity to conduct another review before implementing the interfaces in other languages. This way, resources and effort won’t be wasted on things that will have to be changed. In developing front-end technology, the coding should be carried out according to a hierarchy. The result should be a generic master version, which has a hierarchical parent-child relationship with local versions. In this way, updates and changes may be made on a global level, and the design can be reused. This method minimizes the effort required to make changes and especially the need to recode. Documentation about brand, visual design, interaction flow, and findings from user research is essential for communicating how the system should work to cross- functional international teams and to technical teams. Clear visual style guides and interaction documentation help to maximize global resources, ensure consistency, and provide overall rules for how users interact globally. Language complexities together with communication across vast distances (and time zones) may increase the likelihood that design specifications will not be executed uniformly. The documentation guidelines should also be explicit in the ways the local Web sites may vary and how the overall global program design can be localized and maintained by local teams. Decide up front on unambiguous standard guidelines 310 MONICA CHONG to ensure overall consistency; be specific about where and how international variables are to be accommodated.

Follow-up: Reiterate, Improve, Maintain

The user-centered design process is a continual cycle of trial and error, as it is unlikely that the local and domestic versions will be perfect prior to release. Each of the steps in the design process can be reiterated to enhance the user experience. If user research was not possible in the first release; of the Web site, the time in-between international versions is valuable to learn more about local user behavior for future versions. After the services are made available to the public, the performance of the design can be assessed against the metrics set up in the stakeholders meeting, and the user-centered design process may be repeated for continual improvement. At this point in the process, the Web site service should be useful and usable, but may still not be desirable enough to appeal to and motivate users in local markets. While this phase sets the global infrastructure and local capabilities, the Web service may still need to undergo further localization to appeal to and motivate local users. The next section is the second important phase of the global program, localization, which may be included with the initial public release or over time with the Web service maintenance by local design teams.

DEEPENING THE LOCAL USER EXPERIENCE

Imagine that a foreigner moves into a neighborhood. He eats different food, doesn’t speak the local language well, doesn’t share in the community celebrations, and worst of all he doesn’t attempt to integrate with the rest of society. At first the neighbors may try to approach the foreigner out of curiosity, but if it doesn’t make any attempt to learn the local customs and understand the people of the neighborhood he moved into, or contribute to the livelihood of the community, he’s not going to make many friends and be very popular in the long run. This may also be true for Web services trying to enter a foreign market. At first a foreign service may be a novelty, and people may try it; but, if users are always the ones having to make concessions to bridge communication gaps, they may eventually stop trying. The final version of the Web service reflects what the company has learned about users’ cultural contexts and motivations. In this regard, it is important to remember that the objectives of globalization and those of localization may not Designing the User Experience for International Web Users 311 necessarily or always coincide. While the goal of globalization is to remove locale-specific variables, to create a reusable, modular design, the intent of localization is to reflect in-depth knowledge of local users’ cultural contexts. Sometimes it may make business and design sense to ensure that local versions of a design are highly specific, engaging local users in a manner to which they have become accustomed, even if this is very different to users in other parts of the world. This is the core objective of localization. By making the effort to better localize a Web site, the user experience feels more human.

Decorative Localization

Creating a distinctly local user experience doesn’t have to involve drastic alteration of a backend system. The local Web users just need to feel that the Web site is accommodating them, not the other way around. The Web site needs to find a way to communicate on a common ground. To do this, the visual content and the editorial voice are probably the most important localization elements. One way to find a common ground is to examine common experiences that bring people together emotionally and (or) physically. This is reflected in their attitudes, values and norms. Common local symbols, motifs, colors and imagery, when used with the proper knowledge of their meanings, can be easily added to a Web site to permit a closer identification with local culture. Rituals, festivals, and holidays are an excellent way to express and celebrate a culture’s values. Outwardly, they highlight what is important to society and bring groups closer. Look for opportunities in the interface to add this local flavor. For example, during Chinese New Year, Yahoo! Kimo changes the “theme” of the homepage; the colors change to red for good luck, the graphics are decorated with traditional Chinese symbols, such as fireworks, to ward off evil, red envelopes contain the traditional money gift for prosperity and luck and the animal of the year; this immediately signals to the users that the service is in tune with the local community.

TRANSLATION AND CULTIVATING A LOCAL EDITORIAL VOICE

The editorial voice and tone of a Web site define the personality of the brand and may impact how the user feels toward the service. Rather than straightforward translation, an editorial style needs to be cultivated to appeal to local audiences. To get an idea of strong local voice, think of celebrities, heroes, or other public 312 MONICA CHONG personalities who are well liked and unique to a particular locale. To name a few, Jackie Chan is distinctly Chinese in his art, yet his comedy has a universal appeal; Oprah Winfrey may be representative of U.S. African-American women on TV, but many North American women of all races can identify with her approachable down to earth style; and Richard Branson under the Virgin label, is the hero of the British underdog consumer. As individuals, these figures are unique, but they can also be seen as cultural archetypes, because their public personalities reflect the values and attitudes of their fans in their native countries. Decorative localization of the aesthetics and editorial voice will make a big difference in achieving acceptance among local users. If properly planned for in the global design structure, these adaptations are easy to implement and can be maintained by local teams easily.

Radical Localization

Radical localization incorporates the unstated unspoken and unconscious social rules of a culture in the Web service to the extent the service feels as if it were designed and developed by a native team. Recalling the scenario of a matchmaking service that is available in India and the U.S., a service in a given market may require radical localization in some markets to compete with other local service. For a U.S. based company to enter the matchmaking business in India, it would need to radically “rethink” its service approach and requirements to reflect the social behavior and attitudes to fit the expectation of an Indian user. To radically localize this site, the process would have to go beyond creating a local editorial voice and applying decorative imagery; the service needs to allow users the ability to express and describe themselves and their ideal life partner in a vocabulary that is locally relevant and meaningful. The forms about the ad posters descriptions and criteria fields for searching for a potential partner will have to be altered. There may be other cases in which the visual structural page template system needs to be localized beyond decorative “skins” and “themes” to reflect local tastes and cultural context. For example, an examination of material culture (the visual or physical man-made artifacts) in Hong Kong offers valuable insight into local user preferences. Hong Kong offers a visual environment that would be considered, at minimum, very stimulating by most Western standards (Fig. 9). The flashing neon store signage, the advertising on every building surface, the tantalizing smell of food, the chatty salesmen, and the sheer density of people and automobiles on the street produce an environment in which a dizzying number of elements compete for attention. Hong Kong Web sites such as the home page of Sina.com.hk (left) reflect this aspect of local culture with many busy blinking ads Designing the User Experience for International Web Users 313 Causaway Bay in Hong Kong is on the Right (2002), and Sina.co.hk Web Page on the Left (2003). Fig. 9. 314 MONICA CHONG and dense text that moves around. It is little wonder that Hong Kong Web users appear to show a greater tolerance for visual stimulation and are accustomed to what by Western standards might be considered an inordinately large amount of information on a single page. By looking around one’s environment, the local teams may help to design and articulate the need for radical aesthetic localization. Often, radical localization is not attempted because it can be difficult to anticipate this level of localization, and may require drastic structural changes to the core design and backend technology that are both expensive and time consuming. The benefits need to be carefully weighed against the difficulty in managing this process. However, there are two ways to approach radical localization. Utilizing the benefits of a decentralized team structure, local design teams can design and build two separate products with similar functional goals such as matchmaking. The disadvantage from a global business perspective is that the design and technical effort is not reused, and it is difficult to manage quality and enforce global style and interaction guidelines. The second way would be to rebuild the global program design in a modular structure to completely separate the presentation of the interface from the technical functions. Hopefully, the design will not have to involve this costly retrofitting, as the original global service should have designed with enough flexibility in mind to be able to accommodate the radical localization. This is a test of a good global program.

SUMMARY

Returning to the question of how to design a Web service that attracts a large scale international audience, the answer lies in a carefully managed user-centered process that cultivates a useful, usable and desirable user experience from the perspective of users in their local context. As a global medium, the Internet transcends national borders and offers the possibility of a global user experience that includes all “Net citizens” within Web sites that offer a local and “neighborly” feeling. This undertaking for a company wishing to take advantage of this quality of the Internet is not simple however; an organization has to make a firm strategic commitment to offering international Web services and plan accordingly. Although an organization can incrementally increase global and local resources over time to reflect their international commitment, international expansion needs to be planned for upfront if it is to be done efficiently and be successful financially. The Web service needs to be designed to take advantage of the ubiquitous nature of the Internet by planning for a global program that allows for a global design with significant room to localize. The globalization phase of a design project provides for a global infrastructure and efficient scalability across multiple markets. International user requirements are taken into consideration upfront in Designing the User Experience for International Web Users 315 the core design, rather than retrofitting a design optimized for a single market. An international design specialist may be able to foresee many cultural issues and advise a user experience design team, but the bulk of the requirements should come from research about users in each market. The global design of the user experience should encompass the global design features common to all locales as well as local features that can be accommodated through a modular design structure. Localization, the process of tailoring a product to international users, is in essence the user centered design process applied to multiple markets. A user centered design process ensures that a product meets the users’ needs and expectations by iteratively studying and designing for target users’ behavior and context throughout the design cycle. This method ensures that the Web service will be useful, usable and desirable. If the business goal is to provide satisfactory user experience in the primary market, why shouldn’t the business commitment be made to do the same for every market? The difference between applying a user centered design process to a single market and applying the same methodology to the process of international expansion lies in planning, timing and the allocation of resources. Planning a global program that is built modularly to accommodate international variables allows for a deeper and more locally relevant user experience. Flexible, reusable modular design approach not only saves on design and technology resources, it builds a solid platform for global expansion. International versions of a Web site may start off being generally localized due to time constraints and limited resources, but over time, and as more local resources become available, the local team’s objective may tend toward a deeper or more radical localization in order to engage users in a manner that is uniquely appropriate to the local context. Radical localization doesn’t necessarily mean expensive alterations to backend technology. A more resource efficient alternative is to anticipate and plan for local alterations in visual design and interaction flow. The ability to congregate the largest audience in the world on a Web site is truly a unique trait of the Internet as a medium. At its best, it is a powerful channel for access to information, exchange of goods and social interaction. On the Internet, users do not lack options for things to do; they have the power and autonomy to come and go as they please with just a click of the mouse. The only way to attract and retain a large international audience is to respect user choice and provide a product service that is useful, usable and desirable to their lifestyle.

NOTES

1. Souza, Manning, Sonderegger, Roshan and Doresey (2001). 2. Flogg, Marable, Stanford and Tauber (2002). 316 MONICA CHONG

3. Mueller (2003). 4. Hoft (1995). 5. Harold (1999). 6. Lloyd (1999).

ACKNOWLEDGMENTS

The author would like to acknowledge Patrick Larvie, Manager of International User Research at Yahoo!, for his helpful comments and assistance in editing this chapter.

REFERENCES

Carmel, E. (1999). Global software teams: Collaborating across borders and time zones. Upper Saddle River, NJ: Prentice-Hall PTR. Flogg, B. J., Marable, L., Stanford, J., & Tauber, E. (2002, October 29). How do people evaluate a web site’s credibility? (pp. 23–25). Persuasive Technology Lab, Stanford Univer- sity, Consumer WebWatch, Sliced Bread Design, LC. Retrieved May 5, 2003, from http://www.consumerwebwatch.org/news/report3 credibilityresearch/stanfordPTL TOC.htm. Harold, P. (1999, May). Learning how to use a cellular phone: Comparison between German and Chinese users (pp. 196–204). Technical Communication. Hoft, N. (1995). International technical communication: How to export information about high technology. New York, NY: Wiley. Lloyd, R. (1999, September 23). Metric mishap caused loss of NASA orbiter. CNN Web site. Retrieved February 17, 2003, from http://www.cnn.com/TECH/space/9909/30/mars.metric/index.html. Miller, J. (2003, February). Interview with Justin Miller, ebay Director of User Interface Design, by Monica Chong. San Jose and London. Morrison, T., Conaway, W. A., & Borden, G. A. (1994). Kiss, bow or shake hands. Holbrook, MA: Adams Media Corporation. Mueller, C. (2003, May 20). The internet is the ultimate driving machine, BMW Group. Digital Art Works Conference sponsored by Yahoo! Europe and Campaign. London, England. Souza, R., Manning, H., Sonderegger, P., Roshan, S., & Doresey, M. (2001, June). Get ROI from design. 4,7. The Forrester Report [Electronic version]. Cambridge, MA. 11. “A FRENCHMAN, A GERMAN, AND AN ENGLISHMAN ...”: THE IMPACT OF CULTURAL HETEROGENEITY ON TEAMS

Florian Jentsch, Raegan M. Hoeft, Stephen M. Fiore and Clint A. Bowers

ABSTRACT

Most traditional research on work groups has studied groups and teams that are homogeneous with respect to culture. To alleviate the dearth of material on culturally heterogeneous teams, this chapter provides an overview of the impact of cultural diversity on groups and teams in today’s workforce. First, we focus on the problems involved in defining the constructs of “teams” and “culture.” Second, we provide a brief review of the cultural factors that have been identified as affecting human performance. This review serves as the basis for the third section of this chapter, which investigates if – and how – cultural heterogeneity affects team performance. Finally, we conclude with how culturally diverse workplaces can be managed and how to improve performance when faced with cultural diversity.

Cultural Ergonomics Advances in Human Performance and Cognitive Engineering Research, Volume 4, 317–340 Copyright © 2004 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1479-3601/doi:10.1016/S1479-3601(03)04011-6 317 318 FLORIAN JENTSCH ET AL.

INTRODUCTION

Work groups have become the cornerstone of the American industry, with their emergence seen as early as the late 1970s (Cannon-Bowers, Oser & Flanagan, 1992). Collectivistic approaches to work organization are supposed to increase output, decrease losses, and increase job satisfaction (Harris, 1986). At the same time, the globalization of the market place has increased the opportunities for workers of different cultures to interact and work together. In addition to this exchange among workers from different countries, the increasing proportion of minority workers in American companies has resulted in a culturally diverse workplace. While the percentage of minorities in the workforce was approximately 18.9% in the year 1982, projections for the year 2005 suggest that minorities will encompass 27.6% of the U.S. workforce (U.S. Department of Labor, 1995). Thus, as Ilgen, LePine and Hollenbeck (1997) have argued, heterogeneous teams are created not only for political purposes, but also for the very practical purposes of creating and benefiting from a more diverse pool of knowledge, skills, values, and experiences that can be brought to bear on the team’s task. As a result, work groups in many U.S. organizations are becoming increasingly culturally heterogeneous and, thus, are receiving more attention because a thorough understanding of these groups and their performance can improve overall company productivity (cf. Nemetz & Christensen, 1996; Watson, Johnson & Zgourides, 2002). Traditional research on the dynamics and efficacy of work groups has focused on groups and teams that were relatively homogeneous with respect to culture. In light of the increasing diversity of the American workplace, however, it will also be important in the future to study the effects of cultural heterogeneity on work group processes and performance. The current chapter focuses on this issue, especially with respect to cultural variables that, in the past, have been found to affect individual performance. Examples of cultural variables that may be hypothesized to also affect team performance include socio-cultural factors such as norms and expectations about leadership, as well as culturally based variables such as group orientation, power distance, individualism/collectivism, and context (high/low). In this paper, these variables (and others) are evaluated with respect to their potential impact on work groups, rather than on the organization at large or the individual member of the organization. Before investigating the potential effects of cultural heterogeneity on team performance, however, it is important to identify the research issues underlying the terms “teams,” “groups,” and “culture.” Towards this end, the first section of this chapter focuses on the problems involved in defining these constructs and in performing research in these areas. In particular, the following issues are addressed: (a) What needs to be considered when studying teams and groups?; (b) Teams, Diversity, and Culture 319

What is meant by different “cultures?”; and (c) How can the relationship between culture and team performance be studied? Our second section contains a brief review of the cultural factors that have been identified as affecting human performance. This review serves as the basis for the third section of this chapter, which investigates if – and how – cultural heterogeneity affects team performance. In particular, this third section contains a discussion of the potential utility of culturally diverse work teams. Finally, the last section reviews how organizational scientists have proposed that the culturally diverse workplace be managed and how to improve performance when faced with cultural diversity.

RESEARCH ISSUES

Teams and Groups

Over the last quarter of the 20th Century, “groups” and “teams” have been a major focus of discussion and research in the American workplace. In the quest for a broader knowledge base on the usefulness of collective approaches in the workplace, the terms teams and groups have often been used synonymously to describe a variety of multi-person units. This trend is characterized by an almost inflationary use of terms such as “work teams,” “work groups,” and “self-directed work teams.” In the early nineties, however, there was an effort to discriminate teams from groups. According to these definitions, “teams” differ from “groups” by the interdependence of their members and by their purposiveness (cf. Salas, Dickinson, Converse & Tannenbaum, 1992). That is, teams are conceptualized to have a specific purpose and a membership that is based on a purpose (or task), not merely on shared characteristics, as would be more representative of groups. While the distinction between teams and groups has undoubtedly changed the scope of some recent studies, there still remain questions about the extent to which teams and groups differ and how much of the existing literature applies only to teams, only to groups, or to both. Additionally, it is not clear to what degree the differences between teams and groups moderate the effects of team-level variables on intra-team processes and team (or group) performance. In this paper, therefore, the terms “teams” and “groups” are again used synonymously, unless a further distinction is needed to adequately describe the effects under review. One research issue which repeatedly has raised concern in the study of teams and groups is the temporal duration of membership (McGrath & O’Connor, 1996). Most experimental studies have been conducted in the laboratory with so-called “ad-hoc” teams or groups, that is, teams whose members did not know each other 320 FLORIAN JENTSCH ET AL. before the investigation and who would not see each other afterwards (i.e. teams that had no common history or future). It can, however, be hypothesized that the interactions of team members who have known each other for a while are different than those of team members who are completely unfamiliar with each other. In fact, some research has found just that; teams go through a developmental process in their lifetime during which many team processes and interactions change (Bowers & Jentsch, 1995; Gersick, 1988, 1989). For example, Air Traffic Control teams often work with the same team mates for extended periods of time, which results in learning teammate-specific competencies that ultimately affect team processes and performance (Smith-Jentsch, Baker, Salas & Cannon-Bowers, 2001). As a result, it is important to identify whether teams are newly formed or have existed for a while when investigating their performance. Because culture is supposed to have long-lasting effects, this may be an even more important step if one studies the effects of culture over extended periods of time (as will be discussed in the section on culture as a variable in research). In addition to the temporal dimension of team development, however, another important factor in judging the impact of previous research on teams and groups concerns the tasks that the teams/groups are asked to perform. Here, the main criticism centers on the lack of accountability and response contingency that is involved in performing laboratory tasks. In particular, critics have asked whether laboratory tasks are as meaningful and as important to team members as those found in the real world. Further, do laboratory tasks require the same processes for their solution as real ones? While some researchers, for example, Driskell and Salas (1992), have answered this criticism by pointing out that laboratory research is useful to test theories under controlled conditions and to provide initial insight into team performance processes, other researchers have called for examining groups within the appropriate context (McGrath, 1993). The issue of the generalizability of the results remains an important one that must be considered when reviewing the merits of team performance research. In the review further below, therefore, particular attention is given to the tasks used by researchers in their cross-cultural investigations of team performance.

Culture

The second important component of research into the performance of culturally diverse teams concerns the definition of “culture.” In the past, the term “culture” has been used in a variety of ways and with respect to a number of constructs that appeared to distinguish and discriminate among people. Cultures have been defined on the basis of national origins, languages, socioeconomic strata, religious beliefs, Teams, Diversity, and Culture 321 and ethnicity. Italians (and even Italian Americans), for example, are supposed to have a different culture than Poles (or Polish Americans). Roman Catholics allegedly have a different culture than Southern Baptists; African-Americans vary in culture from Hispanics, etc. In addition to these cultural categorizations that were based primarily on physical, geographical, or historical distinctions, organizational psychology has used the term culture to describe differences in behavior, norms, and expectations between different organizations. According to this extended use of the term culture, the Disney Corporation, for example, can be hypothesized to have a different culture than Texaco, and such differences may produce differences in behavior within the organization (for discussion of “Organizational Culture” at this level of analysis see, for example, Kotter & Heskett, 1992; O’Reilly & Chatman, 1996; Wilkins & Ouchi, 1983). All these examples of cultural diversity are plausible and correct if one defines “cultures” as the “human-made part of the environment” (Triandis, 1987, 1993). That is, one sees cultures as those sets of beliefs, norms, expectations, and traditions which are common to all members of the same population subgroup, but differ among members of different population subgroups. As a result, “culture” can be used to describe members of a wide variety of different population subgroups, as long as their members share a number of characteristics while differing with respect to a number of other factors. At the same time, however, “culture” has become somewhat of a catch-all. Similar to the situation faced by scientists with respect to terms such as “groups” vs. “teams” (see, for example, Salas et al., 1992) or “shared mental model” (e.g. Mohammed, Klimoski & Rentsch, 2000), there appears the possibility of confounding levels of cultural variables as well. This can lead to misunderstandings and miscommunications among those discussing the effects of culture on human behavior. As a result, it is usually impossible to compare across studies that were performed at the different levels of “culture,” be it national, ethnic, or organizational. In addition to the problem of confounding levels of variables, there are additional problems in cross-cultural research, particularly as it relates to cross-national research: National culture (at least to a large degree) is supposed to be the result of long-term processes (such as socialization during upbringing, etc.) that cannot easily be changed. This makes research difficult because pure exper- imental research is virtually impossible. Consequently, only quasi-experimental or correlational methods are left for research, resulting in the typical problems of determining causation. Organizational scientists have recently attempted to address this issue by conducting experimental studies with culture as a factor. For example, a recent study by Brockner and colleagues (2001) manipulated scenarios where procedural justice was high or low and found differences across cultures but also differences within cultures dependent upon participants’ predisposition 322 FLORIAN JENTSCH ET AL. to power distance perceptions. Further, as Child and Tayeb (1987) pointed out, in many early cross-cultural studies,

the operationalization of culture was typically confined to attitudes and opinions that abstracted from, and discouraged attention to, national institutions and sociopolitical systems (p. 23). As the aforementioned discussion illustrates, not everything that has traditionally been attributed to “culture” may really have been a result of cultural differences. Instead, the findings from cross-national research could have been the result of differences in other variables, such as differences in research settings, socioeco- nomic conditions, etc. Consider, for example, the following two studies; one by Birnbaum-More, Wong and Olve (1995), and the other by Kopper (1993). Both studies compared managerial attributes between groups of managers that had the same native language, but came from neighboring countries (P. R. China/Hong Kong for Birnbaum-More et al., and Switzerland/Germany in Kopper’s case, respectively). Although, on the surface, these country-pairs might be seen as identical in culture (having the same language, same ethnicity, and similar histories), there were significant differences in managerial attitudes and behaviors between the Chinese managers from the P. R. China and those from Hong Kong. Likewise, Swiss and German managers varied considerably from each other. This finding thus emphasizes how important it is to clearly identify what is different in two cultures and to only study the effects of these variables on performance, while keeping all other variables constant. Attempts to do just this by more clearly defining the constructs underlying cul- tural variables have been the main focus of recent research. For example, Hofstede (1981) and Triandis (1987), have each provided taxonomies listing a number of variables that can be used to more clearly define cross-cultural differences. Sample items include “power distance,” “uncertainty avoidance” (Hofstede), and “individualism-collectivism,” or “masculinity-femininity” (Triandis). It is this more detailed approach to cross-cultural research that has resulted in a number of useful dimensions along which team heterogeneity may be defined.

CULTURAL DIMENSIONS

To understand the possible impact of cultural variables on human performance, it is useful to briefly identify cultural variables that might be related to individual performance. As stated before, a number of cultural dimensions have been identified. Triandis (1987), for example, has proposed a total of 20 cross-cultural dimensions. As can be seen from Table 1, Triandis grouped these dimensions into three categories: (a) perceptual differentiations (of how people fit into the culture); Teams, Diversity, and Culture 323

Table 1. Dimensions of Cultural Variation (Triandis, 1987, pp. 142–143).

Perceptual differentiations (1) What the other does vs. who the other is (family, in-group, age, race, religion ...). (2) Who is in the in-group? (3) Size of the in-group. (4) Ease of getting into the in-group. (5) Emphasis on age. (6) Emphasis on sex. (7) Emphasis on social class (power distance). (8) High vs. low self-esteem. (9) High vs. low power self-concept. (10) High vs. low activity self-concept. Utilization and evaluation of information (11) Ideologism vs. pragmatism. (12) Associative vs. abstractive communication (field-dependent vs. field-independent). (13) Human nature is good vs. bad. (14) Mastery over nature is good vs. bad. (15) Emphasis on past, present, future. (16) Emphasis on doing, being, being-in-becoming. (17) Individualism vs. collectivism. (18) Uncertainty avoidance. (19) Masculinity vs. femininity. Patterns of actions (20) Contact vs. non-contact.

(b) differences in the utilization and evaluation of information; and (c) patterns of action. Individualism and collectivism, for example, a common dimension of cross-cultural research, are categorized as a difference in the utilization and evaluation of information, because this dimension describes how information about the relationships among people is used to determine one’s own actions. For example, collectivist people tend to emphasize group-values more than individualistic people, even if the group norms and expectations are in conflict with the interests of the individual. This dependence upon the group situation as the guiding principle of one’s own behavior has also often been called “situational ethics” (Triandis), and has been related to conformism and familism (where family ties are more important in selecting business associates, for example, than qualifications). As can be seen from this brief description, the dimension of collectivism and individualism is probably the one that has the most direct link to group and team performance, because it describes the norms and expectations about how individuals fit into a group scheme, what importance collective endeavors and 324 FLORIAN JENTSCH ET AL. consensus have, etc. It has therefore been extensively studied in cross-cultural research on teams (cf. Earley & Gibson, 1998; Erez & Somech, 1996; Gibson, 1999; Jung & Avolio, 1999). Other dimensions from Triandis’ framework that are commonly found in cross-cultural research and are important for team performance research include communication dimensions (e.g. abstractive vs. associative, etc.), uncertainty avoidance, and power distance. Most research with respect to the latter dimensions, uncertainty avoidance and power distance, has been performed by Hofstede (1981). According to Hofstede, power distance is concerned with the way a culture handles the exercise of personal or institutional power. In countries with high power distances (e.g. India), it is accepted and expected that leaders have and exercise strong power, even if that fosters social inequality. In countries with low power-distance, on the other hand (e.g. Austria), superiors and subordinates are not separated by a large gap in power but see themselves more as equals. Clearly, this dimension can be hypothesized to affect team processes, for example, in leader-subordinate exchanges. Uncertainty avoidance, according to Hofstede (1981), describes the way cultures cope with novelty and situational ambiguity. High uncertainty avoidance countries (e.g. Japan, Germany) seek to impose a rigid regulatory framework on everyday life, which avoids situational ambiguities and limits the room for individual interpretations. On the other hand, countries with low uncertainty avoidance (e.g. Denmark), are rather tolerant and seemingly unstructured. As such, they are also often more flexible in responding to changing needs. Again, it is possible to hypothesize that team member differences on this dimension can have an important impact on intra-team processes by creating differences in the ways team members would handle the operations of the team. It is therefore even more surprising that there has been virtually no research on these factors with respect to team and group performance, a topic we turn to next.

CROSS-CULTURAL RESEARCH ON TEAMS AND GROUPS

When reviewing the literature directly related to cultural factors in teams and groups, one quickly notes the dearth of scientific research in this area (cf. Ilgen, LePine & Hollenbeck, 1997; Morgan, Bowers & Jentsch, 1994). In part, this can be explained by the very practical problems involved in cross-cultural studies. For example, McLeod, Lobel and Cox (1996) questioned the extent to which comparisons can be made across cultures when tasks and team performance may not be relevant across cultures. A study that reiterated this notion was performed by Crowley (2001) in which she found that the variables that predicted Teams, Diversity, and Culture 325 performance for U.S. teams were not the same variables that predicted team performance in the other countries (Argentina, Austria, Brazil, and Canada). In addition, as was pointed out in the previous section on research issues, finding matching groups in two different cultures is not an easy task. Those studies that report such investigations, then, are of particular value. A very important empirical study that was published in this area was conducted by Cox, Lobel and McLeod (1991). These authors compared the performance of teams from four ethnic groups, Anglo-American, African-American, Asian- American, and Hispanic-Americans, in performing the “prisoner’s dilemma” task. Cox et al. focused on the construct of individualism-collectivism. They hypothe- sized that: (a) individuals from certain ethnic American groups (i.e. Asian, African, and Hispanic Americans) would be more cooperative than Anglo-Americans (a proposition based on Triandis and Hofstede’s collectivism dimension); and that, as a result, (b) mixed teams with people from various ethnic backgrounds would behave more cooperatively than all-Anglo teams. In particular, Cox et al. expected that minority team members would be especially cooperative when situational cues pointed in a cooperative direction, whereas Anglo-American teams and their team members would not increase their cooperativeness in response to situational cues. In the study, 136 students from the four ethnic groups who attended a large university in the Midwest of the U.S. were distributed into four-person teams. These were either all-Anglo teams or mixed teams in which the Anglo-American team member was in the minority. The teams then performed the prisoner’s dilemma task. In this task, the payoffs received by team members are a function of their own choice, in conjunction with the choices of the other players. If, for example, all players elect to cooperate, they all receive a moderate payoff. If one player is the only one that is not cooperative, but rather competitive, this player receives a higher payoff, while the other players lose their payoff. Finally, if all players elect to be competitive, they all lose. As a result, cooperative strategies only pay off if everybody is also cooperative, whereas competitive strategies only pay if at least one player is cooperative. In order to manipulate the situational cues of cooperativeness, Cox et al. (1991) gave the teams either no feedback or feedback that pointed towards the value of cooperative strategies. Results showed that the feedback itself did not have a main effect. There was, however, a feedback by ethnic group interaction in that mixed teams were more cooperative after receiving feedback which emphasized cooperative behavior than when no feedback was received, whereas all-Anglo groups did not show an increase in cooperativeness. In addition, there was a main effect of ethnic team composition in that, as predicted, ethnically diverse teams were more cooperative (in all conditions) than all-Anglo teams. 326 FLORIAN JENTSCH ET AL.

From their study, Cox et al. (1991) concluded that the cultural dimension collectivism-individualism was related to changes in behavior not only for individuals (their manipulation check), but also for complete teams made of several individuals. Indeed, in this study, all-Anglo teams selected cooperative strategies only 25% of the times, whereas mixed groups elected to cooperate in more than 70% of the cases. Cox et al. concluded that this suggests organizations with an ethnically diverse work force may be better suited for intra-team cooperation than those with teams made up exclusively of inherently less cooperative workers (i.e. Anglo-American). Cox et al. further noted that there was significantly more research needed in this area by stating that “there is a need for studies addressing the differences between homogeneity and heterogeneity more generally” (p. 843). Others have investigated performance differences between culturally heteroge- neous and homogeneous work groups (Watson, Kumar & Michaelsen, 1993). In an attempt to avoid theoretical problems arising from the use of one-time, ad-hoc teams in an artificial task setting (a criticism waged by Watson et al. against Cox and his colleagues), Watson et al. studied group performance on a series of complex problem-solving tasks that were of importance to the group members. Watson et al. tried to achieve the goal of giving their participants meaningful tasks by studying teams of college students who were assigned to work together in grade-relevant team exercises. Further, Watson and colleagues employed a longitudinal design by studying the teams during a complete semester (about 4 months), and they studied groups that were very culturally diverse (i.e. those that contained one white American, one African American, one Hispanic American, and a foreign national from a country in Asia, Latin America, Africa, or the Middle East). Watson et al. (1993) made the following predictions about the performance of the teams. First, they hypothesized that among the newly formed groups (i.e. at the beginning of the study) performance would be better for the homogeneous groups than for the culturally heterogeneous groups, both with respect to the inter-group processes and their actual performance. Later in the semester, however, Watson et al. expected that heterogeneous groups would “catch up,” and even overtake the homogeneous groups, especially in regard to problem-solving tasks. The groups were tested at four times throughout the semester using a variety of criteria (e.g. performance measures, process measures, and with respect to feedback). The results supported Watson et al.’s (1993) hypotheses in that ho- mogeneous groups reported significantly more effective processes during the first three measurement periods than heterogeneous groups. At the last measurement point in the semester, however, heterogeneous and homogeneous groups reported equally effective intra-team processes. Similarly, actual performance was lower for Teams, Diversity, and Culture 327 the heterogeneous groups than for the homogeneous groups during measurement periods 1 and 2. Then, however, the gap was reduced, and heterogeneous groups performed equally well as homogeneous groups during measurement periods 3 and 4. In fact, with respect to problem-solving performance (e.g. generating alternatives and future perspectives), performance at the last measurement point was better for heterogeneous groups than for those teams that were made up of similar individuals. Watson et al. (1993) concluded from the results that culturally diverse groups have the potential to excel on complex problem-solving tasks, but he and his colleagues also pointed out that this potential may only be reached after a considerable time of common team development. This finding led Watson et al. to question the results of other studies which had found that heterogeneous groups and teams may initially show better performance. To reconcile this apparent contradiction, Watson at el. suggested as a future direction in research the investigation of the mechanisms underlying the differences in performance of heterogeneous and homogeneous work teams. Just such a theoretical work was later conducted by Jackson, May and Whitney (1995). In a review of the literature on the performance of diverse teams, these authors provided additional evidence that team member diversity may be useful in improving team performance with respect to, at least, problem solving. Jackson et al. summarized a number of previous studies that investigated diversity with respect to a number of characteristics (e.g. personality, leadership, training background, leadership style, etc.). They found that, generally, team member heterogeneity seemed to improve performance in tasks that required creativity. This is a finding that was also supported by other reviews of studies investigating the team heterogeneity-homogeneity issue: Shaw (1983), for example, noted that it is problem-solving in particular that may benefit from group heterogeneity, whereas heterogeneous groups that perform other tasks may perform worse than homogeneous groups on other tasks. Watson, Johnson and Zgourides (2002) recently conducted a study in which they studied the effects of culture on leadership, group process, and group performance in learning teams. This study focused on learning teams, composed of students working together throughout a four month semester. The learning teams included 75 culturally heterogeneous teams (members from at least three cultures) and 90 culturally homogenous teams (all white Americans). The findings of the study showed that the two types of teams differed in the leadership styles that imme- diately emerged, with culturally heterogeneous teams emphasizing interpersonal leadership and culturally homogenous teams displaying leadership more related to the task at hand. Specifically, the culturally heterogeneous teams took time to deal with problems stemming from different viewpoints, while the culturally 328 FLORIAN JENTSCH ET AL. homogeneous teams were able to begin focusing on their task immediately. In contrast, no differences were initially found between the culturally heterogeneous vs. culturally homogeneous teams on member cohesiveness or self-orientation. During the later stages of the teams’ life cycles, the teams exhibited most of the same characteristics that they had initially, with the exception of the culturally homogeneous teams reporting higher self-orientation scores. In the end, the culturally heterogeneous teams outperformed the culturally homogenous teams, with the cultural differences being predictive of final team project grades. The authors concluded, similar to previous studies, that the culturally heterogeneous teams benefited from their differences only after they were able to take advantage of the unique viewpoints brought to the problem-solving process. A number of studies highlight the importance of considering how the task may influence group processes depending upon the composition of the group. Essentially, heterogeneous groups may be beneficial with tasks that are more divergent in nature because such groups are less likely to become fixated on a particular solution. Fixedness is the inability to reconsider one’s alternatives and to continue to try and use an ineffective means to discover a solution, such as the counterproductive use or undesirable effect of prior knowledge (Smith, 1995). This is particularly problematic when the task requires considering a broad range of alternative issues as was the case for the Watson et al. (1993) study discussed previously. Heterogeneous groups, because of their varied backgrounds, may be better suited to overcome fixation because breadth of experiences, in any form, can be enough to break an impasse. Similarly, a homogeneous group may not be able to overcome fixation and fail to successfully solve a problem. For example, in the context of teams of scientists, Dunbar (1995) showed that labs with members of similar background were less effective problem solvers than labs with members who had diverse backgrounds. Earlier studies found that contact and interaction between scientists was only related to productivity when the scientists differed in their motivations and in their experiences (Pelz, 1956, as cited in Hoffman, 1959). Increases in group diversity can also increase the amount of conflict within the group. Conflict has been shown to influence cognitive change (Levine, Resnick & Higgins, 1993) and to increase group creativity (Nemeth & Kwan, 1985; Nemeth & Wachtler, 1983). Therefore, conflict could actually be beneficial to problems requiring any type of divergent processes. Specifically, conflict could help on these problems by helping the group overcome the consequences of fixation. For example, organizational scientists suggest that the success of the Marshall Plan was due in part to the diverse background of the policy planning staff and the varied input they offered (Janis, 1982; cited in McCauley, 1989). The heterogeneous “backgrounds of the group members were put to use in discussions intense enough Teams, Diversity, and Culture 329 to produce real intellectual agony for the participants, as proposals were criticized and debated” (p. 256; McCauley, 1989). Thus, when it comes to problem solving teams, research on cognitive processes and individual problem solving can illumi- nate this issue. According to Perkins (1995), “the human problem-solver ...needs in addition to conventional clue-oriented expertise another kind of expertise as well. An effective search process ...involves casting a wide net, avoiding redundant coverage of the same regions, searching for new regions altogether, and so on. Roughly, a more divergent style of search is required” (p. 515). Early research conducted for the military used complex laboratory tasks to investigate group process and performance. For example, in the “Mined Road Problem,” members had to devise a means to cross a heavily mined road using various materials at their disposal and all without leaving any trace of their having crossed (Hoffman, 1959; Tuckman & Lorge, 1962). In order to determine whether homogeneous groups would perform better or worse than heterogeneous groups, Hoffman (1959) administered a survey designed to measure a series of differing personality traits. Based upon their responses, subjects were assigned to four-person groups differentiated by either homogeneous or heterogeneous traits. Early studies looking at composition factors found that that heterogeneous groups outperformed homogeneous groups, with a mean score nearly 20 points higher (Hoffman, 1959). Other problem solving tasks show similar patterns. In trying to determine how to best allocate funds in a laboratory task, Hoffman and Maier (1961) found that heterogeneous groups performed better than their counterparts in the homogenous groups. In an early study of managerial decision making laboratory groups role played as factory workers and supervisors (Hoffman, 1959). In this situation, group members had to adopt a new system of work based upon employee skills and the problem allowed for creative solutions that could be rated on quality. Several studies suggested that heterogeneous groups tend to provide more inventive procedures (see also Hoffman & Maier, 1961). Further, Chatman, Polzer, Barsade and Neale (1998) found that cultural diversity in work groups was inversely related to group member focus on organizational objectives, while Harrison, Price and Bell (1998) found that the negative influences of cultural di- versity can weaken as other social differences and categories that were not initially apparent begin to emerge. In a study illustrating the importance of which measure is being used to determine heterogeneity, Williams and Sternberg (1988) investigated both motivational and ability factors. The problems were designed to “encourage divergent, creative solutions.” For example, subjects had to devise a marketing plan for a new sweetener while taking into account a number of factors. The problem clearly has many acceptable answers of low demonstrability. At the same time the problem requires a large search space if the problem solver wishes to 330 FLORIAN JENTSCH ET AL. do well. Results showed that groups heterogeneous in motivation did worse than groups homogeneous in motivation. But, groups heterogeneous in the diversity of previously measured “expressed ideas” did better. Thus, heterogeneity of expressed ideas aided in the creative portion of the task but could be attenuated if members were not homogeneous in motivation. Conversely, heterogeneity may hurt on tasks requiring convergent processes, (i.e. tasks relying more on hypothesis testing). This possible negative effect of group heterogeneity would result from coordination problems which may result from some form of conflict within the group. As the group gets more diverse, the decreasing similarity of members may thus lead to coordination difficulties on more convergent tasks (i.e. tasks for which organization becomes increasingly more important). Coordination problems include factors such as the inability to effectively communicate with all members, or problems in understanding one’s role in the task (Moreland, Levine & Wingert, 1996). There may be decreasing gains in heterogeneity due to coordination losses offsetting any increases in group diversity (Yetton & Bottger, 1983). Steiner (1972) termed such losses as “coordination decrement” and defined it as the difficulty posed by attempting to get all members to work together at their full potential. Thus, problems which require a greater amount of group coordination such as those relying less on divergent processes, are likely to show such coordination decrement. Alterna- tively, homogeneous groups may perform better when convergent processes are required because there is likely to be less coordination difficulty due to less conflict among similar members. Controlled empirical investigations needs to be conducted to more accurately disentangle these process interactions but, as discussed, tangential evidence supports some of this prediction.

Summary

The pattern of findings described above suggests that team member heterogeneity with respect to cultural factors may be considered a mixed blessing. Indeed, these findings are similar to heterogeneity with respect to other factors, such as person- ality, that has been found to affect team performance. For example, Hernandez, Dockery, Bowers and Jentsch (1995) found that aircrew heterogeneity with respect to personality factors such as sociability interacted with work-related factors to affect aircrew performance. It appears that the gains in creativity and unique solu- tions from the heterogeneous makeup of teams are – at least initially – outweighed by the increased process losses associated with the miscommunications, misunder- standings, and lack of shared mental models that result from the differences among the team members. This phenomenon has been termed the “diversity/consensus Teams, Diversity, and Culture 331 dilemma” by Argote and McGrath (1993). On the one hand, it would seem that groups heterogeneous with respect to knowledge and personality would benefit if the task is more divergent. When a divergent thinking style is required, heterogeneous groups tend to perform better than homogeneous groups and they show a trend towards more creative solutions. As Stasson and Bradshaw (1995) concluded in their study of group problem solving, a group will only be beneficial as a problem solving unit when it is made up of individuals with non-overlapping abilities. On the other hand, if the task is more convergent, a heterogeneous mix may be more likely to experience coordination problems and may harm the problem solving process. Further, process losses due to the lack of a common set of language, norms, expectations can be particularly damaging to performance in the first stages of team development (cf. Gersick, 1988, 1989); those phases that are commonly known as “forming,” and “storming.” Especially during the “storming” stage, team members create common models about the functioning of the team on the basis of conflict. That is, they explore the boundaries of the team with respect to its norms and expectations, as well as regarding the role of leadership. Only through this exploratory process do team members assemble an implicit list of acceptable behaviors and interactions that make the team function. It is in later stages of team development (those known as “norming” and “performing”) that team members adapt to the existence of their differences and make positive use of the unique contributions by individual team members. Some researchers (in particular Watson et al., 1993) have therefore suggested to further study the usefulness of homogeneous and heterogeneous teams as a function of the tasks the teams are asked to perform. While these suggestions point to useful additions to our knowledge about team and group performance, there still remain questions as to the exact mechanisms by which cultural diversity affects team processes, and subsequently team performance. One study focusing on the underlying mechanisms at play in culturally heterogeneous vs. culturally homogeneous teams was performed by Chatman and Flynn (2001). They looked at learning teams composed of MBA students as well as business units within a financial services firm in the U.S. The results of the study showed that, compared to culturally homogeneous groups, culturally heterogeneous groups were less likely to develop or notice cooperative norms early in their life cycles and that these cooperative norms were less stable throughout the life cycle of the groups. The groups that were able to develop stable cooperative norms outperformed their counterparts and group members reported higher levels of satisfaction with the group performance. However, no correlation between cultural diversity and reported levels of satisfaction was found. Thus, Chatman and Flynn concluded that the level of cooperation, and 332 FLORIAN JENTSCH ET AL. not the cultural heterogeneity, was directly affecting the satisfaction of the group members. These findings led the authors to caution other researchers in taking the direct effects of cultural diversity for granted, as there may be mediating factors, such as cooperative norms, which are the actual determinants of group processes and performance. In an attempt to identify other underlying mechanisms in heterogeneous teams, Earley (1993) conducted an investigation on the linkage between individualism- collectivism and team performance. Earley, who conducted his study with Chinese, Israeli, and American participants, investigated to what degree membership in a perceived in-group (i.e. a group in which members belong to the same group as the main players) or out-group (i.e. a group where members are not part of the same group as main players) affected performance of individualistic and collectivistic team members. As expected, team members with an individualistic orientation performed best when working alone, and performed worse when they thought they were part of an in- or out-group. Collectivistic team members, on the other hand, performed best when they thought to be part of an in-group, while performing worse when alone or part of a perceived out-group. Earley concluded that culturally diverse orientations not only affected team performance indirectly through group member composition, but also directly through the individuals’ perceptions that mediated the effects of the cultural differences. Thus, these results suggested that diversity in itself may not always lead to increased team performance, but may require a certain team climate for its effectiveness, a topic we turn to next. The suggestion that cultural variables are moderated by situational factors such as climate has been supported by Kirchmeyer and Cohen (1992) who found that changes in group communication structure and climate can significantly affect the efficacy of diversity in increasing team performance. Kirchmeyer and Cohen investigated the role of creating a climate for constructive conflict in increasing the role of minority team members. Because collectivistic team members often put the value of team consensus above that of their unique contributions (Johnston, 1993), they frequently elect not to voice criticism if that endangers the group’s consensus. This, of course, is exactly what is not intended when one tries to reap the benefits of culturally diverse work groups. Creating, however, a climate of constructive criticism (i.e. one in which disagreement was acceptable or even desirable), was found by Kirchmeyer and Cohen to increase minority involvement and diversity of opinions. One can therefore conclude that cultural variations can be ameliorated by the climate in which groups operate. Other factors that have been studied in this regard (i.e. that have been hypothesized to improve the positive effects of diversity in the workplace through the manipulation of climate) are feedback (Earley, 1989), how the work is structured and distributed among team members (Duncan, Teams, Diversity, and Culture 333

1994), and command/leadership style (Johnston, 1993). Based on these findings, then, a team climate which emphasizes the positive effects of cultural diversity by encouraging team members to voice their unique observations and opinions while at the same time reducing intra-team process losses in communication through the effective use of leadership and feedback may be the most important ingredient in making effective use of cultural diversity in the workplace. Future research is clearly needed to identify the exact mechanisms, strategies, and tactics by which such a climate can be achieved. In the time until such research can be completed, a number of practical suggestions are already available that can be used by managers in a variety of settings. Some of these suggestions are reviewed in the following section.

HOW TO BEST MANAGE CULTURAL DIVERSITY IN TEAMS

Considering that culturally diverse workgroups are: (a) more likely to increase in prevalence in the future; and (b) may be beneficial for certain task settings (particularly those requiring creative problem solving), organizational scientists have begun to develop a number of suggestions on how to best manage a culturally diverse workplace. In this last section, a number of these suggestions are briefly reviewed. One main focus of current suggestions of managing cultural diversity in the workplace centers on creating an awareness and acceptance of cultural differences (Rigby, 1987). That such sensitivity to cultural difference is an important prerequisite for the success of multi-national organizations has been shown by Lester (1987) who reviewed several failed transatlantic mergers between European and American corporations. According to Lester, such mergers require adequate preparation and awareness of cultural differences. Others have supported this view. Thorne (1986), for example, demonstrated that cultural stereotypes among managers can be significant barriers for multi-national cooperation when he stated that “French, German, Italian, and British executives are likely to be successful only within the confines of their national boundaries” (p. 90). Likewise, Rosen and Lovelace (1994) pointed out that “a balance between expectations for assimilation and efforts to accommodate diverse groups spells the success of diversity management programs” (p. 86). To reach this goal, Rosen and Lovelace (and others, e.g. Dorgan, 1994) suggested the creation of a climate in which every employee has the potential and opportunity to advance, regardless of his or her cultural background. Finally, Janssens and Brett (1997) proposed the use of meaningful participation, that is, teaching employees to respect each other’s 334 FLORIAN JENTSCH ET AL. unique cultural backgrounds in an effort to build trust, and ultimately performance, within work groups. Another avenue that has been suggested to increase the effectiveness of organizations in managing teamwork lies in personnel selection. Klein (1992), for example, pointed out that, traditionally, in the U.S., employers have made personnel selections primarily on the basis of applicants’ experiences and educa- tion. While this may be a useful strategy when selecting individuals who perform clearly defined tasks without much interaction with others, the use of teams may call for another strategy that focuses on the attitudes, beliefs, and abilities of working with others. According to Klein, such an emphasis on applicant attitudes is characteristic of the Japanese approach to personnel selection. Given the success the Japanese teamwork approach has had, Klein suggested the introduction of the concept of team member attitudes into the selection of individuals for team task settings. At the same time, however, selection is not a panacea. As Ilgen et al. (1997) have pointed out, selection does nothing to help existing teams. Further, there is very little guarantee that, even when teams are newly established, selection on the basis on beliefs about team work alone would be able to overcome other cultural and national differences with respect to language, communication style, background knowledge, etc. Consequently, selection may have limited utility in improving performance in culturally heterogeneous teams. A second approach towards “homogenizing the team” (Ilgen et al., 1997) is socialization, especially through structured training in the areas of language, history, goals and values, and formal and informal work relationships. For example, Klein (1992) suggested the use of training that focuses on general teamwork skills for improving the management of cultural diversity in the workplace. That such skill-based cultural diversity training may be a useful addition to personnel selection and awareness training alone has been shown by Harrison (1992) who investigated the use of several strategies (i.e. behavior role modeling and the use of a so-called “cultural assimilator”) for cultural diversity training. According to Harrison’s study, which was conducted with 65 U.S. government employees, each approach alone was successful in teaching relevant skills, but, additionally, the combination of both strategies led to even better training results. Harrison’s findings and those of others (e.g. Williams, Hayflich & Gastomn, 1986) point towards the contention that working in a culturally diverse workplace requires a certain set of skills and that these skills are modifiable through training. Management style, alone or in combination with personnel selection, aware- ness training, and behavioral training, has been shown to positively affect the performance of culturally diverse groups (e.g. Fine, 1983; Gemmel, 1986). Using a U.S. subsidiary of a Japanese corporation as a case study, Gemmel demonstrated Teams, Diversity, and Culture 335 the synergistic effects of combining the individualistic and direct American management style with the more collectivistic and consensus-building Japanese approach to management. Based on the commercial success of the company in his example, Gemmel concluded that the combination of the two management styles can be of benefit to the company. Similarly, Mai-Dalton (1993) has pointed out the potential gains that such a strategy promises for corporations who aggressively manage cultural diversity to their advantage. There are, however, some who point out that using a mixture of different management styles may not always be successful. Ishida (1986), for example, has suggested that the success of the Japanese management style may be moderated by the socioeconomic status of the workers. According to Ishida, Japanese management doctrines (which are based on egalitarianism, collectivism, and consensus) are much more appealing to American blue-collar workers, who traditionally are more collectivistic and value job-stability. Middle- and upper-managers, however, who are socially and economically more mobile, often resent this component of the Japanese management style. A final approach to alleviating the diversity/consensus dilemma (Argote & McGrath, 1993) has been the suggestion by Ilgen et al. (1997) to increase the structure of the task given to the team. As these authors have pointed out, tasks that are relatively straightforward in terms of structure, require few assumptions, and have unambiguous performance criteria, are probably more suitable for heterogeneous and cross-national teams than tasks that do not share these characteristics. Empirical evidence that such a strategy indeed benefits teams, however, is still scant. Thus, there is not necessarily an easy answer to the question of how to best manage cultural diversity, and there are clearly a number of research issues that still need to be addressed.

CONCLUSIONS

The goal of this chapter was to provide the reader with an understanding of the impact of cultural diversity on groups and teams in today’s workforce. While this is an important topic, the numerous limitations to conducting research on culture and teams that have been outlined above demonstrate the true complexity of determining the effects of cultural heterogeneity on team performance, and consequently, applying those findings to the real world setting. Even with differing opinions on culture and its dimension, proposed differences between groups and teams, and concerns about the generalizability of studies, it is clear that cultural heterogeneity does influence team processes and team performance in some fashion. Thus, clearly this field of study cannot be ignored. 336 FLORIAN JENTSCH ET AL.

On the basis of the literature reviewed in this paper, three main conclusions can be drawn. First, group- or team-based approaches to work organizations are here to stay. Second, because of the internationalization of the marketplace, cultural diver- sity is likely to increase rather then decrease. Third, organizations will have to find ways to reap the potential benefits of cultural diversity (e.g. increased creativity) while avoiding the pitfalls and process-losses associated with a diverse work force. While research in the past has shown that there are cultural differences and that these differences can affect individual performance, future research needs to focus on extending these findings to work teams and groups. In particular, future research needs to identify the mechanisms further by which culture affects intra-team processes and, subsequently, performance. Also, more research is needed to identify the factors that moderate the effects of culture on team performance. Finally, researchers must be able to translate their findings from the realm of research to that of the practitioner by providing managers with the tools and strategies to manage culturally diverse work teams most effectively. Given all these research needs, cultural diversity in teams and work groups is an area of cross-cultural psychology that is ripe for further investigation.

REFERENCES

Argote, L., & McGrath, J. E. (1993). Group processes in organizations: Continuity and change. In: C. Cooper & I. T. Robertson (Eds), International Review of Industrial and Organizational Psychology (Vol. 8, pp. 333–389). New York: Wiley. Birnbaum-More, P. H., Wong, G. Y. Y., & Olve, N.-G. (1995). Acquisition of managerial values in the People’s Republic of China and Hong Kong. Journal of Cross-Cultural Psychology, 26(3), 255–275. Brockner, J., Ackerman, G., Greenberg, J., Gelfand, M. J., Francesco, A. M., Chen, Z. X., Leung, K., Bierbrauer, G., Gomez, C., Kirkman, B. L., & Shapiro, D. (2001). Culture and procedural justice: The influence of power distance on reactions to voice. Journal of Experimental Social Psychology, 37(4), 300–315. Bowers, C. A., & Jentsch, F. (1995). The acquisition and decay of aircrew coordination skills. In: C. A. Bowers & F. Jentsch (Chairs), Empirical Research Using PC-based Flight Simulations. Symposium at the Sixth International Symposium on Aviation Psychology, Columbus, OH. Cannon-Bowers, J., Oser, R., & Flanagan, D. L. (1992). Work teams in industry: A selected review and proposed framework. In: R. W. Swezey & E. Salas (Eds), Teams: Their Training and Performance (pp. 355–377). Norwood, NJ: Ablex. Chatman, J. A., & Flynn, F. J. (2001). The influence of demographic heterogeneity on the emergence and consequences of cooperative norms in work teams. Academy of Management Journal, 44(5), 956–974. Chatman, J., Polzer, J., Barsade, S., & Neale, M. (1998). Being different yet feeling similar: The influence of demographic composition and organizational culture on work processes and outcomes. Administrative Science Quarterly, 41, 423. Teams, Diversity, and Culture 337

Child & Tayeb (1987). Theoretical perspectives in cross-national research. International Studies of Management & Organization, 23–70. Cox, T. H., Lobel, S. A., & McLeod, P. L. (1991). Effects of ethnic group cultural differences on cooperative and competitive behavior in a group task. Academy of Management Journal, 34(4), 827–847. Crowley, A. C. (2001). An investigation of cultural differences in factors related to work team effectiveness [Abstract Only]. Dissertation Abstracts International: Section B: The Sciences & Engineering, 61(8-B), 4456. Dorgan, W. J. (1994, January). Workplace diversity. Modern Machine Shop, 66(8), 110. Driskell, J. E. & Salas, E. (1992). Can you study real teams in contrived settings? The value of small group research to understanding teams. In: R. W. Swezey & E. Salas (Eds), Teams: Their Training and Performance (pp. 101–124). Norwood, NJ: Ablex. Duncan, W. J. (1994, February). Why some people loaf in groups while others loaf alone. The Academy of Management Executive, 8(1), 79–80. Dunbar, K. (1995). How scientists really reason: Scientific reasoning in real-world laboratories. In: R. J. Sternberg & J. Davidson (Eds), Mechanisms of Insight (pp. 365–395). Cambridge, MA: MIT Press. Earley, C. (1989). Social loafing and collectivism: A comparison of the United States and the People’s Republic of China. Administrative Science Quarterly, 34, 565–581. Earley, C. P. (1993). East meets west meets mideast: Further explorations of collectivistic and individualistic work groups. Academy of Management Journal, 36(2), 319–348. Earley, P. C., & Gibson, C. B. (1998). Taking stock in our progress on individualism-collectivism: 100 years of solidarity and community. Journal of Management, 24, 265–304. Erez, M., & Somech, A. (1996). Is group productivity loss the rule or the exception? Effects of culture and group-based motivation. Academy of Management Journal, 39, 1513–1537. Fine, H. R. (1983, Fall). Improving productivity – Western vs. Eastern styles. The Government Accountants Journal, 32, 42–49. [Abstract only: Cited in Business (1982 to 1987)]. Gemmel, A. (1986). Management in a cross-cultural environment: The best of both worlds. Management Solutions, 31(6), 28–33. Gersick, C. J. G. (1988). Time and transition in work teams: Towards a new model of group development. Academy of Management Journal, 31, 9–41. Gersick, C. J. G. (1989). Marking time: Predictable transition in task groups. Academy of Management Journal, 32, 274–309. Gibson, C. B. (1999). Do they do what they believe they can? Group efficacy beliefs and effectiveness across tasks and cultures. Academy of Management Journal, 42, 138–152. Harris, P.R. (1986). Building a high-performance team. Training & Development Journal, 40(4), 28–29. Harrison, J. K. (1992). Individual and combined effects of behavior modeling and the cultural assimilator in cross-cultural management training. Journal of Applied Psychology, 77(6), 952–962. Harrison, D. A., Price, K. H., & Bell, M. P. (1998). Beyond relational demography: Time and the effects of surface and deep-level diversity on work group cohesion. Academy of Management Journal, 41, 96–107. Hernandez, E., Dockery, K., Bowers, C., & Jentsch, F. (1995). Pilot personality, automation, and performance. Paper presented at the 103rd Annual Meeting of the American Psychological Association, August 1995, New York. Hofstede, G. (1981). Culture and organizations. International Studies of Management & Organization, 10(4), 15–41. 338 FLORIAN JENTSCH ET AL.

Hoffman, L. R. (1959). Homogeneity of member personality and its effect on group problem-solving. Journal of Abnormal and Social Psychology, 58, 27–32. Hoffman, L. R., & Maier, R. F. (1961). Quality and acceptance of problems solutions by members of homogenous and heterogenous groups. Journal of Abnormal and Social Psychology, 62, 401–407. Ilgen, D. R., LePine, J. A., & Hollenbeck J. R. (1997). Effective decision making in multinational teams. In: P. C. Earley & M. Erez (Eds), New Perspectives in International Industrial-Organizational Psychology (pp. 377–409). San Francisco, CA: Jossey-Bass. Ishida, H. (1986). Transferability of Japanese human resource management abroad. Human Resource Management, 25(1), 103–120. Jackson, S. E., May, K. E., & Whitney, K. (1995). Understanding the dynamics of diversity in decision making teams. In: R. A. Guzzo & E. Salas (Eds), Team Decision Making Effectiveness in Organizations (pp. 205–261). San Francisco, CA: Jossey-Bass. Janis, I. L. (1982). Groupthink: Psychological studies of policy decisions and fiascoes. Boston: Houghton Mifflin. Janssens, M., & Brett, J. M. (1997). Meaningful participation in transnational teams. European Journal of Work and Organizational Psychology, 6(2), 153–168. Johnston, N. (1993). CRM: Cross-cultural perspectives. In: E. L. Wiener, B. G. Kanki & R. L. Helmreich (Eds), Cockpit Resource Management (pp. 367–398). San Diego, CA: Academic Press. Jung, D. I., & Avolio, B. J. (1999). Effects of leadership style and followers’ cultural orientation on performance in group vs. individual task conditions. Academy of Management Journal, 42, 208–218. Kirchmeyer, C., & Cohen, A. (1992). Multicultural groups: Their performance and reactions with constructive conflict. Group & Organization Management, 17(2), 153–170. Klein, E. (1992). The U.S./Japanese HR culture clash. Personnel Journal, 71(11), 30–34. Kopper, E. (1993). Swiss and Germans: Similarities and differences in work-related values, attitudes, and behavior. International Journal of Intercultural Relations, 17, 167–184. Kotter, J., & Heskett, J. (1992). Corporate culture and performance. New York, NY: Free Press. Lester, T. (1987). The tricky business of transatlantic alliances. International Management, 42(4), 24–28. Levine, J. M., Resnick, L. B., & Higgins, E. T. (1993). Social foundations of cognition. Annual Review of Psychology, 44, 585–612. Mai-Dalton, R. R. (1993). Managing cultural diversity on the individual, group, and organizational levels. In: M. M. Chemers & R. Ayman (Eds), Leadership Theory and Research: Perspectives and Direction (pp. 189–215). San Diego, CA: Academic Press. McCauley, C. (1989). The nature of social influence in groupthink: Compliance and internalization. Journal of Personality and Social Psychology, 57, 250–260. McGrath, J. E. (1993). Time, interaction, and performance (TIP): A theory of groups. In: R. M. Baecker (Ed.), Readings in Groupware and Computer-Supported Cooperative Work, Assisting Human-Human Collaboration (pp. 116–129). San Francisco, CA: Morgan Kaufmann. McGrath, J. E., & O’Connor, K. M. (1996). Temporal issues in work groups. In: M. A. West (Ed.), Handbook of Work Group Psychology (pp. 25–52). Chichester, UK: Wiley. McLeod, P. L., Lobel, S. A., & Cox, T. H. (1996). Ethnic diversity and creativity in small groups. Small Group Research, 27, 248–264. Moreland, R. L., Levine, J. M., & Wingert, M. L. (1996). Creating the ideal group: Composition effects at work. In: E. H. Witte & J. H. Davis (Eds), Understanding Group Behavior: Small Group Processes and Interpersonal Relations (Vol. 2, pp. 11–35). Mahwah, NJ: Erlbaum. Teams, Diversity, and Culture 339

Morgan, B. B., Jr., Bowers, C. A., & Jentsch, F. (1994). Issues in the performance of heterocultural teams. Invited paper presented at the 23rd International Congress of Applied Psychology, July 17–22, 1994, Madrid, Spain. Nemetz, P. L., & Christensen, S. L. (1996). The challenge of cultural diversity: Harnessing a diversity of views to understand multiculturalism. Academy of Management Review, 21, 434–462. Nemeth, C., & Kwan, J. (1985). Originality of word associations as a function of majority vs. minority influence processes. Social Psychology Quarterly, 48, 277–282. Nemeth, C., & Wachtler, J. (1983). Creative problem solving as a result of majority vs. minority influence. European Journal of Social Psychology, 13, 45–55. O’Reilly, C., & Chatman, J. (1996). Culture as social control: Corporations, cults and commitment. Research in Organizational Behavior, 18, 157–200. Perkins, D. N. (1995). Insight in minds and genes. In: R. J. Sternberg & J. E. Davidson (Eds), The Nature of Insight (pp. 495–533). Cambridge, MA: MIT Press. Rigby, J. M. (1987, Fall). The challenge of multinational team development. Journal of Management Development, 6, 65–72. Rosen, B., & Lovelace, K. (1994). Fitting square pegs into round holes. HRMagazine, 39(1), 86–91. Salas, E., Dickinson, T., Converse, S., & Tannenbaum, S. (1992). Towards an understanding of team performance and training. In: R. W. Swezey & E. Salas (Eds), Teams: Their Training and Performance (pp. 3–29). Norwood, NJ: Ablex. Shaw, M. E. (1983). Group composition. In: H. H. Blumberg, A. P. Hare, V. Kent & M. F. Davies (Eds), Small Groups and Social Interaction (Vol. 1). Chichester, England: Wiley. Smith, S. M. (1995). Getting into and out of mental ruts: A theory of fixation, incubation, and insight. In: R. J. Sternberg & J. E. Davidson (Eds), The Nature of Insight (pp. 229–251). Cambridge, MA: MIT Press. Smith-Jentsch, K. A., Baker, D. P., Salas, E., & Cannon-Bowers, J. A. (2001). Uncovering differences in team competency requirements: The case of air traffic control teams. In: E. Salas, C. A. Bowers & E. Edens (Eds), Improving Teamwork in Organizations: Applications of Resource Management Training (pp. 31–54). Mahwah, NJ: Lawrence Erlbaum. Stasson, M. F., & Bradshaw, S. D. (1995). Explanations of individual-group performance differences: What sort of “bonus” can be gained through group interaction? Small Group Research, 26, 296–308. Steiner, I. (1972). Group process and productivity. New York: Academic Press. Thorne, P. (1986). The European manager, a man, or a myth? International Management, 41(10), 90. Triandis, H. C. (1987). Dimensions of cultural variation as parameters of organizational theories. International Studies of Management & Organization, 12(4), 139–169. Triandis, H. C. (1993). Collectivism and individualism as cultural syndromes. Cross-Cultural Research, 27(3/4), 155–180. Tuckman, J., & Lorge, I. (1962). Individual ability as a determinant of group superiority. Human Relations, 15, 45–51. United States Department of Labor (1995). Employment outlook: 1994–2005 job quality and other aspects of projected employment growth. BLS Bulletin 2472. Watson, W. E., Johnson, L., & Zgourides, G. D. (2002). The influence of ethnic diversity on leadership, group process, and performance: An examination of learning teams. International Journal of Intercultural Relations, 26, 1–16. Watson, W. E., Kumar, K., & Michaelsen, L. K. (1993). Cultural diversity’s impact on interaction process and performance: Comparing homogeneous and diverse task groups. Academy of Management Journal, 36(3), 590–602. 340 FLORIAN JENTSCH ET AL.

Williams, E. D., Hayflich, P. F., & Gastomn, J. (1986). The challenges of a multicultural work force. Personnel Journal, 65(8), 148–149. Wilkins, A. L., & Ouchi, W. G. (1983). Efficient cultures: Exploring the relationship between culture and organizational performance. Administrative Science Quarterly, 28, 468–481. Williams, W. M., & Sternberg, R. J. (1988). Group intelligence: Why some groups are better than others. Intelligence, 12, 351–377. Yetton, P. W., & Bottger, P. C. (1983). The relationship among group size, member ability, social decision schemes, and performance. Organizational Behavior and Human Performance, 32, 145–159. 12. CHALLENGES AND APPROACHES TO UNDERSTANDING LEADERSHIP EFFICACY IN MULTI-CULTURAL TEAMS

Eduardo Salas, C. Shawn Burke, Jennifer E. Fowlkes and Katherine A. Wilson

ABSTRACT

Fostered by technological developments and globalization, culturally diverse teams are becoming a mainstay of organizational strategy. As the use of multi-cultural teams continues to increase, it becomes paramount to understand the mechanism(s) by which leaders can promote effectiveness within these teams. Despite this need, there are numerous challenges facing those who seek to understand these phenomena and move science and practice forward. The purpose of this chapter is to present a few of these challenges and approaches which can assist in mitigating these challenges. Finally, we identify what we see as key research needs within this area.

INTRODUCTION

On January 25th 1990 a perfectly operational aircraft fell out of the sky due to fuel exhaustion killing 73 passengers and crew. The Boeing 707 that originated in

Cultural Ergonomics Advances in Human Performance and Cognitive Engineering Research, Volume 4, 341–384 Copyright © 2004 by Elsevier Ltd. All rights of reproduction in any form reserved ISSN: 1479-3601/doi:10.1016/S1479-3601(03)04012-8 341 342 EDUARDO SALAS ET AL.

Columbia was enroute to New Yorkwhen it crashed 15 miles short of its destination (Helmreich, 1994). Investigations into the crash reported that failures in teamwork and national culture were potential casual factors in the crash. While the captain exhibited failures in leadership by neglecting to state an emergency, junior crew members also failed in their duties by not challenging the captain when he failed to make a clear decision (i.e. assertiveness). Moreover, cultural tendencies may have interacted with these failures to produce such a situation. Specifically, Columbia is a culture that has typically been rated as being high in power distance (making it less likely that junior crew members will challenge those more senior) as well as being rated collectivistic (Hofstede, 1980). As the captain was collectivistic he may have been reluctant to put his aircraft before that of other aircraft waiting to land. Finally, the interaction between the Columbian culture of the crew and the American culture of the air traffic controllers might also have played an important role in the crash. Communication between the two cultures resulted in unclear exchanges and a focus of the crew on ensuring a clear understanding of its meaning, leading to a lack of information concerning the actual “state” of the aircraft. In 1996, the military used a team of multi-national forces to help in the recovery effort of an antitank mine explosion in Bosnia. Within this highly critical, time-dependent operation, multi-national forces witnessed several potential decrements in teamwork. For example, team members tended to deviate from the predetermined plan by taking it upon themselves to go beyond their duties, thereby eating into precious time (Cherrie, 1998). In particular, several Swedish personnel began hand sweeping a previously cleared area. Similarly, the Russian team went beyond the specified area and began disarming antipersonnel mines by hand and discarding them in the nearby woods. These actions not only served to make time more limited, but led to ambiguity in terms of team member actions and a lack of shared situational awareness. While the above actions had the potential to lead to disastrous consequences and lives lost, the leadership capabilities of the commander served to quickly correct deviant actions and keep the team on track. As evidenced above, multi-cultural teams: (a) provide many opportunities for misunderstanding and conflict (Helmreich, 2000); (b) result in difficulties in communication; and (c) take extra effort to be understood. However, the trend towards composing and managing multi-cultural teams to tackle global problems, conflicts, and obligations is rising. By multi-cultural teams we are referring to two or more individuals from different national cultures who must interact interdependently and adaptively toward a common goal while embedded within an organizational/environmental context that will influence team process and outcomes (Hackman, 1987; Orasanu & Salas, 1993; Salas, Dickinson, Converse & Challenges and Approaches to Understanding Leadership Efficacy 343

Tannenbaum, 1992). For this paper, we argue that the terms multi-cultural, multi- national, and cross-cultural are synonymous and will use them interchangeably. Often these teams: (a) have members whose roles and expertise are distributed; (b) may be hierarchical in nature; and (c) may have a limited life span. As the trend towards using multi-cultural teams rises, it becomes imperative to better understand the factors that impact their effectiveness. While the last twenty years have witnessed much progress in understanding the factors contributing to team effectiveness, most of the work has been conducted within the United States and doesn’t address the impact that diverse national cultures may have on teamwork processes (Earley & Gibson, 2002; Gibson, 1999; Gibson & Zellmer-Bruhn, 2001; Gomez, Kirkman & Shapiro, 2000; Harrison, McKinnon, Wu & Chow, 2000). Furthermore, although team leadership has often been cited as a key determinant of a team’s effectiveness (Stewart & Manz, 1995), little is known about team leadership within multi-cultural teams. Therefore, the focus of this paper will be to examine team leadership within multi-cultural teams. First, we will outline the many challenges researchers and practitioners will face when trying to understand leadership within multi-cultural teams. Second, we will outline several approaches that can be used to one’s advantage in furthering our understanding of this domain. This will be accomplished by examining the literature on: (a) cultural diversity; (b) multi-cultural/multi-national groups/teams; and (c) the sparse literature on leadership within culturally diverse settings. Third, based on the challenges and approaches identified, research needs and future directions will be identified.

CHALLENGES

We have identified several challenges that researchers and practitioners will need to consider in further investigating leadership within culturally heterogeneous teams, as well as within the development of interventions. We have grouped these challenges into the following four categories: (a) what theories should guide the needed research; (b) how do we define the problem and operationalize relevant constructs; (c) what methods should be used to study leadership within multi-national teams; and (d) what interventions should be used.

Challenge No. 1: What Are the Theories That Should Guide the Research?

Lewin (1951) argued that nothing is more practical than a good theory. This argument remains strong more than half a century later. In order to develop sound 344 EDUARDO SALAS ET AL. research and resultant guidance in terms of training and leadership behaviors, empirical work must be grounded in theory. Therefore, the first challenge that one faces in the examination of team leadership within multi-cultural teams is looking for theoretical guidance. As we are talking about leading teams, it is natural to look to theories of teams and team leadership.

Team Literature

Theories abound within the team performance literature and are primarily illus- trated through input-process-output models (see IPO models, Campion, Medsker & Higgs, 1993; Gladstein, 1984; Marks, Mathieu & Zaccaro, 2001; Morgan, Glickman, Woodard, Blaiwes & Salas, 1986; Tannenbaum, Beard & Salas, 1992). Within these models, inputs (i.e. task characteristics, work characteristics, individual characteristics, and team characteristics) impact team throughput (i.e. team processes, coordination, communication, teamwork), which in turn lead to team outputs (i.e. quality, quantity, satisfaction). For a more complete description and review of the above cited models see Salas, Burke and Stagl (in press). In terms of theory, these models and the empirical work behind them have led to many theoretically-based findings. For example, these models illustrate that teams are dynamic, tend to perform multiple processes sequentially and simultaneously, and that the competencies needed for team effectiveness may vary, depending on the team’s stage of development (i.e. life cycle). Second, research has shown that taskwork competencies (i.e. knowledge, skills, attitudes) are not enough, teams also need to possess requisite teamwork competencies (i.e. adaptability, shared situational awareness, performance monitoring and feedback, leadership/team management, interpersonal relations, coordination, communication, and decision-making, Cannon-Bowers, Tannenbaum, Salas & Volpe, 1995; see also Fleishman & Zaccaro, 1992; Marks et al., 2001). Third, team leaders may play a key role in assisting the team in creating and maintaining the shared affect, behavior, and cognition needed for effective team performance (Burke, 1999; Kozlowski, Gully, Salas & Cannon-Bowers, 1996; Marks, Zaccaro & Mathieu, 2000; Zaccaro, Rittman & Marks, 2001). The above theoretical principles are just a few of those that have emerged within the team literature over the past 25 years; for a more thorough review see Salas et al. (in press).

Team Leadership Literature

Despite the argued impact that team leaders can have on teams, there are very few theories that talk about team leadership. Therefore, it is not surprising that our Challenges and Approaches to Understanding Leadership Efficacy 345 search revealed no theories that specifically addressed team leadership and culture. However, we will briefly illustrate two leadership theories and/or approaches that have been elevated to the team level and may be argued to be helpful in understanding the leadership within multi-national teams (i.e. functional approach to leadership and leader-member exchange).

Functional Approach In terms of team leadership within complex environments perhaps the most popular approach to leadership is the functional approach, which builds upon open systems theory (Katz & Kahn, 1978) to identify critical leadership functions. The key assertion within this approach has been argued to be “that [the leader’s] main job is to do, or get done, whatever is not being adequately handled for group needs” (McGrath, 1962, p. 5, as cited in Hackman & Walton, 1986). Note that within this approach the critical factor is not that the leader himself accomplishes the functions, only that he ensures they are completed. In line with the work of Lord (1977), the functional approach to leadership also recognizes that there exists a generic set of leadership functions that can be tailored to the specific situation (i.e. the same set of behaviors may not have the same instrumentality across situations). The discretion allowed to the leader within this approach, in conjunction with the requirement to tailor generic leadership functions to the particular situation, has led many to define leadership as social problem solving (Fleishman, Mumford, Zaccaro, Kevin, Korotkin & Hein, 1991; Mumford, 1986; Zaccaro, Gilbert, Thor & Mumford, 1991). Much of the work that has been conducted on team leadership has used this approach (e.g. Burke, 1999; Burke, Salas, Stagl & Fowlkes, 2002; Fleishman et al., 1991; Hackman & Walton, 1986; Kozlowski et al., 1996; Kozlowski, Gully & McHugh, 1993; Marks et al., 2000; Zaccaro et al., 2001). The interested reader is referred to Salas et al. (in press) for a more detailed review of this approach.

Leader-Member Exchange In addition to the functional approach to leadership, some of the contingency approaches to leadership may also provide a theoretical basis from which lessons regarding how to lead multi-cultural teams can be extracted. Leader-member exchange (i.e. LMX) theory is one of the few models that have begun to be elevated to the team level. The central premise of this theory is that effective leadership occurs when the leader and followers are able to develop mature working relationships characterized by a high degree of mutual trust, respect, and obligation (Graen & Uhl-Bien, 1995). However, leaders have a strong tendency to develop dyadic relationships with followers that are somewhat different (i.e. forming an in-group and out-group). Most recently the argument has been made that LMX should be viewed as a system of interdependent dyads (see Scandura, 346 EDUARDO SALAS ET AL.

1995), where the focus is on examining how differentiated dyadic relationships combine and interact to form larger systems that impact outcomes (see Tesluk & Gerstner, 2002). This theory has just recently been elevated to a systems perspective that includes teams, but again does not yet consider the impact that national culture may have on the development of these dyadic relationships, nor their interaction within a team environment. However, we would argue that within teams the formation of these types of relationships are crucial in order to form the shared cognitive structures (i.e. mental models) that allow meaning to be assigned to member actions (in terms of both the leader and members).

Challenge No. 2: How Do We Define the Problem and Operationalize Construct(s)?

No one would argue that the study of culturally diverse teams is not a complex issue, as there are a multitude of factors that may interact to determine their effectiveness. Therefore, once theoretical drivers have been investigated, the next challenge is identifying important independent variables that are crucial for the development of models to guide training and research. Therefore, team leadership researchers must determine what variables to focus on and how to isolate these variables from other contextual factors. Researchers must also ensure that the operationalization of critical variables are included in write-ups so that results are comparable across studies and nomological nets can begin to be constructed within this domain. Nomological nets are essential to furthering problem development and understanding in that they depict the representation of concepts of interest, their observable manifestations, and the interrelationships between concepts (Cronbach & Meehl, 1955). Some of the challenges in determining how to define the problem and choose efficacious variables to focus on include, but are not limited to the following: (a) how to define cross-cultural teams; (b) at what level should culture be examined (i.e. individual or national); and (c) what are the important task and team variables to initially focus on?

How to Define Cross-Cultural Teams One of the initial issues that will need to be reconciled in defining the problem is what is meant by “multi-cultural team.” More specifically, are you examining teams that are internally homogeneous with respect to culture, but compare across different teams (e.g. compare an American team to a Japanese team) or teams that have within team cultural diversity (e.g. a team comprised of both Americans and Japanese)? Within much of the current literature, comparisons are made Challenges and Approaches to Understanding Leadership Efficacy 347 between homogeneous groups, providing a picture of interactions within teams of different cultures (if results are compared across teams). For example, Conyne, Wilson, Tang and Shi (1999) examined discussion group member interactions within American and Chinese teams. In Chinese teams, members were more likely to direct communications to the leader vs. to other group members, while the opposite was found in American teams. There were also differences between Chinese and American team members in leadership approaches, team member communication styles, and team members’ willingness to speak up. While this research is instructive for leadership in culturally homogeneous teams it offers little insight into how Chinese and American team members are likely to interact within the same team. Solely studying teams that have within team homogeneity fails to provide information about how culturally distant values and preferences interact within a team setting. For example, what is the impact of having an individualist and collectivist on the same team? Furthermore, how does this factor interact with the situation in which there are differences in uncertainty avoidance or analytic/holistic reasoning (see Table 1 for definitions of these factors)? Researchers and practitioners alike should not narrow their focus to only studying cultural differences across teams as this will limit our knowledge. The researcher must decide and describe which operationalization of multi-cultural teams s/he chooses for the problem definition.

At What Level to Measure Cultural Differences? Another question researchers will have to contend with is, “at what level are cultural differences measured?” For example, it has been argued that there are large individual differences within culture on dimensions that have been used to provide theoretical underpinnings of cultural research – for example, individual- ism (Hofstede, 1980). Similarly, Triandis (1989) has documented within-culture individual differences. Cultural distance is a perceived dissimilarity among team members; a large perceived cultural distance may impede team member interaction. Triandis argues that when cultural distance is large it may be better to segregate ethnic groups and select members from the groups who are more similar to interact. However, much of the research that is within the current literature measures culture at the national level, disregarding the reported findings concerning differences within culture. In addition, others have argued that other factors may override these differences (Hutchins, Holder & Perez, 2002).

What Variables to Examine? A third and final factor that researchers will face is to determine the relevant “contextual” factors to investigate. Trends within the cross-cultural and diversity literature(s) tend to indicate conflicting findings concerning the impact of 348 EDUARDO SALAS ET AL.

Table 1. Cultural Dimensions.

Cultural Definition Dimension

Individualism “A loosely knit social framework in which people are supposed to take care of themselves and of their immediate families only” (Hofstede, 1980, p. 45). Collectivism A tight social framework “in which people distinguish between in-groups and out-groups, they expect their in-group to look out after them, and in exchange for that they feel they owe absolute loyalty to the in-group” (Hofstede, 1980, p. 45). Power distance “The extent to which a society accepts the fact that power in institutions is distributed unequally” (Hofstede, 1980, p. 45). High power distance cultures accept this and social exchanges are based on this fact. Low power distance cultures do not see a strict hierarchy among social exchanges. Uncertainty “The extent to which a society feels threatened by uncertain and ambiguous Avoidance situations and tries to avoid these situations by providing greater career stability, establishing more formal rules, not tolerating deviant ideas and behaviors, and believing in absolute truths and the attainment of expertise” (Hofstede, 1980, p. 45). Ranges from high to low. Masculinity- “The extent to which the dominant values of society are ‘masculine’ – that is, femininity assertiveness, the acquisition of money and things, and not caring for others, the quality of life, or people” (Hofstede, 1980, p. 45). Long-short term Long term orientation refers to a culture with future orientated values, while short orientation term orientation refers to cultures who are driven by past and present-orientated values (e.g. respect for tradition) (Kirkman et al., under review). High-low “Extent to which members are expected to know about rules and procedures context without being told” (Harper & Rifkind, 1995, p. 49). Cultures where information about procedures are rarely communicated are known as high context (much of the information is implicitly implied and exact meaning is determined by context). Within low context cultures communication is very explicit (Hall & Hall, 1990). Analytic/holistic Holistic reasoning (typical of East Asians) holds the assumption that “everything reasoning in the universe is somehow interconnected” and therefore “an event or object cannot be understood in isolation” (Choi, Dalal, Kim-Prieto & Park, 2003, p. 47). Analytic reasoning (typical of Westerners) holds that everything in the universe is “separate from, and independent of, each other” (p. 47). heterogeneity (see Adler, 1997; Daily, Whatley, Ash & Steiner, 1996; Distefano & Maznevski, 2000; Horenczyk & Berkerman, 1997; Jackson, May & Whitney, 1995; Thomas, 1999; Triandis, 2000; Watson, Kumar & Michaelsen, 1993). The conflicting findings may be due to several factors, some of which include task complexity, developmental stage of team, and level of interdependency among members. Therefore, careful specification of variables of interest becomes critical. For example, regarding task complexity, cultural diversity may facilitate task performance for simple tasks (e.g. Jackson, 1992; Ravlin & Meglino, 1993) Challenges and Approaches to Understanding Leadership Efficacy 349 but result in teams not performing at their full potential within complex tasks (Thomas, 1999) due to difficulties in managing the interdependencies within the team and coordinating the actions of several people (i.e. process loss, Steiner, 1972). Similarly, initial research has suggested that multi-cultural teams are likely to have increased process loss as tasks become less structured and more uncertain – what Ilgen and colleagues (Ilgen, LePine & Hollenbeck, 1997) refer to as “weak” tasks. Developmental stage is another variable which has been argued to impact the performance of multi-cultural teams. Watson et al. (1993) observed performance improvements over time in solution quality for case study tasks. Comparing heterogeneous to homogeneous teams, it was found that while performance differ- ences favoring homogeneous teams were evident at the outset, these differences diminished until at the end of the study, no differences could be detected. With regards to the level of interdependence, Gibson (1999) found that task uncertainty and level of interdependence required moderated the relationship between group efficacy and group effectiveness. Group efficacy, the group’s belief that it can perform effectively, was related to performance only for low uncertainty tasks and tasks requiring team member interdependency. Finally, Adler (1997) suggests that diversity may be advantageous to task performance when team members have highly specialized roles and when innovation is required. These are a few of the factors that have emerged within initial research. Researchers will need to consider how to set limits on their investigations in terms of these factors, as well as others that have yet to emerge.

Challenge No. 3: What Methods Are Used to Study Leadership of Multi-cultural Teams?

The complexity contained within the domain of cross-cultural team leadership not only poses a challenge in terms of specifying the problem, but also in determining the methods by which to best study this issue. Gelfand, Raver and Ehrhart (2002) outline desired methodological characteristics for cross-cultural research. These include:
The use of well-developed theories to guide the development of research questions, measures, and the sampling of cultures.
Sampling multiple cultures (i.e. more than two) within a study.
Employment of methods to guard against the use of etic constructs (i.e. constructs that have been developed in one culture but that might not be meaningful, or carry different meaning, in another culture).
Choosing culturally appropriate methods. 350 EDUARDO SALAS ET AL.

Given this guidance, a question we posed was how does the extant research stack up against these recommendations? What we found was that there was lots of room for improvement.

Convenience First, sampling is often dictated by convenience, rather than by theory. We found that the selection of individuals within cultures is mostly limited to undergraduates and in contrast to Gelfand et al.’s (2002) recommendations, comparisons are often made between two cultures. Also in terms of convenience samples, Oyserman, Coon and Kemmelmeier (2002) report that in cross-cultural research comparisons are typically made between European American undergraduates and East Asian undergraduates, thus limiting the generalizability of the research findings. Another aspect of methodology that is often dictated by convenience is the fact that research rarely follows Hofstede’s (1980) lead to provide an actual assessment of how the countries of interest stand on key cultural dimensions. Oyserman et al. (2002) point out that conducting studies at the country level calls for tremendous resources, requiring, for example, sampling of representative organizations and individuals. More common is the use of Hoftstede’s (1980) country-level ratings as proxies for cultural dimensions rather than directly assessing the construct (Oyserman et al., 2002). Another common practice is to assess constructs on an individual level (i.e. through ratings) or by using priming techniques. Priming techniques are based on the assumption that assessable knowledge guides behaviors, whether it is chronically available or temporarily accessible because of a recent experience. Thus, priming studies create temporarily accessible differences on key cultural constructs and assess their impact on behaviors (Oyserman et al., 2002).

Methods Used The study of culture is dominated by survey research (Gelfand et al., 2002). While surveys have the advantage of being fairly simple to administer to a large number of respondents they comprise only one method, and thus bring the strengths and weaknesses inherent in the method. In cross-cultural research, limitations include differences among cultures for how instructions and terms are understood, motivation, and response sets (Gelfand et al., 2002; Triandis, 1983). In short, surveys (as well as any potential method) do not work equally well in all cultures. As an example of this, Gelfand et al. related an incident encountered by Moshinsky (2000):

During survey administration in a cross-cultural study of acculturation, Moshinsky (2000) found that, despite the instructions to work independently, Russian participants were working collaboratively to answer a survey. As a group, they were reading the questions aloud, Challenges and Approaches to Understanding Leadership Efficacy 351

deciding upon a group answer, and all circling the same response. These participants found the individual survey methodology to be inconsistent with their cultural experiences and values, and therefore modified the instructions to the more culturally appropriate group consensus task (p. 226). The dangers of over reliance on a particular method point out the need for employing multiple methodologies, selected for their appropriateness to the stage of research and question(s) being addressed. Other cross-cultural research methodologies include: knowledge elicitation techniques, direct observation, analysis of cultural artifacts (e.g. newspaper articles), and experimentation. While there are a variety of methods to choose from, there has generally been little use of these alternative methods.

Challenge No. 4: In Terms of Interventions, What Are the “Best Practices?”

A fourth challenge confronting researchers and practitioners is to identify the best practices (i.e. interventions) to facilitate effective team leadership within multi-cultural teams. The challenge here is severalfold. First, after searching the literature we found that the predominant number of interventions currently being used focus solely on the multi-cultural aspect of the problem. Most of the training strategies (and resulting training programs) that were identified primarily focused on increasing multi-cultural awareness, with a few offering trainees strategies by which to combat the challenges and potential misinterpretations that occur within a multi-cultural environment. These programs tend to vary in the degree to which they are experientially based, but we found few that focused on the more specific topic of preparation for inclusion in multi-cultural teams, and almost none on preparing leaders to deal with the potential process difficulties within these types of teams. Second, individuals progress through a variety of stages in terms of acculturation when living in another culture (i.e. denial, deference, minimization, acceptance, adaptation, and integration) and may be at different stages of awareness in terms of interacting with cultures different from their own (Bennett, 1986). Training strategies may be differentially appropriate depending on where each individual is in terms of awareness and interactions with individuals from different national cultures. However, this has not yet been systematically investigated. Third, the predominant amount of training programs that were identified within our search revolved around training individuals who were going to live and work in another country (i.e. expatriates). This brings us to the question as to whether individuals who are living in their own culture but working on a multi-national team whose members have different cultural values and orientations need the 352 EDUARDO SALAS ET AL. same type or level of training as do people who are actually to be placed within another culture to work. Fourth, there is the question as to whether cross-cultural training programs should be teaching culturally specific or generic competencies. For example, culture assimilator training (Triandis, 1995) initially focused on teaching cultur- ally specific skills, but this work has now evolved into several generic culture assimilator training programs (Brislin, Cushner, Cherrie & Yong, 1986). While cultural specific programs may represent a more targeted focus, they also assume that there are time and resources to develop the programs and train personnel prior to going into each new situation. Many real world examples indicate that this is not true. For example, members may be distributed, on vacation, assigned to other projects, or otherwise unavailable. The use of culturally specific training programs also suggests that training would have to be tailored to each culture, representing more dollars in terms of development. On the other hand, are there situations where these culturally specific training programs are essential (e.g. consider the high consequences of mistakes)? Finally, while the most common intervention may be training there are also other alternatives (i.e. selection) that may prove more beneficial depending on the flexibility and resources afforded to organizations.

Summary of Challenges

Understanding the dynamics that occur within culturally diverse teams poses several challenges. Throw in team leadership and the complexity grows. However, as the use of such teams continues to rise within industry, as well as the military, and the predominant number of these teams will have some form of leader, it is imperative to better understand how the leader may create and maintain team effectiveness. The first step in understanding these phenomena is the development of theoretical frameworks that will serve to systematically guide the research. As there currently is a lack of such frameworks, this should be the first step. However, there are also several other challenges facing researchers and practitioners who want to understand team leadership within culturally diverse teams. These include: (a) identification of the theories which should guide the research; (b) criteria for problem definition and identification of critical variables; (c) determination of the best methods to use; and (d) choice of the most appropriate interventions. Despite the many challenges, there are also several bodies of work that can generate useful approaches. While these approaches (see below) do not provide full answers they offer a start in the right direction. Challenges and Approaches to Understanding Leadership Efficacy 353

APPROACHES

Though there is not a plethora of theoretical guidance regarding the problem space of multi-cultural team leadership, the picture is not as bleak as it may have first seemed. There are several literatures that while not providing “perfect” answers to our questions can provide initial starting points: (a) culture; (b) culture and groups or teams; and (c) leadership and culture. These literatures can help further efforts in forming a theoretical framework for understanding team leadership within culturally-heterogeneous teams. Furthermore, findings can also be extrapolated from the wealth of information contained within the team performance and measurement literature(s) to further the scientific study of leadership within multi-cultural teams.

Approach No. 1: What Do We Know About Culture and Its Influence?

Culture has been defined as consisting of the shared norms, values, and practices of a nation (Helmreich, 2000). Similarly, others have argued that culture is the shared perception of the self and others, consisting not only of norms and behaviors, but also beliefs which serve to provide structure for member action (Dodd, 1991). Perhaps the most relevant finding in terms of the impact of culture is actually twofold. First, national cultures do differ in terms of their values and preferences of action and cognition (see Hofstede, 1980; Kluckhohn & Strodbeck, 1961; Parsons & Shils, 1951; Triandis, 1989). Second, these differences have a tendency to make multi-cultural interactions frustrating due to misunderstandings revolving around communication and the interpretation of action, as well as differences in the way individuals from different cultures behave. If managed properly, it has been argued that cultural diversity can be a large competitive advantage for organizations (e.g. Earley & Gibson, 2002), but if not, it can result in a large liability.

Value Dimensions Although there are many conceptualizations of cultural dimensions that nations and individuals differ upon (see above), perhaps the most prominently used is the work of Hofstede (1980). While this work has been criticized along several lines, including the following: (a) most of the data was collected from middle-level managers in one multi-national corporation; (b) it was done over 20 years ago; and (c) some of the cultural dimensions proposed are highly inter-correlated, we choose to review it for several reasons. First, it is still the most common metric and therefore has the most empirical work from which practitioners and researchers 354 EDUARDO SALAS ET AL. can draw upon. Second, it has been possible to replicate some of the dimensions within his typology in more recent work, suggesting some stability (Gomez et al., 2000; Helmreich, 2000; Kirkman, Lowe & Gibson, under review; Oyserman et al., 2002). Hofstede and colleagues identified five cultural dimensions: (a) individualism- collectivism; (b) power distance; (c) uncertainty avoidance; (d) masculinity- femininity, and (e) long-short term orientation (see Table 1). We will briefly describe the first three of these dimensions as we feel they are most relevant in terms of team leadership in culturally diverse teams. Of these dimensions, individualism-collectivism has consistently received the most empirical support (see Oyserman et al., 2002). Individualism has been defined as, “a loosely knit social framework in which people are supposed to take care of themselves and of their immediate families only” (Hofstede, 1980, p. 45). Within these cultures the individual’s rights and goals are considered more important than team goals. Conversely, Hofstede describes collectivism as being “...characterized by a tight social framework in which people distinguish between in-groups and out-groups, they expect their in-group to look after them, and in exchange for that they feel they owe absolute loyalty to it” (p. 45). This dimension is hypothesized to have an impact on leadership in that it will impact the way in which the leader needs to build team cohesion, deliver feedback, and promote common goals. The second dimension of culture, power distance, has been defined as “the extent to which a society accepts the fact that power in institutions is distributed unequally” (Hofstede, 1980, p. 45). Individuals from high power distance cultures respect those of higher status and often will not challenge the ideas or opinions of others. This is likely to impact leadership in regards to the leadership prototype held by group members as well as openness of the team’s climate and the willingness of subordinates to question the leader’s assumptions and offer constructive feedback. The third dimension, uncertainty avoidance, has been defined as, “the extent to which a society feels threatened by uncertain and ambiguous situations and tries to avoid these situations by providing greater career stability, establishing more formal rules, not tolerating deviant ideas and behaviors, and believing in absolute truths and the attainment of expertise” (Hofstede, 1980, p. 45). Individuals from high uncertainty avoidance cultures find uncertainty threatening or stressful. Uncertainty avoidance is hypothesized to impact nearly all aspects of leadership as leaders essentially serve to make sense of the environment, integrate this information into their own cognitive structures, and then impart this meaning to team members. Uncertainty avoidance is hypothesized to impact the types of information leaders look for and the frequency of their searches. In addition, the structure of the task, goals, and feedback may have to be differentially tailored Challenges and Approaches to Understanding Leadership Efficacy 355 dependent on the level of uncertainty avoidance in the team in order to provide direction that is perceived as clear and engaging (i.e. motivating). In addition to Hofstede’s dimensions, we add two others that we argue are relevant in terms of leadership within multi-cultural teams: (a) high-low context and (b) analytic/holistic reasoning (see Table 1). A high context culture is one where information about procedures is rarely explicitly communicated because the context serves as the behavioral cue for appropriate action (Harper & Rifkind, 1995). Here the context (e.g. tone, hand gestures, body language) surrounding the message is very important (Triandis, 2000) and verbal style is generally indirect and implicit (Hodgetts & Luthans, 2000). Conversely, communication in low context cultures is: (a) full of information concerning procedures and expectations; (b) generally direct and explicit (Hodgetts & Luthans); and (c) focused on the content of the message (Triandis). As clear communication is a foundation of effective teams and their leadership, this cultural dimension is hypothesized to have a tremendous impact on team leadership. The last dimension concerns cultural differences in reasoning. Holistic reasoning emphasizes the relationships and connectivity among objects (Ji, Penge & Nisbett, 2000). Individuals who reason holistically (e.g. Easterners) often see an object as a part of the larger whole and therefore focus on the field or situation surrounding an object or event when making causal attributions (i.e. situational attribution; Ji et al.; Lee, Hallahan & Herzog, 1996). On the other hand, Westerners are argued to reason analytically and isolate the object as separate from the whole (Ji et al.). Causal attributions are then focused on the properties of the object or situation, rather than the larger field (i.e. dispositional attribution; Choi & Nisbett, 2000). This final dimension is argued to impact leadership and team functioning because it relates to how individuals assign attributions for member actions as well as determining what is “seen.”

Approach No. 2: What Do We Know About Teams and Culture?

Conceptual Models Earlier in the paper we argued for the importance of work being theoretically driven. Often times within the team literature this takes the form of conceptual models. We were able to identify one model that begins to incorporate how culture may impact teams, although it does not heavily focus on team processes. Klein, Pongonis and Klein (2000) proposed a Cultural Lens model for enhancing the interactions of cross-cultural team members. This model is based on notions from Triandis (1994) that culture provides a “lens” through which individuals see the world. The Cultural Lens model holds that if team members are able to view the 356 EDUARDO SALAS ET AL. world through the eyes of their culturally diverse team members, and understand their cultural perspective, then improved interactions may result. As the model is planned for use in driving training for military command and control teams, Klein and colleagues focused the model on cultural dimensions relevant to that domain (i.e. power distance, Hofstede, 1980; uncertainty avoidance and risk assessment, Dorfman & Howell, 1988; Hofstede, 1980; activity orientation, Kluckhohn & Strodbeck, 1961; dialectical reasoning, Peng & Nisbett, 1999; and counterfactual thinking, Tetlock, 1998). The goal is to develop training that will produce leaders and team members who are attuned to cultural differences. While a starting point, the model focuses almost exclusively on knowledge components to the neglect of behavioral and attitudinal components. On the other hand, in support of approaches like the Cultural Lens Model, Conyne et al. (1999) argue that knowledge of cultural differences is a prerequisite for effective leadership in cross-cultural teams, enabling leaders to act with “culture appropriateness,” interpret communication patterns and build and expand the team’s interaction styles. Triandis (1995) argues that the effectiveness of such approaches will depend on the cultural distance among team members. Specifically, when cultural distance is small, then learning about another culture makes individuals in one culture more similar to the members of the other group, and can improve interactions. However, the opposite may occur if cultural distance is large. In this case, members of groups may seem even more different. Ilgen et al. (1997) argue that team building interventions include the assumption that if team members appreciate each other’s differences, improved interactions will result. However, there is little support for the effectiveness of team building. In addition, Ilgen and colleagues argue that appreciating cultural differences does not eliminate differences in perspective, which are still problematic for team functioning. Thus, empirical work is needed to further examine the efficacy of such approaches.

Teamwork Dimensions Although there appears to be a limited amount of conceptual models to drive theoretical thinking concerning team leadership within multi-cultural teams, there has been a limited amount of empirical work conducted on the impact of culture on teamwork. The empirical evidence that was found is organized around teamwork dimensions as argued for by Cannon-Bowers et al. (1995) and may serve as a starting point for hypothesis generation (see Table 2). Table 2 depicts empirical findings regarding the eight process dimensions, but as communication (i.e. information exchange) has been cited as creating the largest problem for culturally diverse teams, we choose to elaborate upon the findings for that dimension within this section. This is also the area where the most work has been conducted. Challenges and Approaches to Understanding Leadership Efficacy 357

Table 2. Research Evidence on Impact of Culture on Teams.

Teamwork • The concept of teamwork varies across national cultures. Five different metaphors emerged (i.e. military, sports, community, family, and associates) (Gibson & Zellmer-Bruhn, 2001). • Employees generally more receptive to teamwork aspect of self-managing teams when they valued collective rather than individual interests (Kirkman et al., under review). • Crews (a) low in power distance, (b) high in uncertainty avoidance, (c) individualistic produced a trend where they had a greater number of positive teamwork behaviors than collectivist teams (Kuang & Davis, 2001). Adaptability • Members of collectivistic cultures have difficulty adjusting to frequent shifts in team membership (Harrison et al., 2000). • High power distances cultures have difficulty adapting when leadership changes from time to time (Harrison et al., 2000). Shared situation • No explicit research found at team level. awareness Performance • Collectivists evaluated in-group members significantly more generously monitoring than did individualists (Gomez, Kirkman & Shapiro, 2000). and feedback Leadership/team • Chinese leader appeared to be more overtly concerned about group’s management success rather than individual member’s feelings about participation (Conyne et al., 1999). • Sensitivity to group norms was a more important element of leader behavior in Iran and Mexico than in U.S. (Ayman & Chemers, 1983). • Charismatic group leadership more prevalent in collectivistic cultures (Pillai & Meindl, 1998). • No differences in leader communication found between individualistic or collectivistic groups (Conyne et al., 1999). • Regardless of members’ ethnicity, transformational leadership is positively correlated with members self-efficacy, group potency, and perceptions of leader effectiveness (Jung & Yammarino, 2001). However, the associations of leadership with group potency and leader effectiveness is stronger for Asian Americans and with self-efficacy for Caucasian Americans. Interpersonal • Collectivistic groups have fewer conflicts (Oetzel, 1998). relations • Collectivists place a higher value on maintenance contributions, while individualists placed higher value on task contributions (Gomez et al., 2000). • Heterogeneity in uncertainty avoidance within top management teams was positively related to issue based conflict (Elron, 1997). Coordination • Teams high in collectivism behave more cooperatively than those high in individualism (Cox, Lobel, & McLeod, 1991; Wade-Benzoni, Okumura, Brett, Moore, Tenbrunsel & Bazerman, 2002). • Team collectivistic orientation was positively related to team cooperation (Eby & Dobbins, 1997). Communication • Most research conducted here. Expounded upon in text. Decision-making • No explicit research found at team level. 358 EDUARDO SALAS ET AL.

Information Exchange and Culture. Information exchange among team members is a key team process and includes at least three components. Specif- ically, information exchange involves individuals and/or teams actively seeking information from all available resources (i.e. people, equipment, environment) (Smith-Jentsch, Johnston & Payne, 1998). The second component of information exchange involves passing relevant information to the appropriate people in a timely manner without being asked (Smith-Jentsch et al.). Finally, the third dimension consists of individuals and/or teams providing situation updates (Smith-Jentsch et al.). Operating from this definition we found results that indicated that within multi-cultural teams, information exchange is:
Slower due to language barriers and the effort it takes for the calibration of meaning (Cherrie, 1998; Hayles, 1982, as cited in Adler, 1997; Helmreich, 2000).
Lower among coworkers (Jackson et al., 1995; Kirchmeyer, 1993). Ilgen et al. (1997) argue that this finding is likely to combine with the well-established finding in teams that communication is likely to center on topics commonly known to team members, vs. important information held by only one team member (e.g. Stasser & Titus, 1987). That is, this propensity may be exacerbated in cross-cultural teams – team members in cross-cultural teams will talk less frequently, and when they do talk, it is likely to be on a topic that team members know about.
Less likely to contribute to group think or the tendency for team members to select and inflate the efficacy of a decision alternative (Ilgen et al., 1997). Group think is more likely when the team is homogeneous and is isolated from outside influences. These conditions would not be expected within multi-cultural teams, especially those with a short history or that are ad hoc in nature. Group think is also less likely when cohesion is low (Janis, 1972), which might be expected in heterogeneous teams in part because of their short history and national differences. Information Exchange and Specific Cultural Dimensions. In addition to the more general findings, there has also been some research conducted on how the various cultural dimensions impact communication. For example as a function of individualism/collectivism, Conyne et al. (1999) reported that collectivist team members were more likely to direct communications to the leader than other team members, while the opposite was found for individualistic groups. This could negatively impact getting information to the right person(s). Collectivist team members were also more hesitant to provide information, possibly due to a culture-related hesitancy to speak. However, when they did speak, it was for longer time periods than for individualist team members (Conyne et al.), possibly to better explain their point and avoid criticism (i.e. to save face) when a risk was taken to speak. Challenges and Approaches to Understanding Leadership Efficacy 359

Differences have also been found regarding the communication styles (i.e. direct vs. indirect communication) of individualists and collectivists, potentially impacting information transfer. Individualists prefer direct communication styles vs. the indirect style that is associated with collectivists (Gudykunst, Matsumoto, Ting-Toomey, Nishida, Kim & Heyman, 1996). Thus, the mode of interaction may serve as a barrier to communication. A second cultural dimension that has been investigated and is likely to affect information exchange is that of high/low context. There is strong evidence supporting the use of decontextualized causal inferencing among individualists. That is, individualists are more likely to give more weight to dispositional cues (e.g. personnel attributes) while collectivists are more likely to pay attention to situation and context cues in inferring causality. Evidence for this effect is also found within a meta-analysis that yielded medium to large effect sizes, suggesting a robust finding. Evidence for this is also found in the work of Morris and Peng (1994). These researchers compared the attributions provided by American and Chinese graduate students asked to rate 28 possible causes of a murder scenario. American students gave greater weight to dispositional factors while Chinese students gave more weight to situational factors. In related research, the authors also examined newspaper accounts of mass murders and found similar findings. Therefore, while not explicitly tested within teams as of yet, we argue that cognitive style will have a profound impact on information exchange behaviors in that cultures differentially seek or sample different types of information in their exchanges. Other cultural dimensions that may impact information exchange are power distance and uncertainty avoidance. A crucial way in which power distance may impact information exchange is that subordinates may fail to provide critical information to leaders, especially when questioning leaders. This has been doc- umented in the airline industry in flight crews (Helmreich, 2000) and elsewhere (e.g. Ilgen et al., 1997). Uncertainty avoidance may also impact information exchange. Situations that deviate from the plan or from the expected may be avoided, unnoticed, or not communicated because these events are threatening to individuals who have high uncertainty avoidance (Ilgen et al.). Thus, uncertainty avoidance may have the effect of limiting the recognition of cues that alert the team of the need to change their behavior.

Approach No. 3: What Do We Know About Leadership and Culture?

The work that is available to be leveraged against in terms of leadership and culture, is almost exclusively focused on the leadership of individuals and focuses on cultural differences in terms of implicit leadership theories (i.e. each 360 EDUARDO SALAS ET AL. culture’s prototype of what an “effective” leader would look like). However, similar to the work conducted in the team domain, we were able to identify a few models that illustrate the relationship between leadership and culture and will briefly describe those prior to discussing implicit leadership theory (see Figs 1 and 2).

Conceptual Models As alluded to in the preceding paragraph there is a plethora of work on leadership across cultures. While this work serves as a starting point, one must be careful to ensure that the work is theoretically driven. In our search, we found two conceptual models that identified the relationship between leadership and culture (Erez & Earley, 1993; House et al., 1999). House and colleagues (1999) developed a theoretical model that has served as the framework within which the GLOBE (Global Leader and Organizational Behavior Effectiveness research program) research is conducted. The central tenet of this model is that differences in societal culture, norms, and practices will be predictive of the practices of organizations and leader attributes most frequently enacted and effective within a given culture. Several propositions flow from this model, but the most important for our current understanding are the following: (a) cultural values and practices affect what leaders do; (b) leadership affects organizational form, culture, and practice and vice versa; (c) societal and orga- nizational culture influence the process by which people come to share implicit theories of leadership; and (d) leader acceptance is a function of the interaction between culturally endorsed implicit leadership theories and leader attributes and behaviors (see Fig. 1). Erez and Earley (1993) developed a model through which to understand leadership and culture that is loosely based on Yukl’s (1989) integration model, as well as the work of Lord and Foti (1985; see Fig. 2). Perhaps one of the most important aspects of this model is that it recognizes that leaders and followers have reciprocal effects on one another that are embedded in the society within which the organization operates (of which national culture is a component). More specifically, the relationship between leader and follower is determined by each person’s self-representation and its relation to culture. Self-representation is comprised of current traits and/or skills, combined with a cognitive framework that determines how information is interpreted (personal, cultural, self-schema). In terms of the leader, this results in the creation of a vision and the corre- sponding behavior to enact and transmit that vision. In terms of followers, this self-representation leads to the degree to which they identify with the leader and resulting actions. This model also notes that leadership is embedded within the structural, relational, and role-based aspects of the work setting. Challenges and Approaches to Understanding Leadership Efficacy 361 GLOBE Theoretical Model (House et al., 1999). Fig. 1. 362 EDUARDO SALAS ET AL. Leadership and Cultural Self-representation Theory (Erez & Earley, 1993). Fig. 2. Challenges and Approaches to Understanding Leadership Efficacy 363

Cultural Differences in Implicit Leadership Theories The predominant amount of work that has been conducted in regard to leadership across cultures has been conducted primarily in terms of individual leadership and deals with how perceptions of leadership effectiveness differ across cultures. This work has primarily been a result of the GLOBE project (see House et al., 1999) which relies heavily on implicit leadership theory (ILT). Within implicit leadership theory it is individual’s perceptions about leadership which are important. As ILT addresses the evaluations that people make about leaders and the cognitive processes underlying these evaluations, it is possible for someone to possess specified leader behaviors, yet not be deemed as a leader (House & Aditya, 1997). Using this approach, results from the GLOBE project indicate that there are globally endorsed theories of implicit leadership, as well as some that are culturally contingent. This project identified six global leadership dimensions of culturally-endorsed implicit leadership theories: charismatic/value based, team orientated, self protective, participative, humane, and autonomous. Findings indicated that: (a) charismatic/value based and team-orientated leadership styles are universally endorsed; (b) humane and participative styles are nearly universally endorsed; and (c) the two remaining dimensions are culturally contingent. Results also suggest high within-culture agreement in terms of implicit leadership styles (see also Brodbeck, Frese, Akerblom, Audia, Bakacsi & Bendova, 2000). Results from GLOBE research also identified 21 leader attributes and behaviors that were universally viewed as contributing to leadership effectiveness, with a predominant number of these falling under the global dimension of charis- matic leadership. A representative sample of these attributes includes: trust, encouraging, motive arouser, communicative, coordinator, and team builder (see also Den Hartog, House, Hanges & Ruiz-Quintanilla, 1999). Research has also indicated eight attributes that were universally viewed as negative (e.g. loner, noncooperataive, nonexplicit, dictatorial) and 35 that were culturally contingent attributes/behaviors (e.g. cautious, formal, independent, indirect, risk taker, sensitive, intra-group conflict avoider). See House et al. (1999) for further information. Others have also found evidence for the idea that different cultures have various prototypes of what constitutes an “effective” leader (Bass, 1990; Hofstede, 1993; Triandis, 1993), as well as what dimensions are viewed negatively across cultures (see laissez-faire leadership; Bass, 1997). Finally, in terms of leadership style, researchers have begun to investigate how cultural value orientations may link to the propensity for certain leadership styles (Sarros & Santora, 2001). One implication of this work is that “what is expected of leaders, what leaders may or may not do, and the status and influence bestowed upon leaders vary considerably as a result of the cultural forces in the countries or regions in which leaders 364 EDUARDO SALAS ET AL. function” (House et al., 1999, p. 178). However, an important point to note is that even if a leadership style is culturally endorsed, that does not imply that the ways the behaviors are enacted across cultures are the same.

Competencies While the predominant amount of work has been done on implicit leadership theories and perceptions of effectiveness, there has been some work done on actual effectiveness. For example, Hazucha, Hezlett, Bontems-Wackens and Ronnqvist (1999) found that management competencies were more similar than cultural values across countries because the nature of the managerial work tends to be similar across countries [USA, Europe – France, Germany, Italy]. These researchers also report more differences between countries in actual behavior (management skill ratings) than in ideal behavior (importance ratings). Conversely, in a comparison of supervisor ratings of managers in the U.S. and Europe, Robie, Johnson, Nilsen and Hazucha (2001) found more similarities than differences in terms of the skill dimensions important for leadership effectiveness across culture. However, they did find that depending on how effectiveness was operationalized, different dimensions were rated as important. Others have begun to investigate how leadership dynamics may differ, depending on whether the leader is operating in a collectivist setting or high power-distance setting, and how the two may interact (see Schermerhorn & Bond, 1997). Researchers are also beginning to examine how different power bases may have differential impacts dependent on national culture (see Stoeberl, Kwon, Han & Bae, 1998).

Approach No. 4: What Do We Know About Cultural Interventions That Are Currently Being Used?

Current theories and related empirical research are not the only things that can serve as starting platforms in the search to understand team leadership within culturally diverse teams. We can also extract methods, tools, and instructional strategies from the training literature. In our review of the cross-cultural literature, we identified many suggestions for the type of information that should be contained within cross-cultural training programs, as well as several interventions (e.g. formal training, informal intervention, selection). We review a representative sample of this work next.

Content In terms of information to include in training, Elmuti (2001) argues that, in dealing with a cross-cultural workforce, managers should develop programs that: (a) Challenges and Approaches to Understanding Leadership Efficacy 365 promote cultural awareness; (b) promote positive attitudes towards differences; (c) recognize common links; (d) encourage flexible communication; and (e) promote a climate where concerns and questions can be expressed. Grahn and Swenson (2000) argue that while understanding the concept and practice of culture is important, trainees must first: (a) understand their own culture and biases and (b) possess culturally appropriate communication skills. Finally, Wentling and Palma-Rivas (1999) interviewed diversity experts regarding content of diversity programs. Experts indicated that programs should: (a) begin with increasing awareness of what the concept of diversity is and why it is important; (b) include skills training and application of strategies; and (c) provide support for new skills.

Formal Training There are a multitude of formal training interventions that are used to train cross-cultural awareness and skills within organizations, as well as within the military. Below is a representative sample of these interventions grouped into four primary areas (i.e. lecture, culture assimilators, cultural self-awareness, and experiential exercises). Lecture. The use of this method entails a trainer lecturing trainees about the target culture (i.e. history, geography, religion, people, business way of life). This method is still commonly used, but when used alone, it has several disadvantages. For example, while simple to implement and inexpensive, this method relies on passive learning. Research on human learning has indicated that information is better encoded when learning is active. Secondly, trainees are typically provided with all the information they need in a tight package, including a well-defined problem. This does not mimic the environment in which they will need to use these skills, for in reality, trainees will be faced with ambiguous problems and situations where they have to collect their own information as well as define the problem. Third, within these types of training programs, trainees are generally taught to be rational and unemotional. However, cross-cultural situations are often charged with emotion. Finally, these types of programs tend to focus on written skills, as opposed to verbal communication or doing. Culture Assimilators. Culture assimilator programs represent a popular type of cross-cultural training. These programs are based on the work of Triandis (1995) and represent a training tool that consists of a number of real life scenarios. Scenarios are created based on critical incidents and represent puzzling cross-cultural interactions. At the end of the critical incident, a question is posed that asks the trainee to reflect on the scenario and think about the source of the misunderstandings. Trainees are then presented with a range of four to five alternatives that differ in behavioral plausibility and told to choose one. For each alternative, an explanation is provided, indicating why it was correct/incorrect in the current situation. 366 EDUARDO SALAS ET AL.

Recently, culture-general assimilators have been developed (see Brislin, Cushner, Cherrie & Yong, 1986) that are more theoretically-driven. These programs cover 18 themes that have appeared in the literature as important in the context of living abroad and are organized around three broad headings: people’s intense feelings (e.g. anxiety, belonging, ambiguity, disconfirmed expec- tations), knowledge areas (e.g. work, time/space, language, roles, importance of group/individual, values), and bases of cultural differences (e.g. categorization, differentiation, in-group/out-group, attribution, learning style). Cultural assimi- lators represent a convenient self-learning tool and have produced a considerable amount of research evidence as to their effectiveness (see Gudykunst & Hammer, 1983; Malpass & Salancik, 1977). These methods have been found: (a) to have an impact on the ability to retain and generalize cultural information (O’Brien & Plooji, 1977); (b) to be effective in many varied situations (target cultures, training lengths, study setting, different populations); and (c) to have greater impact on cognition than on behavior or affect (Bhawuk & Brislin, 2000). Finally, this method has been successfully used both within industry (O’Brien, Fiedler & Hewett, 1970) as well as within the military (Worchel & Mitchell, 1972). Cultural Self-Awareness. Within these types of programs, trainees see a demon- stration of a behavior that is completely opposite to their own. These programs may have participants interact with actors that display these different behaviors or may use behavioral modeling, in that trainees watch videotaped vignettes and have to identify general themes. If the program involves interaction with role players then interaction is videotaped and trainees are debriefed at the end. Conversely, if participants watch taped vignettes and generate themes, these themes are later compared with those developed by the trainer and a discussion/debriefing session follows. Depending on the structure of these programs, they may be very experientially based. Bennett (1986) argues that these programs: (a) help trainees recognize their own cultural values; (b) analyze the contrast with other cultures; and (c) apply insight to cultural interaction. However, trainees generally don’t learn anything about the host culture, but develop more of an awareness of their own. Experiential Exercises. The central characteristic of these types of exercises is that they involve trainees in the learning process. Cultural awareness training may be one example of an experiential exercise, depending on structure. The most popular is the simulation game. While popular, it has been cautioned that simulation is not always effective in the absence of research evidence (Bhawuk & Brislin, 2000). Similarly, outside the cross-cultural training arena, others have argued that the use of simulation in and of itself does not guarantee learning; it must be built using principles of human learning (see Salas, Bowers & Rhodenizer, 1998; Salas & Cannon-Bowers, 2001). Bhawuk and Brislin argue that simulation Challenges and Approaches to Understanding Leadership Efficacy 367 games may not reach their full training potential due to the artificiality of the simulation. Furthermore, debriefing sessions are not as efficacious as they could be, often times resulting in simple conclusions (e.g. cultures are different). Finally, premature use of simulation (prior to introduction of culture-general frameworks) can result in detrimental training effects (Bruschke, Gartner & Seiter, 1993). In addition to simulation games, a technique known as area simulation is also used within cross-cultural training. Within this type of training, the target culture is simulated in a natural setting and trainees learn skills with minimal guidance from training staff (Trifonovitch, 1977). Within area simulation, as with most experiential exercises, doing is focused on as opposed to merely thinking.

Informal Interventions In addition to formal training, there are informal interventions that leaders can institute to help manage cultural diversity. One alternative is to use a recatego- rization strategy (Gaertner, Mann, Murrell & Dovidio, 1989). More specifically, the leader seeks to highlight members’ commonly shared characteristics, while minimizing the characteristics that differentiate them. Furthermore, as it has been shown that over time members begin to focus more on value differences rather than demographic differences when making categorizations (Harrison, Price & Bell, 1998), leaders may be able to broaden in-group membership by positively influencing work related values of all team members (Gomez et al., 2000). Similarly, Earley and Mosakowski (2000) and Graen and Wakabayashi (1994) have argued that in culturally diverse teams, leaders can create a third “hybrid” culture that serves to reduce some of the potential process losses.

Selection Interventions Leaders are in a unique position to arrange task conditions and make team member assignments to tasks in ways that optimize the use of team diversity or minimize its impact. Selection of multicultural team members is potentially useful because of the natural individual differences on cultural dimensions within a culture (Ilgen et al., 1997; Triandis, 1989). Triandis argues that when cultural distance is large, separation of ethnic groups may be best. He suggests selecting individuals with similar objectivity to work as a team and then to have the two separate teams work together to achieve their goals. This may be especially important for unstructured tasks, or tasks in which there is substantial team member interdependency – for such circumstances, the effects of diversity can be expected to be disruptive. For other tasks, diversity may be advantageous. Adler (1997) suggests that cultural diversity may be advantageous when team members have highly specialized roles and when innovation is required. Thus, it may at times be possible to minimize the negative consequences of cultural diversity through selection and 368 EDUARDO SALAS ET AL. by structuring tasks appropriately, minimizing uncertainty, and reducing team member interdependency.

Summary of Cultural Interventions While preliminary evidence seems to indicate that overall trainees feel that cross-cultural training interventions have utility and result in attitudinal changes (Bhawuk & Brislin, 2000; Black & Mendenhall, 1990; Deshpande & Viswesvaran, 1992), there are some concerns in terms of their direct applicability to promoting leadership within multi-cultural teams. First, most of the cross-cultural training programs are not specifically focused on training leadership skills for cultural diversity, but are indicative of culture awareness and interaction skills in general. When trying to extrapolate to the types of training interventions needed in terms of leading culturally diverse teams, these may be of limited value in that most speak nothing of how the leader can manage teamwork processes or interdependent action. Second, since many of the training programs reviewed focus on training people to live in another country, might not training be different if working with a multi-cultural team within one’s own country? Finally, one may argue that there needs to be a firmer grounding in theoretically developed training within this area (for examples see Brislin & Yoshida, 1994; Cushner & Brislin, 1997). Though not perfect, the wealth of cross-cultural training programs do serve as a place from which we can begin to extract knowledge of what works and does not work.

Approach No. 5: What Methodologies Can We Borrow and Adapt From Other Domains to be Used to Study Multi-Cultural Team Leadership?

In the challenges section we mentioned that the study of cultural diversity relies primarily on survey methodology. However, there are a number of other methods within other domains that can be utilized to provide a broader methodology base and eliminate some of the potential for same-source bias. Within this section, we will elaborate on a representative few borrowed from human factors (see Table 3 for more detail). These methods can be grouped into those that pertain to understanding the work environment and measurement.

Understanding the Work Environment In terms of understanding the work environment there are two methodologies employed within the human factors domain that can be used to provide better insight into the cognitive competencies needed for leaders within multi-cultural teams. They are knowledge elicitation and cognitive task analysis. Challenges and Approaches to Understanding Leadership Efficacy 369

Table 3. Examples of Knowledge Elicitation Techniques.

Knowledge Example Application Characteristics Elicitation Method

Unstructured Free form discussions with an expert. Entails little preparation but may interview require extensive domain knowledge on part of analyst to gain useful information. Provides little control over task content. Most applicable early in job analysis. Coding and analysis of data problematic. Structured Discussions with an expert in which the May be useful at any time during a interview topics covered and sequencing are job analysis. Entails more preparation largely predetermined. than unstructured interview, and may require extensive domain knowledge on part of the analyst to gain useful information. Control over the specific aspects of the job treated varies with the application. Familiar tasks Analyst seeks to obtain, via direct Provides a foundation for job observation or verbal protocols, a sample analyses such as task analysis. of experts performing typical tasks. Tough cases Analyst seeks to obtain, via direct Most applicable late in job analysis (Hoffman, observation or verbal protocols, when focused information is sought. 1987) examples of experts performing challenging job situations.

Knowledge Elicitation. Table 3 provides examples of “direct” knowledge elicitation approaches; that is, approaches that obtain knowledge by directly asking or observing the expert. Several general comments can be made about these techniques. First, all are, in one way or another, situated or embedded in a job context. In a structured interview, for example, an expert may be prompted to recall a job situation. In the method of familiar tasks (Hoffman, 1987), an expert is instructed to provide a commentary on his actions as he performs his job. The embedded aspect of knowledge elicitation is critical for revealing meaningful information about the job (Hoffman). As one example, Fowlkes, Salas, Baker, Cannon-Bowers and Stout (2000) used an event-based knowledge elicitation approach. Experts were provided with deliberate and controlled job situations, allowing investigation of specific task aspects and the comparison of expert responses. Specifically, a videotape was developed showing an instructor pilot and student conducting a training mission. Various job situations were depicted in the video to gather information pertinent 370 EDUARDO SALAS ET AL. to understanding team situational awareness. The videotape was shown to experts and novices and responses to the videotape were collected using a questionnaire at predetermined stop points. This allowed the identification of expert-novice differences in important cues and response strategies. Cognitive Task Analysis. Cognitive Task Analysis (CTA) methods are used to analyze the performance of teams conducting complex tasks. CTA methods go beyond traditional task analysis approaches, aiming to describe characteristics of tasks such as goals, strategies, and important environmental cues that are used by focusing on the cognitive competencies needed. For example Neville, Fowlkes, Milham, Bergondy and Glucroft (2001) use a CTA methodology to identify expert-novice differences in distributed team members in terms of the cues, knowledge and team skills used to achieve and maintain team integration. The differences identified suggest the types of training experiences and guidance that would help less experienced team members acquire the adaptive level of expertise required by the dynamic, complex, high-risk, and time-pressured environment in which team members work. In addition, the differences were used to identify which aspects of knowledge and performance should be measured to assess the acquisition of team coordination expertise.

Measurement In terms of measurement there are at least two areas that can be leveraged against: (a) event-based approaches and (b) cognitive-based assessment measures. Event-based approaches. Within the challenges section we identified several challenges to determining and developing “best practices” in terms of cross- cultural training. Event-based approaches can provide a leap in the right direction. Event-based techniques create measurement and training opportunities by system- atically identifying and introducing events within scenario-based exercises that provide known opportunities to observe behaviors of interest (i.e. those that reflect targeted competencies; Dwyer, Fowlkes, Oser, Salas & Lane, 1997; Johnston, Smith-Jentsch & Cannon-Bowers, 1997). They present a useful technique for maintaining links between training objectives, exercise design, and performance assessment. Moreover, instructors find them simple to use for providing highly specific feedback. Finally, they are associated with excellent psychometric prop- erties, a characteristic making them useful for training effectiveness evaluations. The event-based approach to measurement has been used successfully both within the military and elsewhere. Furthermore, it has spawned a number of specific measurement tools (e.g. Johnston et al.; Wigdor & Green, 1991). Cognitive-based assessment measures. Acquiring the knowledge structures necessary for task expertise is at the root of all training programs, yet such measures are just beginning to find their way into training and assessment. Over Challenges and Approaches to Understanding Leadership Efficacy 371 the past decade, there has been substantial methodological progress in assessing knowledge structures (e.g. Glaser, 1992; Nichols, Chipman & Brennan, 1995; Schvaneveldt, 1990). Potentially useful measures include card sorts and measures of structural knowledge (i.e. Pathfinder, scenario-based measurements, concept maps, see Cooke, 1999 for other knowledge elicitation devices).

Summary of Approaches

There are several literatures that can serve as starting points from which to extract information that will aid in understanding team leadership within culturally diverse teams, as well as several new methodologies that can be used to expand what is currently known. While the previously reviewed literature bases provide a foundation to begin hypothesis generation as to the challenges and actions through which leaders may accomplish these functions, there is a serious need to integrate the literature on team leadership with that on cross-cultural teams and cultural diversity. In an effort to further understand effective leadership of culturally diverse teams, we must understand the impact that the leader’s culture will have on his actions and interpretations, as well as what challenges cultural diversity poses in terms of the creation and maintenance of effective teamwork (i.e. interdependent interactions among members). In terms of the first question, the literature on cultural differences in values, cognitive style, and behavioral preferences serves as a framing point. Answering the second question poses a bit more difficulty because information must be extracted from various literature bases none of which fully address this question, albeit parts may be implicitly or explicitly addressed in each. However, existing literature provides some guidance in that most of the differences in cul- tural diversity have been argued to lie within the management of communication and meaning ascription. Therefore, key leadership functions within multi-cultural teams would seem to revolve around minimizing the impact of these two barriers. Extracting from the theoretical drivers underlying functional leadership theory, team leaders should be able to perform these activities through the functional behavior of sense-making (i.e. gathering information from the environment, meaning ascription) and sense-giving (i.e. communication of clear, enabling direction, and creation of enabling performance environment; see Burke et al., 2002), which in turn should result in the shared affect, behavior, and cognition that enables coordinated action. Furthermore, there are several methodologies gathered from domains such as human factors that can be utilized to build empirically sound research that will 372 EDUARDO SALAS ET AL. provide the beginning foundation for a nomological net concerning leadership within multi-cultural teams. Not only can these methodologies provide an additional avenue through which needed empirical research on team leadership within these settings can be conducted, but some can additionally be used in the modification and development of training programs for team leaders. For while there exist many varieties of cross-cultural training programs, many lack the theoretical background and scientific rigor that results in the most efficacious programs possible.

RESEARCH NEEDS AND FUTURE DIRECTIONS

We have previously argued that as culturally diverse teams become more common the need to understand the mechanism(s) by which leaders can promote team effectiveness within these teams becomes of paramount concern. As this is a relatively young area that has not yet been explored the research needs abound. Through examination of the challenges and approaches that have been presented within this chapter, we identify three areas that we feel offer the most promise for furthering our understanding of leadership within culturally diverse teams.

Research Need No. 1: Further Our Understanding of How to Develop Leaders for Multi-Cultural Teams

This line of research is actually twofold: (a) understanding the competencies needed for leaders operating within culturally diverse teams and (b) modifying existing training programs or developing new programs that will foster the identified competencies.

Competencies Needed There is a serious need to integrate the literature on team leadership with that on culture. Therefore, one direction that research might take early on is to focus on understanding how cultural diversity may impact key leadership functions such as: sense-making and sense-giving. For example, what does cultural diversity mean in terms of how the leader provides direction that is interpreted as clear and motivating to all members? Relevant cultural dimensions here may be high-low context, individualism-collectivism, and degree of uncertainty avoidance. More specifically, the team literature has indicated that challenging goals tend to motivate members (see Locke & Latham, 1994). However, in cultures that avoid uncertainty and personal risk, challenging goals may be seen Challenges and Approaches to Understanding Leadership Efficacy 373 as threatening as opposed to motivating. For a further discussion of this line of research see Burke et al. (2002). Furthermore, not only is it important to understand how the cultural diversity of team members may impact the creation and maintenance of team process, it is also important to understand how culture and other individual differences in the leader may impact the enactment of these functional behaviors.

Development and/or Modification of Training At this point in time, there is not a clear understanding of how to prepare teams to work within the complex environments in which cross-cultural teams often operate, not to mention the leadership of these teams. For example, Pierce (n.d.) described several training deficiencies based on observations of a unit preparing for deployment to Bosnia to work within culturally diverse operations. The deficiencies identified ranged from inappropriate training content to lack of diagnostic feedback. Furthermore, based on observations and interviews, Pierce and Klein (n.d.) identify several challenges for teams operating within these environments. Some of these include problems with maintaining situational awareness and shared mental models due to access to different information. These highlight key areas where leaders may need to focus their attention. Our review also indicates that while cultural diversity training programs abound, most are focused at the individual level and many are not theoretically based. Relying on the science of training (see Salas & Cannon-Bowers, 2001), we know that effective training involves providing information, demonstration, practice, and feedback. However, most of the training interventions applied to cross-cultural teams to date are focused on providing information to create an awareness of cross-cultural differences. As we have argued, this may not be enough to produce the requisite changes in cognitions or behaviors. The focus on “awareness” training is not a new state of affairs, as a similar state existed within aviation during the initial development of cockpit resource management (CRM) training. As the research base grew, supporting problem definition and development of theoretical underpinnings for effective crew member interaction, the training matured to include practice and feedback in job-relevant contexts. Now this is considered an essential aspect of CRM training. We anticipate a similar progression in the domain of leadership training for cross-cultural teams. One way to facilitate this progression is to take advantage of training methodologies and practices that have been shown to work in other complex environments. Scenario-Based Training. To increase the efficacy of training, the potential for scenario-based (i.e. event-based) training should be evaluated and requisite scenarios built. Cross-cultural training has been argued to not always live up to 374 EDUARDO SALAS ET AL. its potential. Given the caveat of soundly designed training, we see the use of simulation as being an area that should be explored when training individuals to lead culturally diverse teams. More specifically, scenario-based training can be embedded within simulation to produce more effective training. This offers several potential advantages over current training, in that it: (a) can be offered to members who are distributed, as often in multi-national military settings members may not be able to physically come together until the last minute; (b) can be designed as embedded training for use during downtime, facilitating learning opportunities; and (c) can create an immersive environment for trainees. As cross-cultural teams are often assembled to accomplish specific tasks and are largely ad hoc, the above characteristics are all important. Stress Exposure Training. In addition to scenario-based training research, stress exposure training may be another training technique that can be adapted for use in training leaders within cross-cultural teams. Membership in culturally diverse teams is often seen as challenging, frustrating, and is probably often perceived as being an additional stressor. By extension the leadership of such teams is argued to have many of these same characteristics. One potential training technique that has proven to be effective in training individuals for stressful environments is termed stress exposure training (Driskell, Johnston & Salas, 2001). This technique involves trainees: (a) receiving preparatory information regarding the stressful environment and stressors within that environment; (b) learning the impact on behaviors and techniques to handle stressors; and (c) practicing acquired techniques in a graduated manner. We argue that elevating this technique to the team level and combining it with the use of event-based simulation would result in a high payoff in terms of training. As opposed to traditional stressors (i.e. workload, time, ambiguity), the stressor would be seen as the culturally diverse team and the challenges posed for team leadership within this domain. For more information on the components comprising stress exposure training see Driskell and Johnston (1998). Feedback Approaches. A vital component of any successful training program is feedback. Methods of providing feedback to cross-cultural teams are needed. Feedback is one way that leaders can align the interpretation of events among team members. Watson et al. (1993) observed performance improvements (solution quality in case study analyses) in culturally heterogeneous teams over time. While performance differences favoring homogeneous teams were evident at the outset, these differences diminished, until at the end of the study no differences could be detected. It was argued that part of the reason for improved performance over time was the process feedback (summaries of strengths and weaknesses) provided to the teams. Thomas (1999) argues that feedback can serve to integrate the viewpoints of the team members. Challenges and Approaches to Understanding Leadership Efficacy 375

In addition, leaders must be careful for cultural differences can affect the way in which feedback is perceived, as well as the propensity to provide feedback. For example, individualists are more likely to look for personal attributions (or take feedback as a personal issue) compared to collectivists who are more likely to focus on contextual factors. This may go back to the notion that schemas of teamwork possessed by different cultures will have performance implications. Similarly, the cultural dimension of high-low context may also impact how feedback is perceived, as members of high context cultures tend not to separate the person from the issue (Ting-Toomey, 1985). As such, leaders need to be careful, because even task-based feedback may be interpreted as a personal attack. See Burke et al. (2002) for other cultural implications related to feedback within teams.

Research Need No. 2: Systematically Investigate the Ways in Which Methods Can Be Developed or Adapted in Order to Facilitate the Study of This Phenomenon

Currently, cross-cultural research relies almost exclusively on the use of surveys. While they have many strengths (i.e. easy to administer, inexpensive to develop), when they become the only method used, research findings are subject to common method bias. Furthermore, as environments become increasingly complex other methodologies are needed to enable researchers to better understand how cognition is impacted. Fields such as human factors offer many other methods that might be adapted for use in studying leadership within culturally diverse teams. The following are a few of the methods that may be adapted for use within this domain:
Interviews with successful leaders of cross-cultural teams. To date in this domain, this is an underused technique.
Cognitive task analysis. To our knowledge, this has not been applied and may be especially useful for the identification of: (a) the cognitive competencies that leaders and teams require to operate within these environments and (b) the challenges posed for team leaders in the area of maintaining shared team cognition.
Variation in the use of measurement indices. Observation of teams in context (e.g. performing job tasks or simulated tasks) has proven to be a very useful measurement tool when based on theory and a priori defined dimensions (see Fowlkes, Lane, Salas, Franz & Oser, 1994). Further implications involve the need to use measures that provide diagnostic information concerning team and leader behaviors (i.e. need to capture behavior, cognition, and attitudes). We suspect that team behaviors related to information exchange and supporting behavior will be a key. For the most part, the measures used within current 376 EDUARDO SALAS ET AL.

cross-cultural training programs are primarily focused on outcomes and are not very diagnostic. While methods from other domains can be utilized, Gelfand and Arad (2002) have argued that data collection instruments vary in their instrumentality, partially depending on the culture one is examining. As such, there is a need to begin to investigate whether other data collection methodologies will transfer for use in a cross-cultural domain. If not, new methods need to be developed and evaluated. Good measurement provides the foundation for effective research and understanding.

Research Need No. 3: Development of a Parsimonious Theoretical Framework That Will Serve to Stretch Beyond the Work of Hofstede

The third key need that we have identified, based on the challenges and approaches outlined within the chapter, is the need to develop a parsimonious framework that can serve to guide the systematic study of leadership within cross-cultural teams. In terms of cultural dimensions, while the work of Hofstede (1980) is the most cited work, there are also other cultural typologies that have been less well investigated. Research should begin to examine the impact that the various cultural dimensions have on team process and leadership within culturally diverse teams. Up to this point, these typologies have only been examined in terms of individuals. Perhaps there are other cultural dimensions that are relevant in terms of teams. The same argument can be made in regard to the team and leadership literature. Work needs to be conducted that will serve as the foundation for an initial framework.

CONCLUDING COMMENTS

While global conditions are making it imperative that researchers and practitioners alike understand the factors contributing to the effective leadership of culturally diverse teams, there are many challenges that are to be faced in the search for understanding, as well as some approaches that can be extracted from existing work in other domains. Within the current chapter we have offered what we see as several challenges (i.e. theoretical guidance, problem definition, determination of methods to use, and interventions to develop). However, we have not stopped there for we have also briefly reviewed some areas that can be utilized to help address the challenges posed (i.e. literatures on cultural diversity, teams, leadership, training, Challenges and Approaches to Understanding Leadership Efficacy 377 and measurement). We hope that the content contained within this chapter will serve as “food for thought” and cause researchers and practitioners alike to recognize not only the impact that cultural diversity can have on team leadership, but the complex nature of this topic. Finally, we have offered our view of three primary areas of research needs and some possible suggestions as to how to implement them.

REFERENCES

Adler, N. J. (1997). International dimensions of organizational behavior (3rd ed.). Cincinnati, OH: International Thomson Publishing. Ayman, R., & Chemers, M. M. (1983). Relationship of supervisory behavior ratings to work group effectiveness and subordinate satisfaction among Iranian managers. Journal of Applied Psychology, 68(2), 338–341. Bass, B. (1990). Bass & Stogdill’s handbook of leadership (3rd ed.). New York: Free Press. Bass, B. M. (1997). Does the transactional-transformational leadership paradigm transcend organiza- tional and national boundaries? American Psychologist, 52(2), 130–139. Bennett, J. M. (1986). A developmental approach to training for intercultural sensitivity. International Journal of Intercultural Relations, 10, 179–196. Bhawuk, D. P. S., & Brislin, R. W. (2000). Cross-cultural training: A review. Applied Psychology: An International Review, 49(1), 162–191. Black, J. S., & Mendenhall, M. (1990). Cross-cultural training effectiveness: A review and a theoretical framework for future research. The Academy of Management Review, 15(1), 113–136. Brislin, R., & Yoshida, T. (1994). Improving intercultural interactions: Modules for cross-cultural training programs. Thousand Oaks, CA: Sage. Brislin, R. W., Cushner, K., Cherrie, C., & Yong, M. (1986). Intercultural interactions: A practical guide. Beverly Hills, CA: Sage. Brodbeck, F. C., Frese, Akerblom, Audia, Bakacsi, & Bendova (2000). Cultural variation of leadership prototypes across 22 European countries. Journal of Occupational and Organizational Psychology, 73, 1–29. Bruschke, J. C., Gartner, C., & Seiter, J. S. (1993). Student ethnocentrism, dogmatism, and motivation: A study of BAFA BAFA. Simulation and Gaming, 21(1), 9–20. Burke, C. S. (1999). Examination of the cognitive mechanisms through which team leaders promote effective team processes and adaptive team performance. Unpublished doctoral dissertation, George Mason University, Virginia. Burke, C. S., Salas, E., Stagl, K., & Fowlkes, J. (2002, April). Leading multi-national teams. In: J. C. Zeigert & K. Klein (Co-chairs), Team Leadership: Current Theoretical and Research Perspectives. Symposium conducted at the 17th Annual Meeting of the Society for Industrial and Organizational Psychology, Toronto, Canada. Campion, M. A., Medsker, G. J., & Higgs, A. C. (1993). Relationships between work group characteristics and effectiveness: Implications for designing effective work groups. Personnel Psychology, 46, 823–850. Cannon-Bowers, J. A., Tannenbaum, S. I., Salas, E., & Volpe, C. E. (1995). Defining team competen- cies and establishing team training requirements. In: R. Guzzo, E. Salas & Associates (Eds), 378 EDUARDO SALAS ET AL.

Team Effectiveness and Decision Making in Organizations (pp. 333–380). San Francisco, CA: Jossey-Bass. Cherrie, S. F. (1998). Multinational mine-strike recovery operation. Retrieved on April 25, 2002 from http://call.army.mil/products/trngqtr/tq2–98/cherrie.htm. Choi, I., & Nisbett, R. E. (2000). Cultural psychology of surprise: Holistic theories and recognition of contradiction. Journal of Personality and Social Psychology, 79(6), 890–905. Choi, I., Dalal, R., Kim-Prieto, C., & Park, H. (2003). Culture and judgment of causal relevance. Journal of Personality and Social Psychology, 84(1), 46–59. Conyne, R. K., Wilson, F. R., Tang, M., & Shi, K. (1999). Cultural similarities and differences and differences in group work: Pilot study of a U.S.-Chinese task group comparison. Group Dynamics: Theory, Research and Practice, 3(1), 40–50. Cooke, N. J. (1999). Knowledge elicitation. In: F. T. Durso, R. S. Nickerson, R. W. Schvaneveldt, S. T. Dumais, D. S. Lindsay & M. T. H. Chi (Eds), Handbook of Applied Cognition (pp. 479–509). Sussex, England: Wiley. Cox, T. H., Lobel, S. A., & McLeod, P. L. (1991). Effects of ethnic group cultural differences on cooperative and competitive behavior on a group task. Academy of Management Journal, 34(4), 827–847. Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52, 281–302. Cushner, K., & Brislin, R. W. (Eds) (1997). Improving intercultural interactions: Models for cross cultural training programs (Vol. 2). Thousand Oaks, CA: Sage. Daily, B., Whatley, A., Ash, S. R., & Steiner, R. L. (1996). The effects of a group decision support system on culturally diverse and culturally homogeneous group decision making. Information & Management, 30, 281–289. Den Hartog, D. N., House, R. J., Hanges, P. J., & Ruiz-Quintanilla, S. A. (1999). Culture specific and cross-culturally generalizable implicit leadership theories: Are attributes of charismatic/ transformational leadership universally endorsed? Leadership Quarterly, 10(2), 219–256. Deshpande, S. P., & Viswesvaran, C. (1992). Is cross-cultural training of expatriate managers effective: A meta analysis. International Journal of Intercultural Relations, 16(3), 295–310. Distefano, J. J., & Maznevski, M. L. (2000). Creating value with diverse teams in global management. Organizational Dynamics, 29(1), 45–63. Dodd, C. H. (1991). Dynamics of intercultural communication (3rd ed.). Dubuque, IA: William C. Brown. Dorfman, P. W., & Howell, J. P. (1988). Managerial leadership in the United States and Mexico. In: C. S. Granrose & S. Oskamp (Eds), Cross-Cultural Work Groups (pp. 234–264). Thousand Oaks, CA: Sage. Driskell, J. E., Johnston, J. H., & Salas, E. (2001). Does stress training generalize to novel settings. Human Factors, 43(1), 99–110. Driskell, J. E., & Johnston, J. H. (1998). Stress exposure training. In: J. A. Cannon-Bowers & E. Salas (Eds), Making Decisions Under Stress: Implications for Individual and Team Training (pp. 191–217). Washington, DC: American Psychological Association. Dwyer, D. J., Fowlkes, J. E., Oser, R. L., Salas, E., & Lane, N. E. (1997). Team performance measurement in distributed environments: The TARGET’s methodology. In: M. T. Brannick, E. Salas & C. Prince (Eds), Assessment and Management of Team Performance: Theory, Research, and Applications (pp. 137–154). Mahwah, NJ: Lawrence Erlbaum. Earley, P. C., & Gibson, C. B. (2002). Multinational work teams: A new perspective. Mahwah, NJ: Lawrence Erlbaum. Challenges and Approaches to Understanding Leadership Efficacy 379

Earley, P. C., & Mosakowski, E. (2000). Creating hybrid team cultures: An empirical test of transnational team functioning. Academy of Management, 43(1), 26–49. Eby, L. T., & Dobbins, G. H. (1997). Collectivistic orientation in teams: An individual and group-level analysis. Journal of Organizational Behavior, 18(3), 275–295. Elmuti, D. (2001). Preliminary analysis of the relationship between cultural diversity and technology in corporate America. Equal Opportunities International, 20(8), 1–16. Elron, E. (1997). Top management teams within multinational corporations: Effects of cultural heterogeneity. Leadership Quarterly, 8(4), 393–412. Erez, M., & Earley, P. C. (1993). Culture, self-identity, and work. New York: Oxford University Press. Fleishman, E. A., Mumford, M. D., Zaccaro, S. J., Kevin, K. Y., Korotkin, A. L., & Hein, M. B. (1991). Taxonomic efforts in the description of leader behavior: A synthesis and functional interpretation. Leadership Quarterly, 2(4), 245–287. Fleishman, E. A., & Zaccaro, S. J. (1992). Toward a taxonomic classification of team perfor- mance functions: Initial considerations, subsequent evaluations and current formulations. In: R. W. Sweezy & E. Salas (Eds), Teams: Their Training and Performance (pp. 31–56). New Jersey: ABLEX. Fowlkes, J. E., Lane, N. E., Salas, E., Franz, T., & Oser, R. (1994). Improving the measurement of team performance: The targets methodology. Military Psychology, 6(1), 47–61. Fowlkes, J. E., Salas, E., Baker, D. P., Cannon-Bowers, J. A., & Stout, R. J. (2000). The utility of event-based knowledge elicitation. Human Factors, 42(1), 24–35. Gaertner, S. L., Mann, J., Murrell, A., & Dovidio, J. (1989). Reducing intergroup bias: The benefits of recategorization. Journal of Personality & Social Psychology, 57(2), 239–249. Gelfand, M. J., & Arad, S. (2002). Theoretical and methodological issues in cross-cultural I-O psychology. Tutorial conducted at the 17th Annual Conference for the Society for Industrial and Organizational Psychology, Toronto, Canada. Gelfand, M. J., Raver, J. L., & Ehrhart, K. H. (2002). Methodological issues in cross-cultural organi- zational research. In: S. G. Rogelberg (Ed.), Handbook of Research Methods in Industrial and Organizational Psychology (pp. 216–246). Malden, MA: Blackwell Publisher Ltd. Gibson, C. B. (1999). Do they do what they believe they can? Group efficacy and group effectiveness across tasks and cultures. Academy of Management Journal, 42(2), 138–152. Gibson, C. G., & Zellmer-Bruhn, M. (2001, June). Metaphors and meaning: An intercultural analysis of the concept of teamwork. Administrative Science Quarterly, 1–26. Gladstein, D. L. (1984). Groups in context: A model of task group effectiveness. Administrative Science Quarterly, 29, 499–517. Glaser, R. (1992). Learning, cognition, and education: Then and now. In: H. L. Pick, P. Van Den Broek, D.C. Knill & H. L. Pick, Jr. (Eds), Cognition: Conceptual and Methodological Issues (pp. 239–265). Washington, DC: American Psychological Association. Gomez, C., Kirkman, B. L., & Shapiro, D. L. (2000). The impact of collectivism and in-group/ out-group membership on the evaluation generosity of team members. Academy of Management Journal, 43(6), 1097–1106. Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of leader-member exchange (LMX) theory of leadership over 25 years: Applying a multi-level multi-domain perspective. Leadership Quarterly, 6(2), 219–247. Graen, G. B., & Wakabayashi, M. (1994). Cross-cultural leadership making: Bridging American and Japanese diversity for team advantage. In: H. C. Triandis, M. D. Dunnette & L. M. Hough (Eds), Handbook of Industrial of Organizational Psychology (2nd ed., Vol. 4, pp. 415–446). Palo Alto, CA: Consulting Psychologists Press. 380 EDUARDO SALAS ET AL.

Grahn, J. L., & Swenson, D. X. (2000). Cross-cultural perspectives for quality training. Cross-cultural Management: An International Journal, 7(3), 19–24. Gudykunst, W. B., & Hammer, M. R. (1983). Basic training design: Approaches to intercultural training. In: D. Landis & R. W. Brislin (Eds), Handbook of Intercultural Training: Issues in Theory and Design (Vol. 1, pp. 118–154). Elmsford, NY: Pergamon. Gudykunst, W. B., Matsumoto, Y., Ting-Toomey, S., Nishida, T., Kim, K., & Heyman, S. (1996). The influence of cultural individualism-collectivism, self-construals, and individual values on communication styles across cultures. Human Communication Research, 22, 510–543. Hackman, J. R. (1987). The design of work teams. In: J. W. Lorsch (Ed.), Handbook of Organizational Behavior (pp. 315–342). Englewood Cliffs, NJ: Prentice-Hall. Hackman, J. R., & Walton, R. E. (1986). Leading groups in organizations. In: P. S. Goodman and Associates (Eds), Designing Effective Work Groups (pp. 72–119). San Francisco, CA: Jossey-Bass. Hall, E. T., & Hall, M. R. (1990). Understanding cultural differences. Garden City, NY: Intercultural Press. Harper, L. F., & Rifkind, L. J. (1995). Intercultural communication in the diversified quality workplace. In: L. F. Harper & L. R. Rifkind (Eds), Cultural Collision: Quality Teamwork in the Diverse Workplace (pp. 41–61). Dubuque, IA: Kendall/Hunt Publishing Company. Harrison, G. L., McKinnon, J. L., Wu, A., & Chow, C. W. (2000). Cultural influences on adaptation to fluid workgroups and teams. Journal of International Business Studies, 31(3), 489–505. Harrison, D. A., Price, K. H., & Bell, M. P. (1998). Beyond relational demography: Time and the effects of surface and deep level diversity on work group cohesion. Academy of Management Journal, 41(1), 96–107. Hazucha, J. F., Hezlett, S. A., Bontems-Wackens, S., & Ronnqvist, R. (1999). In search of the Euro- manager: Management competencies in France, Germany, Italy, and the United States. In: W. H. Mobley (Ed.), Advances in Global Leadership (Vol. 1, pp. 267–290). Stamford, CT: JAI Press. Helmreich, R. L. (1994). Anatomy of a system accident: The crash of Avianca Flight 052. International Journal of Aviation Psychology, 4(3), 265–284. Helmreich, R. (2000). On error management: Lessons from aviation. BMJ, 320, 781–785. Hodgetts, R. M., & Luthans, F. (2000). International management: Culture, strategy, and behavior. New York: McGraw-Hill. Hoffman, P. J. (1987). The problem of extracting the knowledge of experts from the perspective of experimental psychology. The AI Magazine, 8, 53–66. Hofstede, G. (1980). Culture’s consequences: International differences in work related values. Beverly Hills, CA: Sage. Hofstede, G. (1993). Cultural constraints in management theories. Academy of Management Executive, 7(1), 81–94. Horenczyk, G., & Berkerman, Z. (1997). The effects of intercultural acquaintance and structured intergroup interaction on ingroup, outgroup, and reflected ingroup stereotypes. International Journal of Intercultural Relations, 21(1), 71–83. House, R. J., et al. (1999). Cultural influences on leadership and organizations: Project globe. In: W. H. Mobley (Ed.), Advances in Global Leadership (Vol. 1., pp. 171–233). Stamford, CT: JAI Press. House, R. J., & Aditya, R. N. (1997). The social scientific study of leadership: Quo vadis? Journal of Management, 23(3), 409–473. Hutchins, E., Holder, B. E., & Perez,´ R. A. (2002). Culture and Flight Deck Operations. Unpublished internal report prepared for the Boeing Company [Sponsored Research Agreement 22-5003]. San Diego: University of California San Diego. Challenges and Approaches to Understanding Leadership Efficacy 381

Ilgen, D. R., LePine, J. A., & Hollenbeck, J. R. (1997). Effective decision making in multinational teams. In: P. C. Earley & M. Erez (Eds), New Perspectives on International Industrial/ Organizational Psychology (pp. 377–409). San Francisco, CA: New Lexington Press/ Jossey-Bass. Jackson, S. E. (1992). Team composition in organizational settings: Issues in managing a diverse workforce. In: S. Worchel, W. Wood & J. Simpson (Eds), Group Process and Productivity (pp. 138–173). Thousand Oaks, CA: Sage. Jackson, S. E., May, K. E., & Whitney, K. (1995). Understanding the dynamics of diversity in decision-making teams. In: R. A. Guzzo, E. Salas & Associates (Eds), Team Effectiveness and Decision Making in Organizations (pp. 204–261). San Francisco: Jossey-Bass. Janis, I. L. (1972). Victims of groupthink: A psychological study of foreign-policy decisions and fiascoes. Boston, MA: Houghton Mifflin. Ji, L., Peng, K., & Nisbett, R. E. (2000). Culture, control, and perception of relationships in the environment. Journal of Personality and Social Psychology, 78(5), 943–955. Johnston, J. H., Smith-Jentsch, K. A., & Cannon-Bowers, J. A. (1997). Performance measurement tools for enhancing team decision-making training. In: M. T. Brannick & E. Salas (Eds), Team Performance Assessment and Measurement: Theory, Methods, and Applications (pp. 311–327). Mahwah, NJ: Lawrence Erlbaum. Jung, D. I., & Yammarino, F. J. (2001). Perceptions of transformational leadership among Asian Americans and Caucasian Americans: A level of analysis perspective [Publication No. 102]. Ciber Working Paper Series, San Diego, CA: San Diego State University. Katz, D., & Kahn, R. L. (1978). The social psychology of organizations. New York: Wiley. Kirchmeyer, C. (1993). Multicultural task groups. Small Group Research, 24(1), 127–148. Kirkman, B. L., Lowe, K. B., & Gibson, C. B. (under review). Two decades of culture’s consequences: A review of the empirical research on Hofstede’s cultural values framework. Klein, H. A., Pongonis, A., & Klein, G. (2000, June). Cultural barriers to multinational C2 decision making. Presented at the 2000 Command and Control Research and Technology Symposium, Naval Postgraduate School, Monterey, CA. Kluckhohn, C., & Strodbeck, F. (1961). Variations in value orientations. Westport, CT: Greenwood Press. Kozlowski, S. W. J., Gully, S. M., & McHugh, P. P. (1993). Leadership and team effectiveness: A developmental-task contingent model. In: J. A. Cannon-Bowers (Chair), Optimizing Team Performance Through Team Leader Behavior. Symposium conducted at the 8th annual conference of the Society for Industrial and Organizational Psychology, San Francisco, CA. Kozlowski, S. W. J., Gully, S. M., Salas, E., & Cannon-Bowers, J. A. (1996). Team leadership and development: Theory, principles, and guidelines for training leaders and teams. In: M. Beyerlein, S. Beyerlein & D. Johnson (Eds), Advances in Interdisciplinary Studies of Work Teams: Team Leadership (Vol. 3, pp. 253–292). Greenwich, CT: JAI Press. Kuang, J. C. Y., & Davis, D. D. (2001). Culture and team performance in foreign flightcrews. Paper presented at the 109th Annual Convention of the American Psychological Association, San Francisco, CA. Lee, F., Hallahan, M., & Herzog, T. (1996). Explaining real-life events: How culture and domain shape attributions. Personality and Social Psychology Bulletin, 22(7), 732–741. Lewin, K. (1951). Field theory in social science; selected theoretical papers. D. Cartwright (Ed.). New York: Harper & Row. Locke, E. A., & Latham, G. P. (1994). Goal setting theory. In: H. F. O’Neil Jr. & M. Drillings (Eds), Motivation: Theory and Research (pp. 13–29). Hillsdale, NJ: Lawrence Erlbaum. 382 EDUARDO SALAS ET AL.

Lord, R. G. (1977). Functional leadership behavior: Measurement and relation to social power and leadership perceptions. Administrative Science Quarterly, 22, 114–133. Lord, R. G., & Foti, R. J. (1985). Schema theories, information processing, and organizational behavior. In: H. P. Sims & D. A. Gioia (Eds), The Thinking Organization (pp. 21–48). San Francisco: Jossey-Bass. Malpass, R. S., & Salancik, G. R. (1977). Linear and branching formats in culture assimilator training. International Journal of Intercultural Relations, 1(2), 76–87. Marks, M. A., Mathieu, J. E., & Zaccaro, S. J. (2001). A temporally based framework and taxonomy of team process. Academy of Management Review, 26(3), 356–376. Marks, M. A., Zaccaro, S. J., & Mathieu, J. E. (2000). Performance implications of leader briefings and team interaction training for team adaptation to novel environments. Journal of Applied Psychology, 85(6), 971–986. Morgan, B. B., Jr., Glickman, A. S., Woodard, E. A., Blaiwes, A. S., & Salas, E. (1986). Measurement of team behaviors in a Navy environment (Technical Report Number 86–014). Orlando, FL: Naval Training Systems Center. Morris, M. W., & Peng, K. (1994). Culture and cause: American and Chinese attributions for social and physical events. Journal of Personality and Social Psychology, 67(6), 949–971. Mumford, M. D. (1986). Leadership in the organizational context: A conceptual approach and its applications. Journal of Applied Social Psychology, 16, 508–531. Neville, K., Fowlkes, J., Milham, L., Bergondy, M., & Glucroft, B. (2001). Team coordination expertise in complex distributed teams: A preliminary cognitive task analysis of the Navy carrier air wing strike team. Proceedings of the 11th International Symposium of Aviation Psychology [CDROM], Columbus, OH. Nichols, P. D., Chipman, S. F., & Brennan, R. L. (Eds) (1995). Cognitively diagnostic assessment. Mahwah, NJ: Lawrence Erlbaum. O’Brien, G. E., Fiedler, F. E., & Hewett, T. (1970). The effects of programmed culture training upon the performance of volunteer medical teams in Central America. Human Relations, 24(3), 209–231. O’Brien, G. E., & Plooji, D. (1977). Comparison of programmed and prose culture training upon attitudes and knowledge. Journal of Applied Psychology, 62, 499–505. Oetzel, J. G. (1998). Culturally homogeneous and heterogeneous groups: Explaining cultural communication processes through individualism-collectivism and self-construal. International Journal of Intercultural Relations, 22(2), 135–161. Orasanu, J., & Salas, E. (1993). Team decision making in complex environment. In: G. Klein, J. Orasanu, R. Calderwood & C. E. Zsambok (Eds), Decision Making in Action: Models and Methods (pp. 327–345). Norwood, NJ: Ablex. Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and collectivism: Evaluation of theoretical assumptions and meta-analyses. Psychological Bulletin, 128(1), 3–72. Parsons, T., & Shils, E. (1951). Toward a general theory of social action. Cambridge, MA: Harvard University Press. Peng, K., & Nisbett, R. E. (1999). Culture, dialectics, and reasoning about contradiction. American Psychologist, 54(9), 741–754. Pierce, L. G. (n.d.). Barriers to adaptability in a multinational team. Unpublished manuscript. Pierce, L. G., & Klein, G. (n.d.). Preparing and supporting adaptable leaders and teams for support and stability operations. Unpublished manuscript. Pillai, R., & Meindl, J. R. (1998). Context and charisma: A “meso” level examination of the relationship of organic structure, collectivism, and crisis to charismatic leadership. Journal of Management, 24(5), 643–671. Challenges and Approaches to Understanding Leadership Efficacy 383

Ravlin, E. C., & Meglino, B. M. (1993). The influence of value congruence on performance of routine and non-routine tasks. Paper presented at the annual meeting of the Academy of Management, Atlanta, GA. Robie, C., Johnson, K. M., Nilsen, D., & Hazucha, J. F. (2001). The right stuff: Understanding cultural differences in leadership performance. Journal of Management Development, 20(7), 639–649. Salas, E., Bowers, C. A., & Rhodenizer, L. (1998). It is not how much you have but how you use it: Toward a rational use of simulation to support aviation training. International Journal of Aviation Psychology, 8, 197–208. Salas, E., Burke, C. S., & Stagl, K. C. (in press). Developing teams and team leaders: Strategies and principles. To appear in: D. Day, S. J. Zaccaro & S. M. Halpin (Eds), Leader Development for Transforming Organizations. Mahwah, NJ: Lawrence Erlbaum. Salas, E., Dickinson, T. L., Converse, S. A., & Tannenbaum, S. I. (1992). Toward an understanding of team performance and training. In: R. J. Swezey & E. Salas (Eds), Teams: Their Training and Performance (pp. 3–29). Norwood, NJ: Ablex. Salas, E., & Cannon-Bowers, J. A. (2001). The science of training: A decade of progress. Annual Review of Psychology, 52, 471–499. Sarros, J. C., & Santora, J. C. (2001). Leaders and values: A cross-cultural study. Leadership & Organizational Development Journal, 22(5), 243–248. Scandura, T. (1995). Leader member exchange model of leadership fairness issues. Unpublished manuscript. Schermerhorn, J. R., & Bond, M. H. (1997). Cross-cultural leadership dynamics in collectivism and high power distance settings. Leadership & Organizational Development Journal, 18(4), 187–193. Schvaneveldt, R. (Ed.) (1990). Pathfinder associative networks: Studies in knowledge organization. Stamford, CT: Ablex Publishing. Smith-Jentsch, K. A., Johnston, J. H., & Payne, S. C. (1998). Measuring team-related expertise in complex environments. In: J. A. Cannon-Bowers & E. Salas (Eds), Making Decisions under Stress: Implications for Individual and Team Training (pp. 61–87). Washington, DC: American Psychological Association. Stasser, G., & Titus, W. (1987). Effects of information load and percentage of shared information on the dissemination of unshared information during group discussion. Journal of Personality and Social Psychology, 53, 81–93. Steiner, I. D. (1972). Group process and productivity. New York: Academic Press. Stewart, G. L., & Manz, C. C. (1995). Leadership for self-managing work teams: A typology and integrative model. Human Relations, 48(7), 747–770. Stoeberl, P. A., Kwon, I. G., Han, D., & Bae, M. (1998). Leadership and power relationships based on culture and gender. Women in Management Review, 13(6), 208–216. Tannenbaum, S. I., Beard, R. L., & Salas, E. (1992). Team building and its influence on team effectiveness: An examination of conceptual and empirical developments. In: K. Kelley (Ed.), Issue, Theory, and Research in Industrial/Organizational Psychology (pp. 117–153). Amsterdam: Elsevier. Tesluk, P. E., & Gerstner, C. R. (2002). Leading self-managing work groups: Effects on processes and performance. Paper presented at the 17th Annual Conference for the Society for Industrial and Organizational Psychology, Toronto, Canada. Tetlock, P. E. (1998). Close-call counterfactuals and belief-system defenses: I was not almost wrong but I was almost right. Journal of Personality & Social Psychology, 75(3), 639–652. Thomas, D. C. (1999). Cultural diversity and work group effectiveness. Journal of Cross-cultural Psychology, 30(2), 242–263. 384 EDUARDO SALAS ET AL.

Ting-Toomey, S. (1985). Toward a theory of conflict and culture. In: W. B. Gudykunst, L. P., Stewart & S. Ting-Toomey (Eds), Communication, Culture, and Organizational Processes. Newbury Park, CA: Sage. Triandis, H. C. (1983). Some dimensions of intercultural variation and their implications for community psychology. Journal of Community Psychology, 11(4), 285–302. Triandis, H. C. (1989). The self and behavior in different cultural contexts. Psychological Review, 96, 506–520. Triandis, H. C. (1993). The contingency model in cross-cultural perspective. In: M. M. Chemers & R. Ayman (Eds), Leadership Theory and Research: Perspectives and Directions (pp. 167–188). San Diego, CA: Academic Press. Triandis, H. C. (1994). Culture and social behavior. New York: McGraw-Hill. Triandis, H. C. (1995). Individualism and collectivism. Boulder, CO: Westview Press. Triandis, H. C. (2000). Culture and conflict. International Journal of Psychology, 35(2), 145–152. Trifonovitch, G. (1977). On cross-cultural orientation techniques. In: R. W. Brislin (Ed.), Culture Learning: Concepts, Applications, and Research (pp. 213–222). Honolulu, HI: University of Hawaii Press. Wade-Benzoni, K. A., Okumura, T., Brett, J. M., Moore, D. A., Tenbrunsel, A. E., & Bazerman, M. H. (2002). Cognitions and behavior in asymmetric social dilemmas: A comparison of two cultures. Journal of Applied Psychology, 87(1), 87–95. Watson, W. E., Kumar, K., & Michaelsen, L. K. (1993). Cultural diversity’s impact on interaction process and performance: Comparing homogeneous and diverse task groups. Academy of Management Journal, 36(3), 590–602. Wentling, R. M., & Palma-Rivas, N. (1999). Components of effective diversity training programmes. International Journal of Training and Development, 3(3), 215–226. Wigdor, A. K., & Green, B. F. (Eds) (1991). Performance assessment for the workplace: Technical issues (Vols 1 and 2). Washington, DC: National Academy Press. Worchel, S., & Mitchell, T. R. (1972). An evaluation of the effectiveness of the culture assimilator in Thailand and Greece. Journal of Applied Psychology, 56, 472–479. Yukl, G. A. (1989). Leadership in organizations (2nd ed.). Englewood Cliffs, NJ: Prentice-Hall. Zaccaro, S. J., Gilbert, J. A., Thor, K. K., & Mumford, M. M. (1991). Leadership and social intelligence: Linking social perceptiveness and behavioral flexibility to leader effectiveness. Leadership Quarterly, 2, 317–342. Zaccaro, S. J., Rittman, A., & Marks, M. A. (2001). Team leadership. Leadership Quarterly, 12, 451–483.