Taylor & Francis Group 270 Madison Avenue · PDF fileTaylor & Francis Group 270 Madison Avenue...

RoutledgeTaylor & Francis Group270 Madison AvenueNew York, NY 10016

RoutledgeTaylor & Francis Group27 Church RoadHove, East Sussex BN3 2FA

© 2010 by Taylor and Francis Group, LLCRoutledge is an imprint of Taylor & Francis Group, an Informa business

Printed in the United States of America on acid-free paper10 9 8 7 6 5 4 3 2 1

International Standard Book Number: 978-0-8058-6437-3 (Hardback)

For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data

Handbook of employee selection / editors, James L. Farr, Nancy T. Tippins.p. cm.

Includes bibliographical references and index.ISBN 978-0-8058-6437-3 (hardcover : alk. paper)1. Employee selection. 2. Employee selection--Handbooks, manuals, etc. I. Farr, James L. II.

Tippins, Nancy Thomas, 1950-

HF5549.5.S38H36 2010658.3’112--dc22 2009052677

Visit the Taylor & Francis Web site athttp://www.taylorandfrancis.com

and the Psychology Press Web site athttp://www.psypress.com

TAF-Y100248_FARR-09-0801-C000.indd ivTAF-Y100248_FARR-09-0801-C000.indd iv 1/20/10 4:03:04 PM1/20/10 4:03:04 PM

http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373

http://www.copyright.com

http://www.copyright.com/

http://www.copyright.com/

http://www.taylorandfrancis.com

http://www.psypress.com


vii

ContentsPreface.............................................................................................................................................xix

Editors .............................................................................................................................................xxi

Contributors ................................................................................................................................. xxiii

Chapter 1 Handbook of Employee Selection: An Introduction and Overview .............................1

James L. Farr and Nancy T. Tippins

Brief History of Science- and Data-Based Employee Selection ..................................1

Structure of Handbook of Employee Selection ............................................................3References ....................................................................................................................6

1 Foundations of Psychological Measurement and PARTEvaluation Applied to Employee SelectionJohn P. Campbell and Frank J. Landy, Coeditors

2 Chapter Reliability and Validity ................................................................................................9

Dan J. Putka and Paul R. Sackett

Overview .................................................................................................................... 10

Reliability ................................................................................................................... 11

Role of Measurement Models ..................................................................................... 14

Estimation of Reliability ............................................................................................22

Emerging Perspectives on Measurement Error ..........................................................36

Closing Thoughts on Reliability ................................................................................ 37

Concept of Validity .................................................................................................... 38

References .................................................................................................................. 45

Chapter 3 Validation Strategies for Primary Studies .................................................................. 51

Neal W. Schmitt, John D. Arnold, and Levi Nieminen

Nature of Validity ....................................................................................................... 51

Validation in Different Contexts ................................................................................ 53

Special Considerations in Criterion-Related Studies ................................................. 59

Concerns About the Quality of the Data: Cleaning the Data ....................................60

Modes of Decision-Making and the Impact on Utility and Adverse Impact ............. 61

Scientifi c or Long-Term Perspective: Limitations of Existing Primary

Validation Studies, Including the Current Meta-Analytic Database ......................64

Conclusions ................................................................................................................ 67

References .................................................................................................................. 67

TAF-Y100248_FARR-09-0801-C000toc.indd viiTAF-Y100248_FARR-09-0801-C000toc.indd vii 1/22/10 12:18:43 PM1/22/10 12:18:43 PM


simsr

Cross-Out

simsr

Replacement Text

Section Editors (place in itals)

watsonp

Sticky Note

Accepted set by watsonp

watsonp

Sticky Note


watsonp

Sticky Note

Completed set by watsonp

watsonp

Sticky Note


watsonp

Sticky Note

MigrationConfirmed set by watsonp

watsonp

Sticky Note



viii Contents

4 Chapter Work Analysis ............................................................................................................ 73

Kenneth Pearlman and Juan I. Sanchez

Traditional Selection-Related Applications of Work Analysis .................................. 73

A Review of Major Work Analysis Methods and Approaches .................................. 75

Key Work Analysis Practice Issues ............................................................................ 83

Frontiers of Work Analysis: Emerging Trends and Future Challenges .....................88

Synopsis and Conclusions ..........................................................................................94

References ..................................................................................................................94

5 Chapter Current Concepts of Validity, Validation, and Generalizability ................................99

Jerard F. Kehoe and Kevin R. Murphy

Converging Trends in Psychometrics—Toward a Unifi ed Theory of Validity ........ 100

Reliability, Validity, and Generalizations From Test Scores.................................... 101

Validation Process: Linking Tests With Their Uses ................................................ 102

Generalizing Research-Based Inferences to Implementation Decisions ................. 108

References ................................................................................................................ 120

2 Implementation and Management of Employee PARTSelection Systems in Work OrganizationsJerard F. Kehoe and Robert E. Ployhart, Coeditors

6 Chapter Attracting Job Candidates to Organizations ............................................................ 127

Ann Marie Ryan and Tanya Delany

Reaching Potential Applicants ................................................................................. 127

Sourcing Globally..................................................................................................... 130

Technology as a Means of Improving Reach ........................................................... 133

Considering Strategic Talent Management ............................................................. 134

Maintaining Interest ................................................................................................. 136

Maintaining Interest Around the Globe ................................................................... 139

Technology’s Role in Maintaining Interest .............................................................. 140

Maintaining the Interest of Targeted Talent ............................................................. 141

Accepting Offers ...................................................................................................... 142

Conclusions .............................................................................................................. 145

References ................................................................................................................ 146

7 Chapter Test Administration and the Use of Test Scores ...................................................... 151

Jeff W. Johnson and Frederick L. Oswald

Use of Test Scores .................................................................................................... 151

Decisions to Make Before Collecting Test Scores ................................................... 151

Collection of Test Scores .......................................................................................... 153

Computation of Test Scores ...................................................................................... 159

TAF-Y100248_FARR-09-0801-C000toc.indd viiiTAF-Y100248_FARR-09-0801-C000toc.indd viii 1/22/10 12:18:43 PM1/22/10 12:18:43 PM


simsr

Cross-Out

simsr

Replacement Text


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note



Contents ix

Making Selection Decisions ..................................................................................... 163

Conclusions .............................................................................................................. 166

References ................................................................................................................ 166

8 Chapter Technology and Employee Selection ........................................................................ 171

Douglas H. Reynolds and David N. Dickter

Introduction: The Changing Relationship Between Industrial-Organizational

Psychology and Information Technology ............................................................. 171

Critical Issues in Technology-Based Selection ........................................................ 176

Implementation Issues .............................................................................................. 186

Future Directions ...................................................................................................... 189

References ................................................................................................................ 192

9 Chapter Strategy, Selection, and Sustained Competitive Advantage ..................................... 195

Robert E. Ployhart and Jeff A. Weekley

Why Personnel Selection Must Show Business-Unit-Level Value ........................... 196

SHRM ...................................................................................................................... 197

Alignment of Selection and Strategy .......................................................................200

Selection’s Potential Contribution to Business Unit Value .......................................203

Selection’s Symbiotic Relationship With Other HR Activities ................................208

Conclusions ..............................................................................................................209

Acknowledgments ....................................................................................................209

References ................................................................................................................ 210

Chapter 10 Managing Sustainable Selection Programs .............................................................. 213

Jerard F. Kehoe, Stefan T. Mol, and Neil R. Anderson

Organization Context for Selection .......................................................................... 213

Defi ning Selection System Sustainability ................................................................ 214

HR Technology.........................................................................................................225

Conclusions .............................................................................................................. 233

References ................................................................................................................ 233

11 Chapter The Business Value of Employee Selection ............................................................. 235

Wayne F. Cascio and Lawrence Fogli

Traditional Model of Employee Selection ................................................................ 235

Challenges to the Business Value of the Traditional Approach ............................... 236

Dynamic, Contemporary Approach to Selection as an Organizational

Process ........................................................................................................237

Importance of Social Context and Interpersonal Processes in Selection

Decisions ............................................................................................................ 240

Conclusions: How Should We Value the Success of Selection in

Organizations? ................................................................................................248

References ................................................................................................................250

TAF-Y100248_FARR-09-0801-C000toc.indd ixTAF-Y100248_FARR-09-0801-C000toc.indd ix 1/22/10 12:18:43 PM1/22/10 12:18:43 PM



x Contents

3 Categories of Individual Difference PARTConstructs for Employee SelectionDavid Chan and Leaetta Hough, Coeditors

12 Chapter Cognitive Abilities.................................................................................................... 255

Deniz S. Ones, Stephan Dilchert, Chockalingam Viswesvaran, and Jesús F. Salgado

History, Current Usage, and Acceptability of Cognitive Ability Measures

in Employee Selection .......................................................................................... 255

Defi nitions and Theoretical Underpinnings ............................................................. 258

Structure ................................................................................................................... 258

Measurement ............................................................................................................ 259

Criterion-Related Validity Evidence ........................................................................ 261

Group Differences on Cognitive Ability Measures..................................................264

Future Challenges for Research and Practice ..........................................................269

Epilogue.................................................................................................................... 270

References ................................................................................................................ 270

13 Chapter Physical Performance Tests ...................................................................................... 277

Deborah L. Gebhardt and Todd A. Baker

Job Analysis for Arduous Jobs ................................................................................. 277

Physical Performance Tests ...................................................................................... 281

Validity of Physical Performance Tests ....................................................................285

Test Scoring and Administration .............................................................................287

Legal Issues .............................................................................................................. 291

Benefi ts of Physical Testing......................................................................................294

References ................................................................................................................294

14 Chapter Personality: Its Measurement and Validity for Employee Selection ........................299

Leaetta Hough and Stephan Dilchert

Structure of Personality Variables ...........................................................................299

Measurement Methods ............................................................................................. 301

Validity of Personality Constructs and Other Factors That Affect

Their Usefulness ...................................................................................................308

Conclusions .............................................................................................................. 312

References ................................................................................................................ 312

15 Chapter Values, Styles, and Motivational Constructs ............................................................ 321

David Chan

Values ....................................................................................................................... 321

Cognitive Styles ........................................................................................................ 324

Motivational Constructs ........................................................................................... 327

Practical Considerations and Future Research Challenges ...................................... 332

Epilogue.................................................................................................................... 335

References ................................................................................................................ 336

TAF-Y100248_FARR-09-0801-C000toc.indd xTAF-Y100248_FARR-09-0801-C000toc.indd x 1/22/10 12:18:43 PM1/22/10 12:18:43 PM


simsr

Cross-Out

simsr

Replacement Text


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note



Contents xi

16 Chapter Practical Intelligence, Emotional Intelligence, and Social Intelligence .................. 339

Filip Lievens and David Chan

Defi nitions and Conceptualizations ......................................................................... 339

Measurement Approaches ........................................................................................344

Conceptual Framework for Examining Practical, Emotional, and

Social Intelligence ................................................................................................ 349

Strategies for Future Research ................................................................................. 351

Epilogue.................................................................................................................... 354

References ................................................................................................................ 355

4 Decisions in Developing, Selecting, Using, and PARTEvaluating PredictorsAnn Marie Ryan and Neal W. Schmitt, Coeditors

17 Chapter Decisions in Developing and Selecting Assessment Tools ...................................... 363

Nancy T. Tippins, Jone M. Papinchock, and Emily C. Solberg

Introduction .............................................................................................................. 363

Which Constructs Should Be Measured? .................................................................364

How Should the Constructs Be Measured? ..............................................................366

How Should Validity Evidence Be Gathered? ......................................................... 371

How Should Scores Be Used? .................................................................................. 373

Conclusions .............................................................................................................. 375

References ................................................................................................................ 375

18 Chapter Administering Assessments and Decision-Making ................................................. 377

R. Stephen Wunder, Lisa L. Thomas, and Zupei Luo

Unproctored Internet Testing ................................................................................... 377

Combination of Candidate Information: The Role of the Decision-Maker .............390

Conclusions .............................................................................................................. 396

References ................................................................................................................ 396

19 Chapter Evaluation of Measures: Sources of Error, Suffi ciency, and Contamination ........... 399

Michael J. Zickar, Jose M. Cortina, and Nathan T. Carter

Reliability .................................................................................................................400

Validity .....................................................................................................................402

Sources of Invariance: A Handful of Hypotheses ....................................................407

CFA MI/E in Selection Research ............................................................................. 411

Conclusions .............................................................................................................. 413

References ................................................................................................................ 414

20 Chapter Assessment Feedback ............................................................................................... 417

Manuel London and Lynn A. McFarland

Some Case Examples ............................................................................................... 417

TAF-Y100248_FARR-09-0801-C000toc.indd xiTAF-Y100248_FARR-09-0801-C000toc.indd xi 1/22/10 12:18:43 PM1/22/10 12:18:43 PM


simsr

Cross-Out

simsr

Replacement Text


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note



xii Contents

Feedback and Professional Standards ...................................................................... 419

Applicants’ Reactions to Feedback .......................................................................... 420

Benefi ts and Costs of Feedback ................................................................................ 426

Implications for Practice .......................................................................................... 431

Implications for Research ......................................................................................... 433

Conclusions .............................................................................................................. 434

References ................................................................................................................ 434

5 Criterion Constructs in Employee SelectionPARTKevin R. Murphy and Elaine D. Pulakos, Coeditors

21 Chapter The Measurement of Task Performance as Criteria in

Selection Research ................................................................................................... 439

Walter C. Borman, Rebecca H. Bryant, and Jay Dorio

Objective Criteria ..................................................................................................... 439

Subjective Criteria .................................................................................................... 441

Dimensionality of Job Performance .........................................................................447

Predictors of Task Performance Dimensions ........................................................... 454

References ................................................................................................................ 458

22 Chapter Adaptive and Citizenship-Related Behaviors at Work .............................................463

David W. Dorsey, Jose M. Cortina, and Joseph Luchman

Conceptualization .....................................................................................................463

Adaptive Behavior Defi ned ......................................................................................463

Citizenship Defi ned ..................................................................................................465

Individual Difference Predictors ..............................................................................468

Measurement ............................................................................................................ 474

Moderators and Mediators—Adaptability ............................................................... 476

Moderators and Mediators—Citizenship ................................................................. 477

Impact on Organizational Outcomes ....................................................................... 479

Too Much of a Good Thing? ....................................................................................480

Conclusions .............................................................................................................. 481

References ................................................................................................................ 482

23 Chapter Counterproductive Work Behavior and Withdrawal ................................................ 489

Maria Rotundo and Paul E. Spector

Nature of CWB ........................................................................................................ 489

Assessment of CWB ................................................................................................. 491

Potential Antecedents of CWB ................................................................................ 492

Environment ............................................................................................................. 499

Person and Environment ..........................................................................................502

Consequences ........................................................................................................... 503

Future Directions ......................................................................................................505

Conclusions ..............................................................................................................506

References ................................................................................................................506

TAF-Y100248_FARR-09-0801-C000toc.indd xiiTAF-Y100248_FARR-09-0801-C000toc.indd xii 1/22/10 12:18:43 PM1/22/10 12:18:43 PM


simsr

Cross-Out

simsr

Replacement Text


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note



Contents xiii

24 Chapter Defi ning and Measuring Results of Workplace Behavior ........................................ 513

Elaine D. Pulakos and Ryan S. O’Leary

Measuring Workplace Behavior Versus Results ...................................................... 514

Defi ning Individual Performance Objectives ........................................................... 516

Challenges Associated With Developing Individual Objectives and

Mitigation Strategies ............................................................................................ 519

Measuring Results of Performance Objectives ........................................................ 522

Challenges Associated With Measuring Results and Mitigation Strategies ............ 524

Individual Difference Predictors of Results ............................................................. 526

Conclusions .............................................................................................................. 527

References ................................................................................................................ 528

25 Chapter Employee Work-Related Health, Stress, and Safety ................................................ 531

Lois E. Tetrick, Pamela L. Perrewé, and Mark Griffi n

Healthy Workers ....................................................................................................... 531

Work Stress............................................................................................................... 535

Occupational Safety ................................................................................................. 541

Conclusions .............................................................................................................. 545

References ................................................................................................................546

26 Chapter Criterion Validity and Criterion Defi ciency: What We Measure

Well and What We Ignore ........................................................................................ 551

Jeanette N. Cleveland and Adrienne Colella

Work and Workforce in the 21st Century: Outmoded Assumptions and

Bases for Change .................................................................................................. 552

Criterion Problem in I-O Psychology ....................................................................... 552

Multilevel Issues in Defi ning Performance and Success ......................................... 558

“Closing In” on Criterion Defi ciency: One Approach to Bridging HR

Systems With Business Unit Strategy .................................................................. 562

Conclusions .............................................................................................................. 563

References ................................................................................................................564

6 Legal and Ethical Issues in Employee SelectionPARTP. Richard Jeanneret and Paul R. Sackett, Coeditors

27 Chapter Ethics of Employee Selection ................................................................................... 571

Joel Lefkowitz and Rodney L. Lowman

Some Meta-Issues ..................................................................................................... 571

Ethical Principles and Dilemmas ............................................................................. 576

Role of Ethical Codes in Professional Practice: Historical and Current

Perspectives .......................................................................................................... 580

Some Specifi c Issues and Sources of Ethical Problems ........................................... 582

Conclusions .............................................................................................................. 587

References ................................................................................................................ 589

TAF-Y100248_FARR-09-0801-C000toc.indd xiiiTAF-Y100248_FARR-09-0801-C000toc.indd xiii 1/22/10 12:18:43 PM1/22/10 12:18:43 PM


simsr

Cross-Out

simsr

Replacement Text


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note



xiv Contents

28 Chapter Professional Guidelines/Standards ........................................................................... 593

P. Richard Jeanneret and Sheldon Zedeck

Introduction ............................................................................................................. 593

Standards for Educational and Psychological Testing ........................................... 595

Principles for the Validation and Use of Personnel Selection Procedures .............604

Uniform Guidelines on Employee Selection Procedures ......................................... 610

Comparisons Among the Three Authorities ............................................................ 615

Final Thoughts ......................................................................................................... 621

Conclusions .............................................................................................................. 621

References ................................................................................................................ 623

29 Chapter A Sampler of Legal Principles in Employment Selection ........................................ 627

Frank J. Landy, Arthur Gutman, and James L. Outtz

Introduction .............................................................................................................. 627

Principles and Exemplar Case Law .......................................................................... 629

Conclusions ..............................................................................................................648

References ................................................................................................................648

30 Chapter Perspectives From Twenty-Two Countries on the Legal Environment

for Selection .............................................................................................................. 651

Paul R. Sackett, Winny Shen, Brett Myors, and Colleagues

Data Collection Methodology .................................................................................. 652

Discussion................................................................................................................. 673

Authors’ Note ........................................................................................................... 675

References ................................................................................................................ 675

7 Employee Selection in Specifi c Organizational PARTContexts

Rick Jacobs and Ann Howard, Coeditors

31 Chapter Selection and Classifi cation in the U.S. Military ..................................................... 679

Wayne S. Sellman, Dana H. Born, William J. Strickland, and Jason J. Ross

Military Personnel System ....................................................................................... 679

Indicators of Recruit Quality ....................................................................................680

Need for Military Selection ...................................................................................... 685

Short History of Military Personnel Testing (Pre-All Volunteer Force) ...................686

Moving to an All-Volunteer Force ............................................................................686

ASVAB Misnorming and Job Performance Measurement Project ..........................688

Enlisted Selection and Classifi cation in Today’s Military ........................................ 691

Enlistment Process ...................................................................................................692

Recruit Quality Benchmarks and Enlistment Standards .......................................... 693

Selection for Offi cer Commissioning Programs ...................................................... 693

Offi cer Retention and Attrition ................................................................................. 695

Offi cer Executive Development ................................................................................696

TAF-Y100248_FARR-09-0801-C000toc.indd xivTAF-Y100248_FARR-09-0801-C000toc.indd xiv 1/22/10 12:18:43 PM1/22/10 12:18:43 PM


simsr

Cross-Out

simsr

Replacement Text


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note



Contents xv

Command Selection and Career Broadening Experiences .......................................696

Defense Transformation in Military Selection .........................................................697

Conclusions .............................................................................................................. 698

References ................................................................................................................699

32 Chapter Public Sector Employment ....................................................................................... 705

Rick Jacobs and Donna L. Denning

Position Classifi cation in the Public Sector ..............................................................706

Civil Service Examinations ......................................................................................706

Linking Tests to Jobs: Validation and Its Many Forms............................................ 708

Creating a Talent Pipeline: Recruiting Candidates .................................................. 711

Promotional Processes: Using What We Know About People and Their

Capabilities to Our Advantage ............................................................................. 712

Personnel Decision-Making and Legal Jeopardy ..................................................... 714

Conclusions .............................................................................................................. 718

References ................................................................................................................ 719

33 Chapter Selection Methods and Desired Outcomes: Integrating Assessment

Content and Technology to Improve Entry- and Mid-Level Leadership

Performance ............................................................................................................. 721

Scott C. Erker, Charles J. Cosentino, and Kevin B. Tamanini

Current Business Trends Affecting Leadership Selection and Development .......... 722

Changing Behavioral Requirements for Leaders ..................................................... 725

Assessment Tools and Techniques ........................................................................... 728

Case Studies of Leadership Selection ...................................................................... 733

Conclusions .............................................................................................................. 737

References ................................................................................................................ 738

34 Chapter Blue-Collar Selection in Private Sector Organizations ............................................ 741

Wanda J. Campbell and Robert A. Ramos

Defi nition of Blue-Collar Jobs ................................................................................. 741

Environment ............................................................................................................. 741

Planning and Developing the Selection Process: Psychometric and

Practical Considerations ....................................................................................... 744

Planning and Developing the Selection Process: The Constituents,

Their Issues, and Preferences ............................................................................... 749

Implementing a Selection Procedure ....................................................................... 754

Maintaining a Selection Procedure .......................................................................... 756

Recruitment and Employee Development ................................................................ 759

References ................................................................................................................ 762

35 Chapter Selection for Service and Sales Jobs ........................................................................ 765

John P. Hausknecht and Angela M. Langevin

Nature of Service and Sales Work ........................................................................... 765

TAF-Y100248_FARR-09-0801-C000toc.indd xvTAF-Y100248_FARR-09-0801-C000toc.indd xv 1/22/10 12:18:43 PM1/22/10 12:18:43 PM



xvi Contents

Research on Selection for Service and Sales Workers ............................................. 768

Implications for Practice and Future Research ........................................................ 774

Conclusions .............................................................................................................. 776

References ................................................................................................................ 776

36 Chapter Selection in Multinational Organizations ................................................................ 781

Paula Caligiuri and Karen B. Paul

Employee Selection in Multinational Organizations ............................................... 781

Challenges of Cross-National Selection and Assessment ........................................ 785

International Assignments: Implications for Employee Selection ........................... 789

Conclusions .............................................................................................................. 795

References ................................................................................................................ 796

37 Chapter Selection for Team Membership: A Contingency and Multilevel Perspective ......... 801

Susan Mohammed, Jan Cannon-Bowers, and Su Chuen Foo

Conceptual Framework for Understanding Selection for Team Membership..........802

Individual-Level Considerations ..............................................................................804

Team-Level Considerations ......................................................................................809

Discussion................................................................................................................. 815

Conclusions .............................................................................................................. 817

References ................................................................................................................ 818

38 Chapter Selecting Leaders: Executives and High Potentials ................................................. 823

George C. Thornton III, George P. Hollenbeck, and Stefanie K. Johnson

Executives and High Potentials: Who Are They? What Do They Do? .................... 823

Executive Competencies and Attributes...................................................................824

Assessment Techniques ............................................................................................826

Executive Selection in an HRM System .................................................................. 830

Performance Reviews: The Role of Top Executives and Boards ............................. 830

Fit: Individuals, Team, Outcome, Followers, and Culture ....................................... 831

Does It Work? ........................................................................................................... 833

Conclusions .............................................................................................................. 834

References ................................................................................................................ 835

8 Milestones in Employee SelectionPARTWalter C. Borman and Benjamin Schneider, Coeditors

39 Chapter The Management Progress Study and Its Legacy for Selection ............................... 843

Ann Howard

The Foundations of the Assessment Center ............................................................. 843

The Beginnings of Managerial Assessment ............................................................. 845

Predicting Managerial Success ................................................................................ 847

Successful Managers’ Development Over Time ...................................................... 852

AC Advantages and Disadvantages .......................................................................... 855

TAF-Y100248_FARR-09-0801-C000toc.indd xviTAF-Y100248_FARR-09-0801-C000toc.indd xvi 1/22/10 12:18:43 PM1/22/10 12:18:43 PM


simsr

Cross-Out

simsr

Replacement Text


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note


watsonp

Sticky Note



Contents xvii

The MPS Selection Legacy ...................................................................................... 858

References ................................................................................................................ 863

40 Chapter Project A: 12 Years of R & D ................................................................................... 865

John P. Campbell and Deirdre J. Knapp

Origins of Project A ................................................................................................. 865

Enabling of Project A ...............................................................................................866

Specifi c Research Objectives ................................................................................... 867

Overall Research Design .......................................................................................... 867

Research Instrument Development: Predictors ........................................................ 869

Job Analyses and Criterion Development ................................................................ 873

Modeling the Latent Structure of Performance ....................................................... 875

Correlations of Past Performance With Future Performance .................................. 878

Criterion-Related Validation .................................................................................... 879

Some Broader Implications ...................................................................................... 882

Conclusions ..............................................................................................................884

References ................................................................................................................ 885

41 Chapter The Dictionary of Occupational Titles and the Occupational

Information Network ................................................................................................ 887

Norman Peterson and Christopher E. Sager

Milestones ................................................................................................................ 887

Organizing Framework ............................................................................................ 887

DOT .......................................................................................................................... 889

DOT to O*NET™ Transition ...................................................................................894

Development and Testing of a Prototype O*NET .................................................... 895

O*NET Becomes Operational ..................................................................................899

Conclusions ..............................................................................................................905

References ................................................................................................................906

42 Chapter Situational Specifi city and Validity Generalization .................................................909

Lawrence R. James and Heather H. McIntyre

Situational Specifi city .............................................................................................. 910

VG ............................................................................................................................ 912

Situational Specifi city Response to VG ................................................................... 914

Substantive Tests of Situational Specifi city.............................................................. 917

Conclusions .............................................................................................................. 918

References ................................................................................................................ 919

43 Chapter Employee Selection in Europe: Psychotechnics and the

Forgotten History of Modern Scientifi c Employee Selection .................................. 921

Jesús F. Salgado, Neil R. Anderson, and Ute R. Hülsheger

Introduction .............................................................................................................. 921

The European Origin of Scientifi c Employee Selection: The Years of Success

(1900–1945) .......................................................................................................... 921

TAF-Y100248_FARR-09-0801-C000toc.indd xviiTAF-Y100248_FARR-09-0801-C000toc.indd xvii 1/22/10 12:18:43 PM1/22/10 12:18:43 PM



xviii Contents

The Fall of the European Employee Selection Enterprise After

World War II: The Years of Decline and Stagnation (1945–1975) ...................... 931

The Resurgence of European Employee Selection: The Years of

Optimism and Growth (Since 1975).....................................................................934

Conclusions .............................................................................................................. 936

Acknowledgments .................................................................................................... 937

References ................................................................................................................ 937

44 Chapter Employee Selection: Musings About Its Past, Present, and Future .......................... 943

Robert M. Guion

Psychometric Musings .............................................................................................. 943

Musings on the Perpetual Disconnect: Research and Practice ................................ 950

Is Education Really Broadening? .............................................................................954

A Final Musing ......................................................................................................... 956

References ................................................................................................................ 956

Author Index................................................................................................................................. 959

Subject Index ................................................................................................................................ 983

TAF-Y100248_FARR-09-0801-C000toc.indd xviiiTAF-Y100248_FARR-09-0801-C000toc.indd xviii 1/22/10 12:18:43 PM1/22/10 12:18:43 PM



51

3 Validation Strategies for Primary Studies

Neal W. Schmitt, John D. Arnold, and Levi Nieminen

NATURE OF VALIDITY

Most early applications of the use of tests as decision-making tools in the selection of personnel in

work organizations involved a validation model in which the scores on tests were correlated with

some measure or rating of job performance, such as the studies of salespersons by Scott (1915) and

streetcar motormen by Thorndike (1911). This view of validity was reinforced in books by Hull

(1928) and Viteles (1932). Subsequent reviews by Ghiselli (1966, 1973) were similarly focused on

what was by then known as criterion-related validity.

During this time, there was a recognition that tests could and should be based on other logical

arguments as well. Standards for Educational and Psychological Tests (American Psychological

Association [APA], American Educational Research Association [AERA], and National Council

on Measurement in Education [NCME], 1954) identifi ed four aspects of validity evidence: content,

predictive, concurrent, and construct. With time, the predictive and concurrent aspects of validity

became seen as simply different research designs, the purpose of which was to establish a predictor-

criterion relationship; hence, they became known as criterion-related validity. Content and construct

validation were seen as alternate methods by which one could validate and defend the use of test

scores in decision-making. A much broader view of the nature of validity is accepted today, and in

general it is seen as the degree to which the inferences we draw from a set of test scores about job

performance are accurate.

TRINITARIAN VERSUS UNITARIAN IDEAS OF VALIDITY

The “trinitarian” approach to validity was popularized and incorporated in the 1974 revision

of the APA Standards. This conceptualization of validity holds that there are three approaches

to the validation of tests. Content validation involves a demonstration that test items are a

representative sample of the behaviors to be exhibited in some performance domain. Content

validation typically depends heavily on a job analysis that specifi es the tasks performed on

a job and how those tasks (or very similar tasks) are refl ected in the tests used to make deci-

sions. Criterion-related validation involves the demonstration that scores on a test are related

to job performance measures. If job performance measures and test scores are obtained from

job incumbents at approximately the same time, the study involves what has become known

as concurrent criterion-related validity. If test scores are collected from job applicants and

performance measures are collected some time after these individuals are hired, the study

represents a test of predictive criterion-related validity. Thousands of criterion-related studies

have been summarized in meta-analyses over the last 30 years, and Schmidt and Hunter (1998)

have summarized these meta-analyses. Construct validation often includes a series of studies or

collections of evidence that a psychological concept or construct explains test performance. In

TAF-Y100248_FARR-09-0801-C003.indd 51TAF-Y100248_FARR-09-0801-C003.indd 51 1/20/10 5:11:07 PM1/20/10 5:11:07 PM



52 Handbook of Employee Selection

addition, there should be support for the notion that the construct is central to the performance

of job tasks.

This separation of approaches to validity produced numerous problems, not the least of which

was the notion that there were times when one approach was to be preferred over another or that

there were different acceptable standards by which these different aspects of validity were to be

judged. Most important, however, was the realization on the part of measurement scholars that all

were aspects of construct validity—the theoretical reasonableness of our explanations of job behav-

ior. There was a realization that the inferences we derive from test scores was central to all valida-

tion work. Content validity and the practices usually associated with it were recognized as desirable

practices in the development of any test. Careful consideration of the “theory” and hypotheses that

underlie our conceptualization of performance and how the constructs central to job performance

are represented in our tests is always important and unifying insofar as our validation efforts are

concerned. Traditional criterion-related research represents one type of evidence that can be col-

lected to confi rm/disconfi rm these hypotheses. This “unitarian” approach to validity was strongly

argued in the 1985 Standards and has been incorporated most completely in the 1999 version of the

Standards (AERA, APA, & NCME, 1999).

DIFFERENT APPROACHES TO THE COLLECTION OF DATA ABOUT VALIDITY

Validity, as defi ned in the most recent version of the Standards (1999), is the degree to which

evidence and theory support the interpretation of test scores proposed by the test user. The user

must state explicitly what interpretations are to be derived from a set of test scores including the

nature of the construct thought to be measured. The document goes on to describe a variety of

evidence that can support such an interpretation. An evaluation of test themes, wording, item

format, tasks, and administrative guidelines all comprise the “content” of a test, and a careful

logical or empirical analysis of the relationship of this content to the construct measured as well

as expert judgments about the representativeness of the items to the construct measured supports

validity.

Validity evidence can also take the form of an examination of the response processes involved

in responding to an item. For example, in evaluating the capabilities of an applicant for a mechanical

job, we might ask the person to read a set of instructions on how to operate a piece of equipment

and then ask the applicant to demonstrate the use of the equipment. Because the equipment is

used on the job, it would seem valid, but suppose we also fi nd that test scores are highly related

to examinees’ vocabulary level. We would then want to know if vocabulary is necessary to learn

how to use this equipment on the job and depending on the answer to that question, we may want

to revise the test.

Yet a third piece of evidence might be to collect data regarding the internal structure of a test.

We would examine the degree to which different items in a test (or responses to an interview) yield

correlated results and whether items designed to measure different constructs can be differentiated

from items written to assess a different construct. Researchers interested in these questions use item

means, standard deviations, and intercorrelations as well as exploratory and confi rmatory analyses

to evaluate hypotheses about the nature of the constructs measured by a test.

Similar to looking at the internal structure of a test, researchers can also examine its external

validity by correlating the test with measures of theoretically similar or dissimilar constructs as

well as job performance measures that are hypothesized correlates of test scores. Validity in the

personnel selection area has been almost synonymous with the examination of the relationship

between test scores and job performance measures, most often referred to as criterion-related valid-

ity. Because there is a large body of primary studies of many job performance-test relationships,

one can also examine the extent to which tests of similar constructs are related to job performance

and generalize in a way that supports the validity of a new measure or an existing measure in a new

context. These are studies of validity generalization.




Validation Strategies for Primary Studies 53

Finally, and somewhat controversial among industrial-organizational (I-O) psychologists, the

Standards (1999) also suggest that researchers examine the intended and unintended consequences

of test use to make decisions. This evidence is referred to as consequential validity (Messick, 1998).

The consequences of most concern are the degree to which use of test scores results in dispro-

portionate hiring of one or more subgroups (e.g., gender, race, disabled). The 1999 version of the

Standards clearly refl ects a unitarian view of validation; the old tripartite notions and terms are not

mentioned.

The Society for Industrial and Organizational Psychology’s Principles (2003) described the

sources of validity evidence in a way that more refl ects the trinitarian approach to validation in that

it includes a relatively lengthy discussion of criterion-related evidence, evidence based on content,

and evidence based on the internal structure of tests and their relationships to other variables. The

Principles, like the Standards, also recognized the central role of psychological constructs in all

validation research.

Finally, some I-O psychologists have also noted that the traditional separation of reli-

ability and validity concepts may be inadequate (Lance, Foster, Gentry, & Thoresen, 2004;

Lance, Foster, Nemeth, Gentry, & Drollinger, 2007; Murphy & DeShon, 2000). Their ideas are

addressed in Chapter 2 of this volume. It is also the case that technology affords the opportunity

to make the traditional one-time criterion-related validity study an ongoing effort in which the

accumulation of predictor and criterion data can be collected and aggregated across time and

organizations.

VALIDATION IN DIFFERENT CONTEXTS

This chapter will discuss validation largely within the context of personnel selection. This is the

most common application of the various approaches to validation. It is also the most straightforward

example of how validation approaches can be applied.

There is a wide range of contexts in which the validation of measures is desirable; however,

organizations should, for example, ensure they are using “validated” tools and processes in their

performance management systems, in their assessments of training and development outcomes, in

their promotion and succession planning processes, etc.

Each of these circumstances is associated with its own set of challenges as the researcher designs

an appropriate validation study. However, the design of the well-constructed study by necessity will

follow the same logic as will be discussed for the personnel selection context. Following this logic,

the studies should be structured to include the following three elements:

1. Job analysis: The foundation of validation in employment settings always involves the

development of a clear understanding of job and organizational requirements. For exam-

ple, for promotion purposes these would be the requirements of the target job(s) into which

a person might be promoted. For training and development purposes, these would be the

meaningful outcomes in terms of on-the-job performance that are the focus of the training/

development efforts.

2. Systematic development: As measures are developed, they need to follow an architecture

that is fi rmly grounded in the results of the job analysis. As the development of the mea-

sures is planned and as the tools are being constructed, activities need to be focused on

ensuring that the measures are carefully targeted to address the intended constructs.

3. Independent verifi cation: Once the measures are developed, they need to be subjected

to independent verifi cation that they measure the intended constructs. At times, this can

involve statistical studies to determine whether the measures exhibit expected relation-

ships with other independent measures (e.g., Does the 360-degree assessment of leader-

ship behavior correlate with an independent interview panel’s judgment of a leader’s





behavior?). Often, the independent verifi cation is derived from structured expert reviews

of the measures that are conducted prior to implementation. Regardless of the method, this

“independent verifi cation” is a necessary aspect of verifying the validity of a measure.

STRONG VERSUS WEAK INFERENCES ABOUT VALIDITY

The fi eld’s evolving conceptualization of validity has important implications for I-O researchers

concerned with designing and conducting primary validation studies. Given that validation is a

process of collecting evidence to support inferences derived from test scores (e.g., that a person

will perform effectively on a job), the confi dence with which inferences are made is a function of

the strength of the evidence collected. In this way, the strength of the validity evidence refers to the

probability that the inference is correct, with “stronger” evidence connoting higher probabilities.

Consistent with the unitarian view, validation can be viewed as a form of hypothesis testing (Landy,

1986; Messick, 1975), and judgments of validity are to be based on the same host of considerations

applicable to judgments concerning the veracity with which a null hypothesis would be rejected in

any psychological research (e.g., the extent to which threats to validity are ruled out; see Cook &

Campbell, 1979). Thus, it is critical for researchers designing and conducting local validation stud-

ies to concentrate their efforts on ensuring that studies result in strong evidence for the inferences

they wish to make in much the same way that they would otherwise “defend” their conclusions in a

hypothesis testing situation.

Gathering stronger evidence of validity almost always necessitates increased effort, resources,

and/or costs (e.g., to gain larger sample sizes or expand the breadth of the criterion measures).

Thus, a key decision for researchers designing primary validation studies involves determin-

ing how to optimize the strength of the study (assurance that inferences are correct) within the

bounds of certain practical limitations and organizational realities. Situations may vary in terms

of the extent to which feasibility drives the researcher’s choice among validation strategies. In

some cases, situational limitations may be the primary determinant of the validation strategies

available to researchers. For example, for situations in which adequately powered sample sizes

cannot be achieved, validation efforts may require use of synthetic validity strategies (Scherbaum,

2005), transporting validity evidence from another context that is judged to be suffi ciently simi-

lar (Gibson & Caplinger, 2007), generalizing validity across jobs or job families on the basis of

meta-analytic fi ndings (McDaniel, 2007; Rothstein, 1992), or relying on evidence and judgments

that the content of the selection procedures is suffi ciently similar to job tasks to support their use

in decision-making. Other factors noted by the Principles that may serve to limit the feasibility

of certain validation strategies include unavailability of criterion data, inaccessibility to subject

matter experts (SMEs) as might be the case when consulting SMEs would compromise test secu-

rity, dynamic working conditions such that the target job is changing or does not yet exist, and

time and/or money. Clearly then, validation strategy needs to account for feasibility-driven con-

siderations and researchers’ judgment about the strength of evidence required. Further, because

these demands are often competing, researchers are frequently required to make the best of a less

than optimal situation.

Given the need to balance several competing demands (e.g., issues of feasibility limiting the

approach that can be taken vs. upholding high standards of professionalism and providing strong

evidence to support key inferences), it is essential that researchers understand the various factors

that have potentially important implications for the strength of evidence that is required in a given

validation scenario. In other words, part of the decision process, with regard to planning and imple-

menting validation strategy, is a consideration of how strong the evidence in support of key infer-

ences ought to be. The basic assumption here is that different situations warrant different strategies

along several dimensions (Sussman & Robertson, 1986), one of which has to do with the strength of

evidence needed in support of inferences. Consideration of how certain situational factors inform

the adequacy of a validation effort does not imply that the validation researcher adopt a minimalist





approach. Rather, all validation studies and selection practices should aspire to the ethical and

professional guidelines offered in the Principles, which means using sound methods rooted in sci-

entifi c evidence and exhibiting high standards of quality. However, the Principles’ guidelines are

not formulaic to the exclusion of professional judgment, nor are their applications invariant across

circumstances. In the following paragraphs, several factors are identifi ed that have potential impli-

cations for the strength of the evidence needed by a local validation effort. In general, these factors

represent situations in which conclusions regarding the validity of a selection practice need to be

made with a higher degree of confi dence than usual. In turn, these situations tend to warrant intensi-

fi ed research strategies.

SITUATIONAL FACTORS INFLUENCING THE STRENGTH OF EVIDENCE NEEDED

Although it is beyond the scope of this chapter to describe in full detail the legal issues surrounding

validation research and selection practice (see Chapters 29 and 30, this volume, for further discus-

sions of legal issues), it would be diffi cult if not impossible to describe applied validation strategy

without underscoring the infl uence of litigation or the prospect of litigation. It is becoming almost

cliché to state that, in circumstances in which there is a relatively high probability of litigation

regarding selection practices, validation evidence is likely to function as a central part of defending

selection practices in the courtroom. Indeed, much validation research is stimulated by litigation,

whether post facto or in anticipation of lawsuits. Within this context, researchers make judgments

regarding the potential for litigation and adapt their validation strategies accordingly. Numerous

contextual factors contribute to the probability that litigation will occur. A primary example has to

do with the type of selection instrument being validated and the potential for adverse impact, or the

disproportionate rejection of identifi able subgroups. Tests that have historically resulted in adverse

impact, such as tests of cognitive ability (Schmitt, Rogers, Chan, Sheppard, & Jennings, 1997)

or physical ability (Arvey, Nutting, & Landon, 1992; Hogan & Quigley, 1986) tend to promote

more litigation, and researchers validating these instruments in a local context should anticipate

this possibility. Similarly, selection instruments with low face validity (i.e., the test’s job relevance

is not easily discerned by test-takers) are more likely to engender negative applicant reactions

(Shotland, Alliger, & Sales, 1998), and decisions based on such tests may lead to applicant percep-

tions of unfairness (Cropanzano & Wright, 2003). In their recent review of the antecedents and

consequences of employment discrimination, Goldman, Gutek, Stein, & Lewis (2006) identifi ed

employee perceptions of organizational and procedural justice as important antecedents of dis-

crimination lawsuits. In addition to considering characteristics of the selection instrument(s) being

validated, lawsuits over selection practice are more frequent in some industry (e.g., government)

and job types (Terpstra & Kethley, 2002).

Researchers should also consider the implications and relative seriousness of selection systems

resulting in hiring decisions that are false positives or false-negative errors. A false positive is made

by selecting an unqualifi ed individual whose performance on the job will be low, whereas a false-

negative error is made by rejecting a qualifi ed individual whose performance on the job would have

been high. Making an error of either type can be potentially costly to the organization. However, the

relative impact of such errors can differ by occupation type and organizational context. For exam-

ple, the negative impact of a false positive in high-risk occupations (e.g., nuclear power plant opera-

tor or air-traffi c controller) or high visibility occupations (e.g., Director of the Federal Emergency

Management Agency [FEMA]) can be catastrophic, threaten the organization’s existence, and so

on (Picano, Williams, & Roland, 2006). Alternatively, for occupations that are associated with less

risk, such that failure on the job does not have catastrophic consequences for the organization or

larger society, or when organizations use probationary programs or other trial periods, the cost of

false-positive errors may be relatively low. Although validation efforts in both situations would be

concerned with selection errors and demonstrating that use of tests can reduce the occurrence and

negative consequences of such errors, clearly there are some situations in which this would be more





of a central focus of the validation effort. It is our contention that validating selection systems for

high-risk occupations are a special circumstance warranting particularly “watertight” validation

strategies in which strong evidence should be sought to support the inferences made. In these cir-

cumstances, a test with low validity (e.g., less than r = .10) might be used to make hiring decisions

if that relationship is found to be statistically signifi cant.

In some circumstances, the cost of false negatives is more salient. For example, strong evidence

of a test’s validity may be warranted when an organization needs to fi ll a position or several posi-

tions, but applicants’ test scores are below some acceptable standard, indicating that they are not fi t

to hire (i.e., predicted on-the-job performance is low or very low). In this case, the organization’s

decision to reject an applicant on the basis of their test scores would leave a position or several

positions within the organization vacant, a costly mistake in the event that false-negative errors

are present. Demonstrable evidence to support the test’s validity would be needed to justify such a

decision, and in essence, convince the organization that they are better off with a vacant position

than putting the wrong person in the job. In these instances, one might want evidence of a larger

test-criterion relationship (perhaps greater than r = .30) to warrant use of this test and the possible

rejection of competent applicants.

The possibility of false negatives becomes a special concern when members of some subgroup(s)

are selected less frequently than members of another subgroup. When unequal ratios of various

subgroups are selected, the organization must be prepared to show that false negatives are not

primarily of one group as opposed to another. When this is impossible, the legal and social costs

can be very high. Concern about these costs is another reason to be concerned about the false

negatives aside from the concerns associated with possible vacant positions.

Personnel psychologists have long been aware of the fact that the utility of selection systems

increase as a function of selectivity, such that selection instruments even modestly related to impor-

tant outcomes can have large payoffs when there are many applicants from which only a few are to

be selected (Brogden, 1951, 1959). On the other hand, as selection ratios become extremely liberal,

such that nearly all applicants are accepted, even selection instruments highly related to perfor-

mance have less positive implications for utility. From a purely utilitarian perspective, it would

seem logical that demonstrating test validity is less of an impetus when selection ratios are liberal

(because even the best tests will have little effect) and more of an impetus when selection ratios are

low. The consequences of large-scale applicant rejections would also seem to justify more rigorous

validation methods from societal and legal perspectives. However, these extremes likely occur less

frequently in practice, as indicated by Schmidt and Hunter (1998), who cite typical selection ratios

that are moderate, ranging on average from .30 to .70.

In licensing examinations, this utility perspective takes a different form because the major pur-

pose of these examinations is to protect the public from “injuries” related to incompetent practice.

In this case, the license-no license decision point using test scores is usually set at a point that is

judged to indicate “minimal competence.” Depending on the service provided (e.g., hairdresser vs.

surgeon), the cost of inappropriately licensing a person could be very different. On the other hand,

certifi cation examinations are usually oriented toward the identifi cation of some special expertise in

an area (e.g., pediatric dentistry or forensic photography), hence a decision as to a score that would

warrant certifi cation might result in the rejection of larger numbers or proportions of examinees.

The cost-benefi t balance in this situation (assuming all are minimally competent) might accrue

mostly to the individual receiving the certifi cation in the form of greater earning power.

Another factor that can affect the extent of the local validation effort that is required is the avail-

ability of existing validation research. The Principles describes three related validation strategies

that can be used as alternatives to conducting traditional local validation studies or to support the

conclusions drawn at the primary study level. First, “transportability” of validity evidence involves

applying validity evidence from one selection scenario to another, on the basis that the two con-

texts are judged to be suffi ciently similar. Specifi cally, the Principles note that researchers should

be concerned with assessing similarity in terms of job characteristics [e.g., the knowledge, skills,





and abilities (or KSAs) needed to perform the job in each context], job tasks and content, applicant

pool characteristics, or other factors that would limit generalizability across the two contexts (e.g.,

cultural differences). Assessing similarity in this manner usually requires that researchers conduct

a job analysis or rely on existing job analysis materials combined with their own professional exper-

tise and sound judgment.

Second, synthetic validity is a process in which validity for a test battery is “synthesized” from

evidence of multiple predictor-job component relationships (Peterson, Wise, Arabian, & Hoffman,

2001; Scherbaum, 2005). Job analysis is used to understand the various components that comprise

a particular job, and then predictor-job-component relationships are collected for all available jobs

with shared components. Because evidence can be drawn from other jobs besides the focal job, syn-

thetic validity may be a particularly useful strategy for organizations that have too few incumbents

performing the focal job to reach adequate sample sizes for a traditional criterion-related study

(Scherbaum, 2005).

Third, validity generalization involves using meta-analytic fi ndings to support the conclusion

that predictor-criterion validity evidence can be generalized across situations. Like transportability

strategies, meta-analytic fi ndings provide researchers with outside evidence to support inferences

in a local context. The argument for validity generalization on the basis of meta-analyses is that

some selection tests, such as cognitive ability tests (Ones, Viswesvaran, & Dilchert, 2005), are valid

across selection contexts. Thus, the implication is that with validity generalization strategies, unlike

transportability, in-depth job analyses or qualitative studies of the local organizational context are

unnecessary. In support of this assertion, the program of research initiated by Schmidt and Hunter

and colleagues (for review, see Schmidt & Hunter, 2003) has argued that between-study variability

in validity coeffi cients can be largely attributed to statistical artifacts, such as range restriction,

unreliability, or sampling error. However, caution is warranted to the extent that meta-analyses have

identifi ed substantive moderators, or in the presence of strong theory indicating that some variable

may moderate the magnitude of validity. Further, with regard to generalization across contexts,

inferences drawn from meta-analytic fi ndings are limited to the contexts of those studies included

in the meta-analysis. At minimum, meta-analytic fi ndings should be referenced in test development

and can be used to supplement evidence at the local level, either via theoretical or statistical means

(Newman, Jacobs, & Bartram, 2007). The argument for more direct use of validity generalization

strategies is dependent on the strength of the meta-analytic fi ndings.

DESIGN CONSIDERATIONS WHEN STRONG EVIDENCE IS NEEDED

On the basis of the preceding discussion, situational factors can affect the feasibility and appropri-

ateness of the validation models applied to a given selection context. Moreover, researchers should

be particularly attuned to contextual variables that warrant an increased concern for demonstrating

the strength of evidence collected and high levels of confi dence in the inferences to be made. The

validity strategies used refl ect consideration of these contextual factors and others. For instance, in

apparent response to increasing litigation, Boehm (1982) found longitudinal trends suggesting that

published validation studies were using larger sample sizes, relying on supervisory rating criteria

less frequently, and utilizing more predictive (as opposed to concurrent) criterion-related designs.

Similarly, the discussion that follows is focused on identifying a handful of actionable validation

strategies to be considered by researchers when particularly strong evidence is needed.

Importance of the Nomological NetBinning and Barrett (1989) offered a thorough conceptualization of the nomological network

implicit in validity models. Their model identifi es multiple inferential pathways interrelating

psychological constructs and their respective operational measures. Inferential pathways in the

model are empirically testable using observed variables (e.g., linkages between operationalized

measures of constructs and linkages between constructs and their operationalized measures).





Others may be theoretically or rationally justifi ed (e.g., construct to construct linkages) or tested

using latent variable models, although these applications are relatively rare in personnel selec-

tion research (see Campbell, McHenry, & Wise, 1990, for an attempt to model job performance).

Consistent with the unitarian conceptualization of validity, all validity efforts in a selection con-

text are ultimately concerned with demonstrating that test scores predict future job performance,

and each of the various inferential pathways represents sources or types of evidence to support

this common inference. Binning and Barrett (1989, p. 482) described how “truncated” validation

strategies often concentrate exclusively on demonstrating evidence for a single inferential path-

way and as a result provide only partial support for conclusions regarding test validity. A more

cogent argument for validity is built upon demonstration of strong evidence for several inferential

pathways within the nomological network. For example, in addition to demonstrating a statisti-

cal relationship between operational measures from the predictor and performance domain, as

is commonly the main objective in criterion-related validity studies, researchers should provide

evidence of the psychological constructs underlying job performance (as well as the predictor

measures) and demonstrate that the criterion measure adequately samples constructs from the

performance domain.

Criterion ConcernsIn practice, criterion-related validity studies are often criticized for failing to adequately address

validity issues surrounding the criterion measure(s) used. The relative lack of scientifi c scrutiny

focused on criteria, termed the “criterion problem” (Austin & Villanova, 1992), has been a topic

of discussion among personnel psychologists for years (Dunnette, 1963; Fiske, 1951; Guion, 1961).

Universal to these discussions is the call for more rigorous validation evidence with respect to

the criteria that are used. Binning and Barrett (1989) outlined this task, underscoring two inter-

related goals for the validation researcher. First, they suggested that the selection of criteria should

be rooted in job analysis to the same extent that selection of predictors traditionally are (i.e.,

more attention to rigorous “criterion development”). Other considerations relevant to the task of

criterion development and validation include the use of “hard” or objective criteria versus more

proximal behaviors that lead to these outcomes (Thayer, 1992), use of multiple relevant crite-

ria as opposed to a single overall criterion (Dunnette, 1963), and the possibility that criteria are

dynamic (Barrett, Caldwell, & Alexander, 1985). Second, researchers should be concerned with

demonstrating evidence of construct-related validity for the criterion. Investigators must specify

the latent dimensions that underlay the content of their criterion measures. This involves expansion

of the nomological network to include inferences that link the criterion measure(s) to constructs in

the performance domain (e.g., by demonstrating that criterion measures are neither contaminated

nor defi cient with respect to their coverage of the intended constructs in the performance domain)

and link constructs in the performance domain to job demands that require specifi c ability or moti-

vational constructs (e.g., by demonstrating through job analysis that constructs in the performance

domain are organizationally meaningful). Campbell and his colleagues (e.g., Campbell, McCloy,

Oppler, & Sager, 1993) have repeatedly emphasized this emphasis on the nature of criteria or per-

formance constructs. These authors make the somewhat obvious, although often overlooked, point

that performance should be defi ned as behavior (“what people actually do and can be observed”);

the products of one’s behavior, or what are often called “hard criteria,” are only indirectly the

result of one’s behavior and other factors that are not attributable to an individual job incumbent.

Further, we may consider relatively short term or proximal criteria or distal criteria, such as the

impact of one’s career on some fi eld of interest. Any specifi cation of a performance or criterion

domain must also consider the impact of time (Ackerman, 1989; Henry & Hulin, 1989). In any

study of performance, these various factors must be carefully considered when one decides on the

nature of the performance constructs and actual operationalizations of the underlying constructs

and how those measures might or might not be related to measures of other constructs in the

domain of interest.





Multiple Inferences in Validation ResearchGathering evidence to support multiple inferences within a theoretically specifi ed nomological net-

work resembles a pattern-matching approach. The advantage of pattern-matching research strate-

gies is that stronger support for a theory can be gained when complex patterns of observed results

match those that are theoretically expected (Davis, 1989). Logically, it would be less likely that a

complex pattern of results would be observed simply because of chance. In addition, when experi-

mental control of potentially confounding variables is not possible, pattern matching can be used

to preempt alternative explanations for the observed relationships (i.e., threats to validity; Cook &

Campbell, 1979).

Campbell and Fiske (1959) described a specifi c application of pattern matching that can be used

to support inferences regarding the appropriateness with which a construct is operationalized in

measurement. A slight variation of the convergence/divergence idea is included in Binning and

Barrett’s (1989) elaborated nomological network model. The elaborated model includes inferences

regarding the proximal versus distal nature of relationships between constructs. A sound argument

for validity can be made on the basis of results that indicate a reliable pattern, in which strong sta-

tistical relationships are obtained for constructs that are theoretically proximal to one another and

weaker statistical relationships are obtained for constructs that are theoretically more distal. This, of

course, is the rationale underlying mediation (Baron & Kenny, 1986) and the idea of causal chains.

Even without testing mediation directly, a strong argument can be made by demonstrating that a

predictor correlates highly with proximal operational measures of the criterion construct, and that

this relationship is attenuated as criterion measures become more distally related to the construct

space. For example, Thayer (1992) noted that in many cases objective outcome criteria, although

important at the organizational level, are distal indicators of the job performance construct because

they are contaminated by factors outside of the employee’s control. Also, low reliability was found

for the theoretically removed, “hard” criteria measures discussed by Thayer. In contrast, criteria

that assess employee behavior directly as defi ned above are more proximal to the job performance

construct. To the extent that these proximal criteria can be identifi ed and measured, stronger support

for validity is gained in the event that test scores correlate more highly with proximal as compared

to distal criteria.

A more extensive form of pattern matching involves the use of multiple studies, or research pro-

grams, to corroborate evidence of validity. Again, the logic is straightforward; stronger evidence

is gained when a constellation of fi ndings all lead to the same conclusion. Sussman and Robertson

(1986) suggested that programs of research could be undertaken, “composed of multiple stud-

ies each utilizing a different design and aimed at collecting different types of evidence” (p. 467).

Extending the rationale of the multi-trait multi-method (MTMM) (Campbell & Fiske, 1959), con-

vergent evidence across studies may indeed be stronger if gained through different research designs

and methods. Landy’s (1986) assertion that test validation is a form of hypothesis testing, and that

judgments of validity are to be based on a “preponderance of evidence” (p. 1191; Guion, as cited in

Landy, 1986), provides the context for consideration of research strategies such as quasi-experimen-

tal designs (Cook & Campbell, 1979) and program evaluation research (Strickland, 1979). Binning

and Barrett (1989) presented a similar rationale by calling for “experimenting organizations”

(p. 490) in which local validation research is treated as an ongoing and iterative process. Published

research on use of these research methods in a selection-validation context remains sparse to date.

SPECIAL CONSIDERATIONS IN CRITERION-RELATED STUDIES

In the event that a criterion-related strategy is part of the validation effort undertaken, a special set

of considerations is relevant. Power analysis is a useful framework for interrelating the concepts

of statistical signifi cance, effect size, sample size, and reliability (Cohen, 1988) and has design

and evaluation implications for the statistical relationships sought in criterion-related studies. For





instance, the sample size needed to demonstrate a statistically signifi cant predictor-criterion rela-

tionship decreases as the magnitude of the relationship that exists between predictor and criterion

(i.e., effect size) increases. Sussman and Robertson (1986), in their assessment of various predic-

tive and concurrent validation designs, found that those strategies that allowed larger sample sizes

gained a trivial increment in power. This suggests that, as long as sample sizes can support the

use of a criterion-related design, further attention toward increasing N may not reap large benefi ts.

Other factors affecting power include the interrelatedness and number of predictors used, such that

the addition of nonredundant predictors increases power (Cascio, Valenzi, & Silbey, 1978). The

reliability of the predictors and criteria and the decision criteria used for inferring that a relation-

ship is nonzero also impact power.

By incorporating power analysis in validation design, researchers can increase the likelihood

that relationships relevant to key inferences will be tested with suffi cient sample size upon which to

have confi dence in the results. However, from a scientifi c standpoint, the importance of demonstrat-

ing that predictor-criterion relationships are statistically signifi cant may be overstated, given that

relationships, which may not be practically meaningful, can reach statistical signifi cance with large

enough sample sizes. For instance, a statistically signifi cant relationship, in which a test accounts for

less than 5% of the variance in job performance, is not unequivocal support for the test’s use. This

is especially evident when there is reason to suggest that other available tests could do a better job

predicting performance. Nonetheless, we agree with Cortina and Dunlap’s (1997) overall contention

that statistical signifi cance testing remains a useful tool to researchers when decision criteria are

needed. For this reason, strong evidence in support of the predictor-criterion relationship should be

derived based on the size and signifi cance of the validity coeffi cients.

Operationally, there are several other important considerations in criterion-related research (e.g.,

job analyses that support the relevance of predictor and criterion constructs and the quality of the

measures of each set of constructs). However, those concerns are addressed repeatedly in textbooks

(e.g., Guion, 1998; Ployhart, Schneider, & Schmitt, 2006). In the next section of this chapter, we

address a very important concern that is rarely discussed.

CONCERNS ABOUT THE QUALITY OF THE DATA: CLEANING THE DATA

Once data have been collected, quality control techniques should be applied to ensure that the data

are clean before proceeding to statistical analysis. Some basic quality control techniques include

double-checking data for entry errors, spot checks for discrepancies between the electronic data

and original data forms, inspecting data for out-of-range values and statistical outliers, and visual

examination of the data using graphical interfaces (e.g., scatter plots, histograms, stem-and-leaf

plots). Special concern is warranted in scenarios with multiple persons accessing and entering data

or when data sets from multiple researchers are to be merged. Although these recommendations

may appear trite, they are often overlooked and the consequence of erroneous data can be profound

for the results of analyses and their interpretations.

A study by Maier (1988) illustrated, in stepwise fashion, the effects of data cleaning procedures

on validity coeffi cients. Three stages of data cleaning were conducted, and the effects on correla-

tions between the Armed Services Vocational Aptitude Battery (ASVAB) and subsequent perfor-

mance on a work sample test for two military jobs (radio repairers and automotive mechanics) were

observed. Selection was based on the experimental instrument (the ASVAB) and the work sample

criterion tests were administered to incumbents in both occupations after some time had passed. In

Phase 1 of the data cleaning process, the sample was made more homogenous for the radio repairers

group by removing the data of some employees who received different or incomplete training before

criterion data collection. In comparison to the total sample, the validity coeffi cient for the remain-

ing, more representative group that had received complete training before criterion collection was

decreased (from .28 to .09). The initial estimate had been infl ated because of the partially trained

group having scored low on the predictor and criterion.





In Phase 2, scores on the criterion measure (i.e., ratings from a single rater on a work sample)

were standardized across raters. There were signifi cant differences between raters that were attrib-

uted to different rating standards and not to group differences in ratees, such as experience, rank,

or supervisor performance ratings. The raters were noncommissioned offi cers and did not receive

extensive training in the rating task, so that differences between raters in judgmental standards

were not unexpected. As a result, the validity coeffi cients for both jobs increased (radio repairers,

from .09 to .18; automotive mechanics, from .17 to .24). In Phase 3, validity coeffi cients were cor-

rected for range restriction, which again resulted in an increase in the observed validity coeffi cients

(radio repairers, from .18 to .49; automotive mechanics, from .24 to .37). Maier noted that the fi nal

validity coeffi cients were within the expected range on the basis of previous studies.

The Maier (1988) study is illustrative of the large effect that data cleaning can have for attaining

more accurate estimates of validity coeffi cients in a predictive design scenario. Several caveats are

also evident, so that researchers can ensure that data cleaning procedures conducted on sound pro-

fessional judgment are not perceived as data “fudging.” First, the cleaning procedures need to have

a theoretical or rational basis. Researchers should document any decision criteria used and the sub-

stantive changes that are made. For example, researchers should record methods used for detecting

and dealing with outliers. In addition, a strong case should be built in support of any decisions made.

The researcher bears the burden of defending each alteration made to the data. For example, in the

Maier study, the decision to standardize criterion data across raters (because raters were relatively

untrained and used different rating standards) was supported by empirical evidence that ruled out

several alternative explanations for the mean differences observed between raters.

MODES OF DECISION-MAKING AND THE IMPACT ON UTILITY AND ADVERSE IMPACT

If we have good quality data, it still matters how we use those data in making decisions as to

whether or not use of the test produces aggregated performance improvements. In this section, we

will discuss the impact of various modes of decision-making on two outcomes that are of concern

in most organizations: Overall performance improvement or utility and adverse impact on some

protected group defi ned as unequal proportions of selection across subgroups. Advancing both out-

comes is often in confl ict, especially when one uses cognitive ability tests to evaluate the ability of

members of different racial groups or physical ability when evaluating male and female applicants

for a position. Measures of some other constructs (e.g., mechanical ability) produce gender or race

effects, but the subgroup differences that are largest and affect the most people are those associated

with cognitive and physical ability constructs.

TOP-DOWN SELECTION USING TEST SCORES

If a test has a demonstrable relationship to performance on a job, it is the case that the optimal util-

ity in terms of expected employee performance will occur when the organization selects the top

scoring persons on the test to fi ll its positions (Brown & Ghiselli, 1953). Expected performance is

a direct linear function of the test score-performance relationship in the situation in which the top

scoring individuals are selected. However, use of tests in this fashion when it is possible will mean

that lower scoring subgroups will be less likely to be selected (Murphy, 1986). This confl ict between

maximization of expected organizational productivity and adverse impact is well known and has

been quantifi ed for different levels of subgroup mean differences in ability and selection ratios

(Sackett, Schmitt, Ellingson, & Kabin, 2001; Sackett & Wilk, 1994; Schmidt, Mack, & Hunter,

1984). For social, legal, and political reasons as well as long-term organizational viability in some

contexts, the adverse impact of a strict top-down strategy of test use often cannot be tolerated. For

these reasons as well as others, researchers and practitioners have often experimented with and used

other ways of using test scores.





BANDING AND CUT SCORES

One method of reducing the consequences of subgroup differences in test scores and top-down

selection is to form bands of test scores that are not considered different usually using a statistical

criterion known as the standard error of the difference, which is based on the reliability of the test.

Most of us are familiar with a form of banding commonly used in academic situations. Scores on

tests are usually grouped into grades (e.g., A, B, C, etc.) that are reported without specifi c test score

information. So persons with scores of 99 and 93 might both receive an A in a course just as two

with scores of 88 and 85 would receive a B. The theory in employment selection use of banding is

that the unreliability inherent in most tests makes the people within a band indistinguishable from

each other just as occurs when grades are assigned to students.

Because minorities tend to score lower on cognitive ability tests, creating these bands of indis-

tinguishable scores helps increase the chances that minority applicants will fall in a top band and be

hired. There are two ways in which banding can increase minority hiring. One is to make the bands

very wide so that a greater number of minority test scorers will be included in the top bands. Of

course, a cynic may correctly point out that a test of zero reliability will include everyone in the top

band and that this approach supports the use of tests with low reliability. A second way in which to

impact the selection of minority individuals is the manner in which individuals are chosen within a

band. The best way to increase the selection of minority individuals is to choose these persons fi rst

within each band before proceeding to consider other individuals in the band, but this has proven

diffi cult to legally justify in U.S. courts (Campion et al., 2001). Other approaches to selection within

a band include random selection or selection on secondary criteria unrelated to subgroup status, but

these procedures typically do not affect minority hiring rates in practically signifi cant ways (Sackett

& Roth, 1991). A discussion of various issues and debates regarding the appropriateness of banding

is contained in an edited volume by Aguinis (2004).

An extreme departure from top-down selection occurs when an organization sets a minimum

cutoff test score such that individuals above some score are selected whereas those below that score

are rejected. In essence, there are two bands of test scores—those judged to represent a passable

level of competence and those representing a failing level of performance. Perhaps the most com-

mon use of cutoff scores is in licensing and credentialing, in which the effort is usually to identify

a level of expertise and knowledge of the practice of a profession below which a licensure candidate

is likely to bring harm to clients. In organizational settings a cutoff is often required when selection

of personnel is done sequentially over time rather than from among a large number of candidates

at a single point in time. In this case, hire-reject decisions are made about individuals and a pass

score is essential.

Use of a single cutoff score will certainly reduce the potential utility inherent in a valid test

because it ignores the individual differences in ability above the test score cutoff. There exists a

great deal of evidence (e.g., Coward & Sackett, 1990) that test score-job performance relationships

are linear throughout the range of test scores. However, using a minimum cut score on a cognitive

ability test on which we usually see the largest minority-majority differences to select employees

and selecting above that cutoff on a random basis or on the basis of some other valid procedure that

does not display subgroup differences may very much reduce the adverse impact that usually occurs

with top-down selection using a cognitive ability test.

Perhaps the biggest problem with the use of cutoff scores is deriving a justifi able cut score.

Setting a cutoff is always judgmental. Livingston (1980) and Cascio, Alexander, and Barrett (1988)

among others have usually specifi ed the following as important considerations in setting cutoffs:

the qualifi cations of the experts who set the cutoff, the purpose for which the test is being used, and

the consideration of the various types of decision errors that can be made (i.e., denying a qualifi ed

person and accepting an unqualifi ed individual). One frequently used approach is the so-called

Angoff method, in which a representative sample of experts examines each test item and deter-

mines the probability that a minimally competent person (the defi nition and experts’ understanding





of minimally competent is critical) would answer the question correctly. These probabilities are

summed across experts and across items. The result is the cutoff score. A second approach to the

setting of cut scores is to set them by reference to some acceptable level of performance on a crite-

rion variable. In this case, one could end up saying that an individual with a score of 75 on some test

has a 10% (or any percent) chance of achieving success on some job. However, this “benchmark-

ing” of scores against criteria does not resolve the problem because someone will be asked to make

sometimes equally diffi cult decisions about what constitutes acceptable performance. Cizek (2001)

provided a comprehensive treatment of methods of setting performance standards.

The use of score cutoffs to establish minimum qualifi cations or competency is common in licens-

ing exams. Credentialing exams may require evidence of a higher level of skill or performance

capability in some domain, but they too usually require only a “pass-fail” decision. Validation of

these cutoffs almost always relies solely on the judgments of experts in the performance area of

interest. In these cases, careful explication of the behaviors required to perform a set of tasks and

the level of “acceptable” performance is essential and likely the only possible form of validation.

USING PROFILES OF SCORES

Another possibility when scores on multiple measures of different constructs are available is that a

profi le of measured abilities is constructed and that this profi le is matched to a profi le of the abilities

thought to be required in a job. In this instance, we might measure and quantify the type of job expe-

riences possessed by a job candidate along with their scores on various personality tests, and their

oral communications and social skills as measured in an interview and scores on ability tests. If this

profi le of scores matches that required in the job, the person would be selected. This contrasts with

the traditional approach described in textbooks in which the person’s scores on these tests would be

linearly related to performance and combined using a regression model so that each score was opti-

mally linearly related to job performance. In using profi les, one is interested in patterns of scores

rather than an optimally weighted composite. Use of profi les of scores presents various complex

measurement and statistical problems of which the user should be aware (Edwards, 2002). Instances

in which selection decisions are made in this fashion include individual assessments (Jeanneret &

Silzer, 1998), which involve the use of multiple techniques using multiple methods of assessment

and a clinical judgment by the assessor that a person is qualifi ed for some position (Ryan & Sackett,

1987; 1992; 1998). Another venue in which profi les of test scores are considered is in assessment

centers in which candidates for positions (usually managerial) are evaluated in various exercises on

different constructs and assessors make overall judgments that are then used in decision-making.

Overall judgments based on these procedures have shown criterion-related validity [see Ryan &

Sackett (1998) for a summary of data relevant to individual assessment and Gaugler, Rosenthal,

Thornton, & Bentson (1987) or Arthur, Day, McNelly, & Edens (2003) on assessment center valid-

ity], but we are aware of no evidence that validates a profi le or confi gural use of scores.

Perhaps the best description of the research results on the use of profi les to make high-stakes

decisions is that we know very little. The following would be some of the issues that should

receive research attention: (a) Is a profi le of scores actually used, implicitly or explicitly, in com-

bining information about job applicants and what is it? (b) What is the validity of such use and

its incremental validity over the use of individual components of the profi le or linear composites

of the scores in the profi le? and (c) What is the adverse impact on various subgroups using profi le

judgments?

CLINICAL VERSUS STATISTICAL JUDGMENT

Clinical judgment refers to the use and combination of different types of information to make a

decision or recommendation about some person. In psychology, clinical judgment may be most

often discussed in terms of diagnoses regarding clinical patients (Meehl, 1954). These judgments





are likely quite similar to those made in the individual assessments discussed in the previous section

of this chapter but also may occur when judgments are made about job applicants in employment

interviews, assessment centers, and various other instances in which human resource specialists

or psychologists make employment decisions. Clinical judgment is often compared with statisti-

cal judgment in which test scores are combined on the basis of an arithmetic formula that refl ects

the desired weighting of each element of information. The weights may be determined rationally

by a group of job experts or by using weights derived from a regression of a measure of overall

job success on scores on various dimensions using different methods of measurement. Meehl’s

original research (1954) showed that the accuracy of the worst regression estimate was equal to

the judgments made by human decision-makers. A more recent treatment and review of this litera-

ture by Hastie and Dawes (2001) has reaffi rmed the general conclusion that predictions made by

human experts are inferior to those based on a linear regression model. However, human experts are

required to identify the types of information used in the prediction task. The predictions themselves

are likely best left to some mechanical combination rule if one is interested in maximizing a perfor-

mance outcome. The overall clinical judgment when used to make decisions should be the focus of

the validation effort, but unless it is clear how information is combined by the decision-maker, it is

unclear what constructs are playing a role in their decisions. The fact that these clinical judgments

are often not as highly correlated with externally relevant and important outcomes suggests that the

constructs these decision-makers use are not relevant.

In clinical judgment, the presence or absence of adverse impact can be the result of a combina-

tion of information that does not display sizable subgroup differences or a bias on the part of the

person making the judgment. Psychologists making clinical judgments may mentally adjust scores

on the basis of their knowledge of subgroup differences on various measures. There are again no

studies of which we are aware that address the use or appropriateness of such adjustments.

SCIENTIFIC OR LONG-TERM PERSPECTIVE: LIMITATIONS OF EXISTING PRIMARY VALIDATION STUDIES, INCLUDING THE CURRENT META-ANALYTIC DATABASE

There are a great many meta-analyses of the criterion-related validity of various constructs in the

prediction of job performance and many thousands of primary studies. Secondary analyses of meta-

analyses have also been undertaken (e.g., Schmidt & Hunter, 1998). The studies that provided these

data were nearly all conducted more than 30 years ago. Although it is not necessarily the case that

the relationships between ability and performance documented in these studies have changed in

the last half-century or so, this database has some limitations. In this section, we describe these

limitations and make the case that researchers continue their efforts to evaluate test-performance

relationships and improve the quality of the data that are collected.

CONCURRENT VALIDATION DESIGNS

In criterion-related validation research, concurrent validation studies in which predictor and cri-

terion data are simultaneously collected from job incumbents are distinguished from predictive

designs. In the latter, predictor data are collected before hiring from job applicants and criterion

data are collected from those hired presumably on the basis of criteria that are uncorrelated with

the predictor data after some appropriate period of time when job performance is thought to have

stabilized. Defects in the concurrent design (i.e., restriction of range and a different motivational set

on the part of incumbents versus applicants) have been described frequently (Barrett, Phillips, &

Alexander, 1981). Most comparisons of predictive and concurrent designs indicate that they provide

similar estimates of validity. However, it is probably the case that tests more susceptible to motiva-

tional differences between job incumbents and applicants, as might be the case for many noncog-

nitive measures which would display differences in validity when the participants in the research





were actually being evaluated for employment versus a situation in which they were responding “for

research purposes.” To our knowledge, this comparison has not been made frequently, and when it

has been done in meta-analyses cognitive and noncognitive test validities have not been separated

(Schmitt, Gooding, Noe, & Kirsch, 1984). It is the case that practical considerations have made the

use of concurrent designs much more frequent than predictive designs (Schmitt et al., 1984).

Meta-analytic data suggest that there are not large differences in the validity coeffi cients resulting

from these two designs. Further, range restriction corrections can be applied to correct for the fact

that data for lower-scoring persons are absent from concurrent studies, but these data are often

absent in reports of criterion-related research. Nor can we estimate any effects on test scores

that might result from the fact that much more is at stake in a testing situation that may result in

employment as opposed to one that is being done for research purposes. Moreover, as Sussman

and Robertson (1986) maintained, the manner in which some predictive studies are designed and

conducted make them little different than concurrent studies.

UNIDIMENSIONAL CRITERION VERSUS MULTIDIMENSIONAL PERSPECTIVES

Over the last two decades, the view that job performance is multidimensional has become much

more widely accepted by I-O psychologists (Borman & Motowidlo, 1997; Campbell, Gasser, &

Oswald, 1996). Early validation researchers often used a single rating of what is now called task

performance as a criterion, or they combined a set of ratings into an overall performance measure.

In many cases a measure of training success was used as the criterion. The Project A research

showed that performance was comprised of clearly identifi able dimensions (Campbell et al., 1990)

and subsequent research has very often included the use of measures of contextual and task perfor-

mance (Motowidlo, 2003). Some researchers also argue that the nature of what constitutes perfor-

mance has changed because jobs have changed (Ilgen & Pulakos, 1999). In all cases, the underlying

performance constructs should be specifi ed as carefully as possible, perhaps particularly so when

performance includes contextual dimensions, which, as is true of any developing literature, have

included everything that does not include “core” aspects of a job. Validation studies (and meta-

analyses) that include this multidimensional view of performance are very likely to yield informa-

tion that updates earlier validation results.

SMALL SAMPLE SIZES

The limitations of small sample sizes in validity research have become painfully obvious with the

development of meta-analyses and validity generalization research (Schmidt & Hunter, 1977) as

well as the recognition that the power to reject a null hypothesis that there is no test score-perfor-

mance relationship is very low in much early validation work (Schmidt, Hunter, & Urry, 1976).

Although methods to correct for the variability in observed validity coeffi cients are available and

routinely used in meta-analytic and validity generalization research, the use of small samples does

not provide for confi dence in the results of that research and can be misleading in the short term

as enough small sample studies are conducted and reported to discern generalizable fi ndings. This

may not be a problem if we are satisfi ed that the relationships studied in the past are the only ones in

which our fi eld is interested, but it is a problem when we want to evaluate new performance models

(e.g., models that include a distinction between task, contextual dimensions, or others), new predic-

tor constructs (e.g., some noncognitive constructs or even spatial or perceptual measures), or when

we want to assess cross- or multilevel hypotheses.

INADEQUATE DATA REPORTING

The impact of some well-known defi ciencies in primary validation studies is well known. Corrections

for range restriction and criterion unreliability (in the mean and variance of validity coeffi cients)





and for the variability due to small sample size are also well known and routinely applied in validity

generalization work. However, most primary studies do not report information that allows for

sample-based corrections for criterion unreliability or range restriction. Schmidt and Hunter (1977)

in their original meta-analytic effort used estimates of the sample size of the validity coeffi cients

they aggregated because not even sample size was available in early reports. Consequently, in

estimating population validity coeffi cients, meta-analysts have been forced to use assumed artifact

distributions based on the small amount of data that are available. There is some evidence that

these assumptions are approximately correct (e.g., Alexander, Carson, Alliger, & Cronshaw, 1989;

Sackett & Ostgaard, 1994) for range restriction corrections, but the use of such assumed artifact

distributions would not be necessary with adequate reporting of primary data. Unfortunately, such

information for most of our primary database is lost. In addition, researchers disagree regarding

the appropriate operationalization of criterion reliability (Murphy & DeShon, 2000; Schmidt,

Viswesvaran, & Ones, 2000).

CONSIDERATION OF MULTILEVEL ISSUES

As described in the section above on the utility and adverse impact associated with selection proce-

dures, selection researchers have made attempts to estimate the organizational outcomes associated

with the use of valid tests (Boudreau & Ramstad, 2003). Utility is linearly related to validity minus

the cost of recruiting and assessing personnel. When multiplied by the number of people and the

standard deviation of performance in dollar terms, the estimates of utility for most selection instru-

ments are very large (e.g., see Schmidt, Hunter, Outerbridge, & Trattner, 1986).

Another body of research has focused on the relationship between organizational human resource

practices such as the use of tests and measures of organizational success. The organizational-level

research has documented the usefulness of various human resource practices including test use.

Terpstra and Rozell (1993) reported correlational data that supported the conclusion that organiza-

tions that used various selection procedures such as interviews, cognitive ability tests, and biodata

had higher annual levels of profi t, growth in profi t, and overall performance.

Various other authors have called for multilevel (individuals, work groups, organizations) or

cross-level research on the relationship between knowledge, skills, abilities, and other characteristics

(KSAOs) and organizational differences (Schneider, Smith, & Sipe, 2000). Ployhart and Schmitt

(2007) and Schneider et al. (2000) have proposed a series of multilevel questions that include consid-

erations of the relationships between the variance of KSAOs and measures of group and organiza-

tional effectiveness. In the context of the attraction-selection-attrition model (Schneider, 1987), there

are many issues of a multilevel and longitudinal nature that researchers are only beginning to address

and about which we have very little or no data. These questions should be addressed if we are to fully

understand the relationships between KSAOs and individual and organizational performance.

VALIDATION AND LONG-TERM OR SCIENTIFIC PERSPECTIVE

Given the various limitations of our primary database noted in the previous sections of this chapter,

we believe selection researchers should aim to conduct additional large-scale or consortium studies

like Project A (Campbell, 1990; Campbell & Knapp, 2001). These studies should include the fol-

lowing characteristics:

1. They should be predictive (i.e., longitudinal with data collection at multiple points), con-

current, and of suffi cient sample size to allow for adequate power in the tests of hypoth-

eses. Large-scale studies in which organizations continue data collection over time on an

ever-expanding group of participants should be initiated.

2. Multiple criteria should be collected to allow for evaluation of various KASO-performance

relationships.





3. Data should be collected to allow for artifact corrections such as unreliability in the criteria

and range restriction.

4. Unit-level data should be collected to allow for evaluation of multilevel hypotheses. These

data should include basic unit characteristics and outcome data.

5. Demographic data should be collected to allow for evaluation of subgroup differences

in the level of performance and differences in KASO-performance relationships across

subgroups.

6. Data on constructs thought to be related (and unrelated) to the target constructs of interest

should be collected to allow for evaluation of broader construct validity issues.

Obviously, these studies would necessitate a level of cooperation and planning not characteris-

tic of multiple researchers, much less multiple organizations. However, real advancement in our

understanding of individual differences in KSAOs and performance will probably not come from

additional small-scale studies or meta-analyses of primary studies that address traditional questions

with sample sizes, research designs, and measurement characteristics that are not adequate.

CONCLUSIONS

It is certainly true that meta-analyses have provided our discipline with strong evidence that many

of the relationships between individual differences and performance are relatively strong and gener-

alizable. However, many situations where validation is necessary do not lend themselves to validity

generalization or the use of meta-analytic databases. As a result, practitioners frequently fi nd them-

selves in situations where well-designed primary studies are required. A focus on the appropriate

designs for these studies is therefore important.

Additionally, without primary studies of the relationships between individual differences and

performance, there can be no meta-analyses, validity transport, or validity generalization. The qual-

ity and nature of the original studies that are the source of our meta-analytic database determine to

a great extent the currency and quality of the conclusions derived from the meta-analyses, statistical

corrections notwithstanding.

We argue that the fi eld would be greatly served by large-scale primary studies of the type con-

ducted as part of Project A (see Sackett, 1990 or Campbell & Knapp, 2001). These studies should

begin with a clear articulation of the performance and predictor constructs of interest. They should

involve the collection of concurrent and predictive data and improve upon research design and

reporting issues that have bedeviled meta-analytic efforts for the past three decades. Demographic

data should be collected and reported. Data should be collected across multiple organizational units

and organizations (and perhaps globally), and data describing the organizational context should be

collected and recorded. We know much more about the complexities of organizational behavior,

research design, measurement, and individual differences than we did 80–100 years ago, and this

should be refl ected in how we collect our data and make them available to other professionals.

REFERENCES

Ackerman, P. L. (1989). Within-task intercorrelations of skilled performance: Implications for predicting

individual differences? (A comment on Henry & Hulin, 1987). Journal of Applied Psychology, 74,

360–364.

Aguinis, H. (Ed.). (2004). Test score banding in human resource selection: Legal, technical and societal issues.

Westport, CT: Praeger.

Alexander, R. A., Carson, K. P., Alliger, G. M., & Cronshaw, S. F. (1989). Empirical distributions of range

restricted SDx ion validity studies. Journal of Applied Psychology, 74, 253–258.

American Educational Research Association, American Psychological Association, & National Council on

Measurement in Education. (1985). Standards for educational and psychological testing. Washington,

DC: American Psychological Association.





American Educational Research Association, American Psychological Association, & National Council on

Measurement in Education. (1999). Standards for educational and psychological testing. Washington,

DC: American Educational Research Association.

American Psychological Association, American Educational Research Association, and National Council on

Measurement in Education. (1954). Technical recommendations for psychological and diagnostic tech-

niques. Psychological Bulletin, 51, 201–238.

Arthur, W., Jr., Day, E. A., McNelly, T. L., & Edens, P. S. (2003). A meta-analysis of the criterion-related valid-

ity of assessment center dimensions. Personnel Psychology, 56, 125–153.

Arvey, R. D., Nutting, S. M., & Landon, T. E. (1992). Validation strategies for physical ability testing in police

and fi re settings. Public Personnel Management, 21, 301–312.

Austin, J. T., & Villanova, P. (1992). The criterion problem: 1917–1992. Journal of Applied Psychology, 77,

836–874.

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychologi-

cal research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182.

Barrett, G. V., Caldwell, M. S., & Alexander, R. A. (1985). The concept of dynamic criteria: A critical reanalysis.

Personnel Psychology, 38, 41–56.

Barrett, G. V., Phillips, J. S., & Alexander, R. A. (1981). Concurrent and predictive validity designs: A critical

reanalysis. Journal of Applied Psychology, 66, 1–6.

Binning, J. F., & Barrett, G. V. (1989). Validity of personnel decisions: A conceptual analysis of the inferential

and evidential bases. Journal of Applied Psychology, 74, 478–494.

Boehm, V. R. (1982). Are we validating more but publishing less? The impact of governmental regulation on

published validation research—An exploratory investigation. Personnel Psychology, 35, 175–187.

Borman, W. C., & Motowidlo, S. J. (1997). Task performance and contextual performance: The meaning for

personnel selection research. Human Performance, 10, 99–109.

Boudreau, J. W., & Ramstad, P. M. (2003). Strategic industrial and organizational psychology and the role

of utility. In W. C. Borman, D. R. Ilgen, & R. J. Klimoski (Eds.), Handbook of psychology (Vol. 12,

pp. 193–221). Hoboken, NJ: Wiley.

Brogden, H. E. (1951). Increased effi ciency of selection resulting from replacement of a single predictor with

several differential predictors. Educational and Psychological Measurement, 11, 173–195.

Brogden, H. E. (1959). Effi ciency of classifi cation as a function of number of jobs, percent rejected, and the

validity and intercorrelation of job performance estimates. Educational and Psychological Measurement, 19, 181–190.

Brown, C. W., & Ghiselli, E. E. (1953). Percent increase in profi ciency resulting from use of selection devices.

Journal of Applied Psychology, 37, 341–345.

Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait-multimethod

matrix. Psychological Bulletin, 56, 81–105.

Campbell, J. P. (1990). An overview of the Army Selection and Classifi cation Project (Project A). Personnel Psychology, 43, 231–240.

Campbell, J. P., Gasser, M. B., & Oswald, F. L. (1996). The substantive nature of job performance variabil-

ity. In K. R. Murphy (Ed.), Individual differences and behavior in organizations (pp. 258–299). San

Francisco, CA: Jossey-Bass.

Campbell, J. P., & Knapp, D. J. (Eds.) (2001). Exploring the limits in personnel selection and classifi cation. Mahwah, NJ: Lawrence Erlbaum.

Campbell, J. P., McCloy, R. A., Oppler, S. H., & Sager, C. E. (1993). A theory of performance. In N. Schmitt &

W. C. Borman (Eds.), Personnel selection in organizations. San Francisco, CA: Jossey-Bass.

Campbell, J. P., McHenry, J. J., & Wise, L. L. (1990). Modeling job performance in a population of jobs.


Campion, M. A., Outtz, J. L., Zedeck, S., Schmidt, F. L., Kehoe, J. F., Murphy, K. R., & Guion, R. M. (2001).

The controversy over score banding in personnel selection: Answers to 10 key questions. Personnel Psychology, 54, 149–185.

Cascio, W. F., Alexander, R. A., & Barrett, G. V. (1988). Setting cut scores: Legal, psychometric, and profes-

sional issues and guidelines. Personnel Psychology, 41, 1–24.

Cascio, W. F., Valenzi, E. R., & Silbey, V. (1978). Validation and statistical power: Implications for applied

research. Journal of Applied Psychology, 63, 589–595.

Cizek, G. J. (Ed.). (2001). Setting performance standards. Hillsdale, NJ: Lawrence Erlbaum.

Cohen, J. (1988). Statistical power for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.





Cook, T. D., & Campbell, D. T. (1979). Quasi-experimentation: Design and analysis issues for fi eld settings.

Chicago, IL: Rand-McNally.

Cortina, J. M., & Dunlap, W. P. (1997). On the logic and purpose of signifi cance testing. Psychological Methods, 2, 161–172.

Coward, W. M., & Sackett, P. R. (1990). Linearity of ability-performance relationships: A reconfi rmation.


Cropanzano, R., & Wright, T. A. (2003). Procedural justice and organizational staffi ng: A tale of two paradigms.

Human Resource Management Review. Special Issue: Fairness and Human Resources Management, 13,

7–39.

Davis, J. E. (1989). Construct validity in measurement: A pattern matching approach. Evaluation and Program Planning. Special Issue: Concept Mapping for Evaluation and Planning, 12, 31–36.

Dunnette, M. D. (1963). A note on the criterion. Journal of Applied Psychology, 47, 251–254.

Edwards, J. R. (2002). Alternatives to difference scores: Polynomial regression analysis and response surface

methodology. In F. Drasgow & N. Schmitt (Eds.), Measuring and analyzing behavior in organizations.

San Francisco: Jossey-Bass.

Fiske, D. W. (1951). Values, theory, and the criterion problem. Personnel Psychology, 4, 93–98.

Gaugler, B. B., Rosenthal, D. B., Thornton, G. C., & Bentson, C. (1987). Meta-analyses of assessment center

validity. Journal of Applied Psychology, 72, 493–511.

Ghiselli, E. E. (1966). The validity of occupational aptitude tests. New York, NY: Wiley.

Ghiselli, E. E. (1973). The validity of aptitude tests in personnel selection. Personnel Psychology, 26,

461–478.

Gibson, W. M., & Caplinger, J. A. (2007). Transportation of validation results. In S. M. McPhail (Ed.),

Alternative validation strategies: Developing new and leveraging existing validity evidence: The profes-sional practice series. (pp. 29–81). Hoboken, NJ: Wiley.

Goldman, B. M., Gutek, B. A., Stein, J. H., & Lewis, K. (2006). Employment discrimination in organizations:

Antecedents and consequences. Journal of Management, 32(6), 786–830.

Guion, R. M. (1961). Criterion measurement and personnel judgments. Personnel Psychology, 14, 141–149.

Guion, R. M. (1998). Assessment, measurement, and prediction for personnel decisions. Mahwah, NJ: Lawrence

Erlbaum.

Hastie, R., & Dawes, R. M. (2001). Rational choice in an uncertain world. Thousand Oaks, CA: Sage.

Henry, R. A., & Hulin, C. L. (1989). Changing validities: Ability-performance relations and utilities. Journal of Applied Psychology, 54, 365–367.

Hogan, J., & Quigley, A. M. (1986). Physical standards for employment and the courts. American Psychologist, 41, 1193–1217.

Hull, C. L. (1928). Aptitude testing. Yonkers, NY: World Book.

Ilgen, D. R., & Pulakos, E. D. (Eds.). (1999). The changing nature of performance. San Francisco: Jossey-

Bass.

Jeanneret, P. R., & Silzer, R. (Eds.). (1998). Individual psychological assessment: Predicting behavior in orga-nizational settings. San Francisco, CA: Jossey-Bass.

Lance, C. L., Foster, M. R., Gentry, W. A., & Thoresen, J. D. (2004). Assessor cognitive processes in an opera-

tional assessment center. Journal of Applied Psychology, 89, 22–35.

Lance, C. L., Foster, M. R., Nemeth, Y. M., Gentry, W. A., & Drollinger, S. (2007). Extending the nomological

network of assessment center construct validity: Prediction of cross-situationally consistent and specifi c

aspects of assessment center performance. Human Performance, 20, 345–362.

Landy, F. J. (1986). Stamp collecting versus science: Validation as hypothesis testing. American Psychologist, 41, 1183–1192.

Livingston, S. A. (1980). Comments on criterion-referenced testing. Applied Psychological Measurement, 4,

575–581.

Maier, M. H. (1988). On the need for quality control in validation research. Personnel Psychology, 41,

497–502.

McDaniel, M. A. (2007). Validity generalization as a test validation approach. In S. M. McPhail (Ed.),

Alternative validation strategies: Developing new and leveraging existing validity evidence: The profes-sional practice series (pp. 159–180). Hoboken, NJ: Wiley.

Meehl, R. J. (1954). Clinical versus statistical prediction: A theoretical analysis and a review of the evidence.

Minneapolis: University of Minnesota Press.

Messick, S. (1975). The standard problem: Meaning and values in measurement and evaluation. American Psychologist, 30, 955–966.

Messick, S. (1998). Test validity: A matter of consequence. Social Indicators Research, 45, 35–44.





Motowidlo, S. J. (2003). Job performance. In W. C. Borman, D. R. Ilgen, & R. J. Klimoski (Eds.), Handbook of psychology (Vol. 12, pp. 39–53). Hoboken, NJ: Wiley.

Murphy, K. R. (1986). When your top choice turns you down: The effects of rejected offers on the utility of

selection tests. Psychological Bulletin, 99, 133–138.

Murphy, K. R., & DeShon, R. P. (2000). Interrater correlations do not estimate the reliability of job perfor-

mance ratings. Personnel Psychology, 53, 873–900.

Newman, D. A., Jacobs, R. R., & Bartram, D. (2007). Choosing the best method for local validity estima-

tion: Relative accuracy of meta-analysis versus a local study versus Bayes-analysis. Journal of Applied Psychology, 92, 1394–1413.

Ones, D. S., Viswesvaran, C., & Dilchert, S. (2005). Cognitive ability in selection decisions. In O. Wilhelm

(Ed.), Handbook of understanding and measuring intelligence (pp. 431–468). Thousand Oaks, CA:

Sage.

Peterson, N. G., Wise, L. L., Arabian, J., & Hoffman, R. G. (Eds.). (2001). Synthetic validation and valid-

ity generalization: When empirical validation is not possible. In J. P. Campbell & D. J. Knapp (Eds),

Exploring the limits in personnel selection and classifi cation (pp. 411–452). Mahwah, NJ: Lawrence

Erlbaum.

Picano, J. J., Williams, T. J., & Roland, R. R. (Eds.). (2006). Assessment and selection of high-risk operational personnel. New York, NY: Guilford Press.

Ployhart, R. E., & Schmitt, N. (2007). The attraction-selection-attrition model and staffi ng: Some multilevel

implications. In D. B. Smith (Ed.), The people make the place: Exploring dynamic linkages between individuals and organizations (pp. 89–102). Mahwah, NJ: Lawrence Erlbaum.

Ployhart, R. E., Schneider, B., & Schmitt, N. (2006). Staffi ng organizations: Contemporary practice and the-ory. Mahwah, NJ: Lawrence Erlbaum.

Rothstein, H. R. (1992). Meta-analysis and construct validity. Human Performance, 5, 71–80.

Ryan, A. M., & Sackett, P. R. (1987). A survey of individual assessment practices by I/O psychologists.


Ryan, A. M., & Sackett, P. R. (1992). Relationships between graduate training, professional affi liation, and

individual psychological assessment practices for personnel decisions. Personnel Psychology, 45,

363–385.

Ryan, A.M., & Sackett, P. R. (1998). Individual assessment: The research base. In R. Jeanneret & R. Silzer

(Eds.), Individual psychological assessment: Predicting behavior in organizational settings (pp. 54–87).

San Francisco, CA: Jossey-Bass.

Sackett, P. R. (Ed.) (1990). Special Issue: Project A: The U. S. Army Selection and Classifi cation Project.


Sackett, P. R., & Ostgaard, D. J. (1994). Job-specifi c applicant pools and national norms for cognitive ability

tests: Implications for range restriction corrections in validation research. Journal of Applied Psychology, 79, 680–684.

Sackett, P. R., & Roth, L. (1991). A Monte Carlo examination of banding and rank order methods of test score

use in personnel selection. Human Performance, 4, 279–295.

Sackett, P. R., Schmitt, N., Ellingson, J. E., & Kabin, M. B. (2001). High-stakes testing in employment cre-

dentialing, and higher education: Prospects in a post-affi rmative action world. American Psychologist, 56, 302–318.

Sackett, P. R., & Wilk, S. L. (1994). Within-group norming and other forms of score adjustment in preemploy-

ment testing. American Psychologist, 49, 929–954.

Scherbaum, C. A. (2005). Synthetic validity: Past, present, and future. Personnel Psychology, 58, 481–515.

Schmidt, F. L., & Hunter, J. E. (1977). Development of a general solution to the problem of validity generaliza-

tion. Journal of Applied Psychology, 62, 529–540.

Schmidt, F. L., & Hunter, J. E. (1998). The validity and utility of selection methods in personnel psychol-

ogy: Practical and theoretical implications of 85 years of research fi ndings. Psychological Bulletin, 124,

262–274.

Schmidt, F. L., & Hunter, J. E. (2003). History, development, evolution, and impact of validity generalization

and meta-analysis methods, 1975–2001. In K. R. Murphy (Ed.), Validity generalization: A critical review. Applied Psychology Series (pp. 31–65). Mahwah, NJ: Lawrence Erlbaum.

Schmidt, F. L., Hunter, J. E., Outerbridge, A. N., & Trattner, M. H. (1986). The economic impact of job selec-

tion methods on size, productivity, and payroll costs of the federal work force: An empirically based

demonstration. Personnel Psychology, 39, 1–30.

Schmidt, F. L., Hunter, J. E., & Urry, V. W. (1976). Statistical power in criterion-related validity studies. Journal of Applied Psychology, 61, 473–485.





Schmidt, F. L., Mack, M. J., & Hunter, J. E. (1984). Selection utility in the occupation of U.S. park ranger for

three modes of test use. Journal of Applied Psychology, 69, 490–497.

Schmidt, F. L., Viswesvaran, C., & Ones, D. S. (2000). Reliability is not validity and validity is not reliability.


Schmitt, N., Gooding, R., Noe, R. A., & Kirsch, M. (1984). Meta-analyses of validity studies published between

1964 and 1982, and the investigation of study characteristics. Personnel Psychology, 37, 407–422.

Schmitt, N., Rogers, W., Chan, D., Sheppard, L., & Jennings, D. (1997). Adverse impact and predictive

effi ciency of various predictor combinations. Journal of Applied Psychology, 82, 717–730.

Schneider, B. (1987). The people make the place. Personnel Psychology, 40, 437–454.

Schneider, B., Smith, D., & Sipe, W. P. (2000). Personnel selection psychology: Multilevel considerations.

In K. J. Klein & S. W. J. Kozlowski (Eds.), Multilevel theory, research, and methods in organizations (pp. 91–120). San Francisco, CA: Jossey-Bass.

Scott, W. D. (1915). The scientifi c selection of salesmen. Advertising and Selling, 5, 5–7.

Shotland, A., Alliger, G. M., & Sales, T. (1998). Face validity in the context of personnel selection: A multime-

dia approach. International Journal of Selection and Assessment, 6, 124–130.

Society for Industrial and Organizational Psychology. (2003). Principles for the validation and use of person-nel selection procedures. Bowling Green, OH: Author.

Strickland, W. J. (1979). The relationship between program evaluation research and selection system validation:

Application to the assessment center method. Dissertation Abstracts International, 40(1-B), 481–482.

Sussman, M., & Robertson, D. U. (1986). The validity of validity: An analysis of validation study designs.


Terpstra, D. E., & Kethley, R. B. (2002). Organizations’ relative degree of exposure to selection discrimination

litigation. Public Personnel Management, 31, 277–292.

Terpstra, D. E., & Rozell, E. J. (1993). The relationship of staffi ng practices to organizational level measures of

performance. Personnel Psychology, 46, 27–48.

Thayer, P. W. (1992). Construct validation: Do we understand our criteria? Human Performance, 5, 97–108.

Thorndike, E. L. (1911). Individuality. Boston, MA: Houghton-Miffl in.

Viteles, M. S. (1932). Industrial psychology. New York, NY: Norton.




651

30 Perspectives From Twenty-Two Countries on the Legal Environment for Selection

Paul R. Sackett, Winny Shen, Brett Myors, Filip Lievens, Eveline Schollaert, Greet Van Hoye, Steven F. Cronshaw, Betty Onyura, Antonio Mladinic, Viviana Rodríguez, Dirk D. Steiner, Florence Rolland, Heinz Schuler, Andreas Frintrup, Ioannis Nikolaou, Maria Tomprou, S. Subramony, Shabu B. Raj, Shay Tzafrir, Peter Bamberger, Marilena Bertolino, Marco Mariani, Franco Fraccaroli, Tomoki Sekiguchi, Hyuckseung Yang, Neil R. Anderson, Arne Evers, Oleksandr Chernyshenko, Paul Englert, Hennie J. Kriek, Tina Joubert, Jesús F. Salgado, Cornelius J. König, Larissa A. Thommen, Aichia Chuang, Handan Kepir Sinangil, Mahmut Bayazit, Mark Cook, and Herman Aguinis1

In the United States, the legal context plays a major role in how psychologists approach selection

system development. Psychologists know well the set of protected groups, the approaches to making

an a priori case of discrimination (e.g., differential treatment vs. adverse impact), the key court cases

infl uencing selection, and the prohibitions against preferential treatment (e.g., the 1991 ban on score

adjustment or within-group norming). Selection texts (e.g., Guion, 1998) and human resource man-

agement texts (e.g., Cascio & Aguinis, 2008) give prominent treatment to the legal context. In recent

years, there has been a growing internationalization of industrial-organizational (I-O) psychology

such that psychologists from all over the world work with clients in other countries and contribute

to our journals and to our conferences. Test publishers and consulting fi rms establish offi ces all

over the world. As this internationalization continues to increase, it becomes increasingly useful to

take a broader look at the legal environment for selection, examining similarities and differences

in various countries. For example consider a U.S fi rm with operations in several other countries.

Although U. S. fair employment law applies only to those overseas employees who are U.S. citizens,

1 All authors contributed equally to this chapter. Paul R. Sackett and Winny Shen integrated the text materials provided

by each author. Portions of this chapter were drawn from an article by the same set of authors: Myors, B., Lievens, F.,

Schollaert, E., Van Hoye, G., Cronshaw, S. F., Mladinic, A., et al. (2008). International perspectives on the legal environ-

ment for selection. Industrial and Organizational Psychology: Perspectives on Science and Practice, 1, 200–256. Used

by permission of the Society for Industrial and Organizational Psychology and Wiley Blackwell.

TAF-Y100248_FARR-09-0801-C030.indd 651TAF-Y100248_FARR-09-0801-C030.indd 651 1/22/10 10:58:53 AM1/22/10 10:58:53 AM




the employment by U.S. fi rms of host country nationals or third country nationals is subject to the

legal environment of the host country.

DATA COLLECTION METHODOLOGY

To compare and contrast the legal environment for selection in various countries, the senior author

prepared a set of questions about the legal environment for selection, prepared model answers

describing the legal environment in the United States, and contacted psychologists in various coun-

tries, asking them to prepare a document responding to each question and describing the legal

environment in their country. They were also invited to suggest additional project participants in

other countries. Some invitees declined; some initially agreed, but subsequently did not participate.

The goal was to obtain a range of perspectives by sampling about 20 countries, thus, this is by no

means a complete catalog of the legal environment around the world. Researchers and practitioners

who are experts on the topic of selection participated from the following 22 countries: Australia,

Belgium, Canada, Chile, France, Germany, Greece, India, Israel, Italy, Japan, Kenya, Korea, The

Netherlands, New Zealand, South Africa, Spain, Switzerland, Taiwan, Turkey, the United Kingdom,

and the United States. As the list indicates, the countries covered do broadly sample the world.

Because of space constraints, the write-up for each country was subsequently summarized and

organized by issue (e.g., what groups are protected; is preferential treatment of minority group

members permitted) rather than by country to create this chapter. For more context on the legal,

social, and political environment of the countries surveyed, see Myors et al. (2008). Contributing

authors from each of the 22 countries responded to several questions, nine of which are addressed

in turn in this chapter.

Question 1: Are there racial/ethnic/religious subgroups such that some are viewed as “advan-taged” and others as “disadvantaged”?

Table 30.1 identifi es the major groups viewed as “disadvantaged” in each country (note that gen-

der is treated separately in the next section, and specifi c legal protections for disadvantaged groups

are treated under Question 4). As Table 30.1 indicates, the disadvantaged groups differ on several

dimensions. First, the basis for disadvantaged status varies: (a) native/aboriginal people in a setting

where colonizers became the majority group (e.g., Native Americans in the United States, Maori in

New Zealand, First Nations Peoples in Canada), (b) recent immigrants (e.g., many European coun-

tries), (c) racial groups either native to or with long histories in the country (e.g., African Americans

in the United States; Blacks, colored individuals, and Indians in South Africa), (d) religious groups

(e.g., India), and (e) language groups (e.g., Francophones in Canada, Rhaeto-Romanic speakers in

Switzerland). Second, the size of the minority population varies, from a very small percentage of

the population in some countries to the South African extreme of a previously disadvantaged Black

majority. These fi ndings illustrate that there is considerable variability from country to country in

what constitutes a disadvantaged group.

Question 2: What is the general picture regarding women in the workplace (e.g., historical trends regarding employment for women; current data on percentage of women in the work-force; and current status regarding occupational segregation, such as gender representation in various job classes and at various organizational levels)?

Among the countries surveyed, women make up a substantial portion of the workforce. In gen-

eral, women make up from over one quarter to slightly less than one half of the working population

(see Table 30.2). Great strides have been made such that women are being increasingly involved

in the workforce across all countries surveyed, as evidenced by reports of the increased rate of

women’s participation in the workforce, with the exception of Turkey, who reports a slight decline

in the recent years (34% in the early 1990s down to 25.4% in 2004; State Institute of Statistics,

2006). There is substantial variability among countries in terms of the percentage of women who




Perspectives From Twenty-Two Countries on the Legal Environment for Selection 653

TABLE 30.1Disadvantaged Groups Within Each Country

Country Group Percentage of Population

Australia Indigenous Australians 2.5

Belgium Non-Western immigrants

Moroccan

Turkish

0.8

0.4

Canada Immigrants

Visible minorities

First Nations peoples

Francophones

18.4

13.4

2.1

15.7

Chile Recent immigrants

Argentinean

Peruvian

Bolivian

Ecuadorian

1.2

France Immigrant groups

European

North African

Other African

Asian

7.4

3.33

2.22

0.67

0.96

Germany Migrant workers/immigrants

Turkish

Southern European countries

Reimmigrants (Volga-Germans)

3.7

2.8

Greece Immigrants

Albanian

Bulgarian

Georgian

Romanians

7.0

India Within Hindu Castesa

Scheduled castes

Scheduled tribes

Other backward classes

Muslims

15.06

7.51

43.70

13.0

Israel Palestinian Arabs

Druze

Sephardic Jews

Iraq

Iran

Morocco

Ethiopia

22.0

2.0

31.0

Italy Albanian

Rumanian

Moroccan

Ukrainian

Chinese

1.0

0.9

0.9

0.4

Japan North and South Korean

Chinese

Brazilians

Philippines

0.5

0.4

0.2

0.1

continued





TABLE 30.1 (continued)Disadvantaged Groups Within Each Country


Kenya Foreigners

Asians

Europeans

Muslims

Less populous Kenyan tribes

(Swahili, Kalenjin, Kamba,

Kisii, Ameru, Embu, Maasai, Somali,

Turkana, Taita, and Samburu)

1.5

7.0

51.5

Korea Foreigners 0.8

The Netherlands Non-Western immigrants

Turkish

Moroccan

Surinamese

Antillean/Aruban

10.5

2.2

2.0

2.0

0.8

New Zealand Pacifi c peoples

Maori

6.4

13.5

South Africa Black (disadvantaged majority)

African

Colored (mixed race)

Indian

79.5

8.9

2.5

Spain Immigrant groups

Moroccan

Ecuadorian

Rumanian

Colombian

Argetinean

Bolivian

Chinese

Peruvian

9.25

1.16

1.01

0.89

0.59

0.43

0.31

0.22

0.21

Switzerland Immigrant groups

Ex-Yugoslavia

Italians

Portuguese

Germans

Rhaeto-Romanic-speaking

Swiss

21.9

4.7

4.1

2.5

2.4

0.5

Taiwan Taiwanese aborigines 2.0

Turkey Religious minorities

Alevi

Christian and Jewish

Kurdish

Arabic

Other

Armenian

Greek

Jewish

20.0

0.3

11.0

1.5

1.8





participate in the workforce, ranging from approximately one quarter of women in Turkey (State

Institute of Statistics, 2006) to between 60 and 70% in France (Attal-Toubert & Lavergne, 2006),

Kenya (primarily due to the high involvement of women in small-scale farming and pastoralism),

New Zealand, and the United Kingdom. These differences are undoubtedly at least partially due

to the multitude of differences among countries including those in history, culture and values, eco-

nomic conditions, and political conditions. It is interesting to note that in no instance is the female

participation rate higher than the male participation rate; this may partially refl ect the traditional

division of labor between men and women, such that women are more likely to have household and

childcare duties.

Although women are less likely to participate in the workforce than their male counterparts, it

appears that there tend to be no or small differences in the unemployment rate for men and women

(usually within 1 or 2 percentage points). In fact, in recent years in Taiwan, the unemployment

rate for women has been lower than that for men. Exceptions to this trend include Greece (where

the unemployment rate of women is often 2 to 3-fold that of men), Kenya, and Switzerland, where

women are still substantially more likely to be unemployed then male workers. However, it must

be noted that even small changes in the unemployment rate may have strong repercussions for the

economic, political, and social situation of a country.

Among all nations surveyed, there is still gender disparity in pay, and this disparity continues to

be substantial in magnitude. Among all countries where gender disparity information was available,

women earned between 11 and 34% less than men. However, this fi gure may be lower or higher

among countries where we currently do not have the information available. Although it is unclear as

to whether these estimates take into account factors such as differences in occupations, differences

in full- versus part-time work, differences in educational attainment, etc., other research has shown

that even taking into account some of these extraneous factors, women still earn less than their male

counterparts (although the magnitude does decrease slightly). The U.S. General Accounting Offi ce

(2003) reported that women still only earn 80% of what men earn (compared to 75% when not tak-

ing into account differences) in 2000 after taking into account occupation, industry, race, marital

status, and job tenure. Currently, the most positive outlook for women’s earning are in Belgium,

France, Israel, New Zealand, Switzerland, and the United Kingdom, where women earn 80 cents or

more for every dollar earned by men (Equal Opportunities Commission, 2004).

There continues to be occupational segregation to some extent in all 22 countries. Across the

board, women are still more likely to be found in clerical or secretarial, retail or sales, healthcare

TABLE 30.1 (continued)Disadvantaged Groups Within Each Country


United Kingdom Indian

Pakistani

Black Caribbean

Black African

Bangladeshi

Chinese

Other

1.78

1.26

0.95

0.82

0.48

0.41

2.1

United States Black/African American

Hispanic/Hispanic American

Native American and Alaskan Native

12.3

12.5

0.9

a The Hindu caste system differentiates between “forward” (advantaged) and “backward” (disadvantaged) groups. A national

“schedule” or classifi cation of castes differentiates between scheduled castes (previously “untouchable” castes), scheduled

tribal groups, and other backward castes.





TAB

LE 3

0.2

Wom

en’s

Sta

tus

in t

he W

orkp

lace

Wit

hin

Each

Cou

ntry

Cou

ntry

Perc

enta

ge o

f Wor

kfor

ce

Popu

lati

on

Perc

enta

ge o

f Men

Pa

rtic

ipat

ion

in

Wor

kfor

ce

Perc

enta

ge o

f Wom

enPa

rtic

ipat

ion

in

Wor

kfor

ceM

ale

Une

mpl

oym

ent

Rat

eFe

mal

e U

nem

ploy

men

t R

ate

Wag

e D

iffer

enti

ala

Aust

rali

a72.0

57.0

66.0

Bel

giu

m73.6

58.3

7.6

9.6

85.0

Can

ada

64.0

Chil

e38.2

Fra

nce

46.4

74.5

63.8

81.0

Ger

man

y47.0

Gre

ece

64.1

–65.0

38.9

–42.7

5.1

–8.2

13.0

–18.6

India

30.0

Isra

el81.6

Ital

y69.7

45.3

Japan

30.0

b73.2

–79.4

48.0

–50.0

67.1

Ken

ya

29.0

d74.7

72.6

25.0

c38.0

c

Kore

a41.7

74.4

50.0

3.8

3.1

66.2

The

Net

her

lands

70.0

–77.0

54.0

4.5

6.8

New

Zea

land

61.2

81.0

–87.0

Spai

n42.2

3.5

4.8

South

Afr

ica

45.7

Sw

itze

rlan

d72.8

56.9

79.0

–89.0

Tai

wan

67.6

48.1

4.3

3.9

76.9

Turk

ey36.3

72.3

25.4

Unit

ed K

ingdom

78.0

69.0

83.0

Unit

ed S

tate

s46.4

74.0

59.0

4.7

4.5

77.0

The

auth

ors

rep

rese

nti

ng t

he

var

ious

countr

ies

hav

e u

nder

taken

to r

eport

the

most

rec

ent

dat

a av

aila

ble

fro

m t

hei

r co

untr

y;

ther

e m

ay b

e sl

ight

dis

crep

anci

es b

etw

een t

he

yea

rs r

eport

ed f

or

each

countr

y.a

Per

cent

of

wom

en’s

sal

ary c

om

par

ed t

o m

en’s

sal

ary (

men

’s s

alar

y =

100%

).b

P

erce

nt

of

full

-tim

e w

ork

forc

e.c

In u

rban

are

as.

d

Wit

hin

the

moder

n w

age

sect

or.





(e.g., nursing, childcare services), education (e.g., elementary school teachers), public services, or

small-scale agricultural farming occupations (e.g., Kenya and Turkey) than their male counterparts.

Furthermore, the occupations that women are most heavily concentrated in tend to be in the lower

income segment. Women remain underrepresented in business and management positions, particu-

larly higher levels of management. In most countries, women continue to lag behind in representa-

tion for technical and scientifi c positions, professional jobs, higher-level governmental positions

(e.g., judges, cabinet members, etc.), and most higher-level jobs across sectors.

Authors for several countries note that women are more likely to join the workforce as part-time

workers (e.g., Belgium, France, Germany, Japan, Switzerland, and the United Kingdom) to bet-

ter balance work and family demands or leave the workforce because of childcare demands (e.g.,

Japan, Korea, and the United States). The latter trend is particularly pronounced in Japan, where

the participation ratio by age groups shows an M-shaped curve, because labor force participation

rate declines in women’s early 30s because of childcare responsibilities. During the period of 1970–

2004, the valley section of this M-curve has shifted northeastward due in part to the trend of late

marriage and late childbirth. In addition, both peaks of this M-curve have become higher, indicating

that women’s workforce participation has substantially increased in their 20s and late-30s or older

(Japan Institute of Labor Policy and Training, 2007). However, some countries also indicated that

the wage gap between men and women may be even more pronounced among part-time workers.

For example, in the United Kingdom, women are paid 17% less then men in full-time work and 38%

less in part-time work (Equal Opportunities Commission, 2004).

Question 3: Is there research documenting mean differences between groups identifi ed above on individual difference measures relevant to job performance?

Mean differences on ability and personality measures are commonly examined in the United

States, with enough data for large-scale meta-analytic summaries. Mean differences on tests of

developed abilities of roughly 1.00 standard deviation (SD) between Whites and African Americans

and roughly 0.67 SD between Whites and Hispanics have been consistently reported. The largest-

scale summary of this literature is a meta-analysis by Roth, Bevier, Bobko, Switzer, and Tyler

(2001). This abundance of data proves to be in marked contrast to the pattern of fi ndings in the

countries examined here. In fact, for most countries, the authors reported fi nding either no research

or research with samples so small that they refrained from drawing conclusions (i.e., Chile, France,

Greece, Italy, Japan, Korea, Spain, Switzerland, Turkey, and the United Kingdom). Although lim-

ited, there are some data on group differences in some countries.

Two countries (Australia and Taiwan) report research on cognitive ability differences between

aborigines and the advantaged group. The lower cognitive ability scores for Australian aborigi-

nes may refl ect differences in language and culture. Aborigines in Taiwan, who typically have

lower educational attainment (Council of Indigenous Peoples, 2002), also score lower than nona-

borigines on several cognitive ability tests. Data from the United Arrangement Commission for

college entrance examinations in Taiwan in 2006 showed d values between 0.44 and 0.68 in favor

of nonaborigines, depending on the particular test subject (A. Chuang, personal communication,

May 1, 2007).

Cognitive ability mean score differences have been reported of d = 1.39 between Turkish/

Moroccan immigrants and Dutch test-takers and d = 1.08 between Surinamese/Antillean and Dutch

test-takers, in both cases favoring the majority group (te Nijenhuis, de Jong, Evers, & van der Flier,

2004). Language differences appear to contribute to these fi ndings because higher scores are found

for second-generation than fi rst-generation immigrants. Studies in Belgium also report mean differ-

ences of about 1.00 SD on cognitive tests between Belgians and Turkish and Moroccan immigrants

in samples of children (Fontaine, Schittekatte, Groenvynck, & De Clercq, 2006).

In South Africa, mean score differences on cognitive tests between Black and White groups are

normally larger than U.S. studies and have d values of approximately 1.00–1.50, with Whites obtain-

ing the higher mean scores. In a study performed in a South African fi nancial services organization,





d values of 0.99 for averbal ability, 1.03 for a numerical ability, and 1.14 for a diagrammatic ability

test were found (SHL, 2006). In South Africa, these differences are largely ascribed to the differ-

ences in the educational level of the racial groups. In the 2001 census, it was determined that 22.3%

of Africans, 8.3% of Colored (mixed race), 5.3% of Indians, and 1.4% of Whites had no schooling

(Statistics South Africa, 2001).

Limited data report lower scores for Arabs than Jews in Israel (Zeidner, 1986), for Canadian

First Nations people than for Whites, for New Zealand Maori than for Whites (Chernyshenko,

2005; Guenole, Englert, & Taylor, 2003), and differences between individuals in various provinces

in Kenya (Kinyungu, 2006). Data on personality measures are even more limited than for cogni-

tive ability, with authors reporting personality data from only two countries: a large-scale study of

Black-White differences in South Africa (Kriek, 2006) showing small differences and several stud-

ies of Dutch-immigrant differences in the Netherlands showing much larger differences (van Leest,

1997; te Nijenhuis, van der Flier, & van Leeuwen, 1997, 2003).

Overall, several fi ndings of interest emerge. First, it is clear that gathering data and reporting

mean differences by group is generally far more common in the United States than in virtually all

of the countries contributing to this report. This is likely the result of the legal scrutiny to which

tests are held in the United States. The Uniform Guidelines on Employee Selection Procedures

(U.S. Equal Employment Opportunity Commission, 1978) use adverse impact computations as the

basis for a prima facie case of discrimination, and thus, adverse impact resulting from test use is

routinely examined, with mean differences between groups and the method of test use (e.g., a high

or a low cutoff) functioning as key determinants of adverse impact. Second, although data tend to

be more sparse than in the United States, group differences are studied and observed in various

settings involving different types of disadvantaged groups (e.g., immigrant groups in Belgium and

The Netherlands; native peoples in Australia, New Zealand, and Canada; tribal and provincial dif-

ferences in Kenya; the native Black population in South Africa; and Arab groups in Israel). Third,

as in the United States, there is interest not only in whether there are group differences, but also

in understanding the basis for these differences. Language, culture, and differences in educational

access and attainment are seen as key concerns in understanding differences in test scores across

groups.

In the United States, disparate impact is the basis for a prima facie case of discrimination. The

implicit assumption is that various groups are expected to obtain similar mean scores absent bias in

the measure. Our data suggest that many European countries target certain groups as immigrants

to meet specifi c labor shortages. Thus, immigrants might have higher or lower abilities, depending

whether a country tried to attract highly skilled people (e.g., recent immigrants into Switzerland

from northern and western Europe) or tried to attract people with low skills (e.g., Turkish immi-

grants to Germany). In other words, even if one has a general expectation of no group differences

at the population level, a fi nding of differences between locals and immigrants would be expected

given this targeted immigration.

Question 4: Are there laws prohibiting discrimination against specifi c groups and/or man-dating fair treatment of such groups? Which groups are protected? Which employers are covered? Which employment practices are covered (e.g., selection, promotion, dismissal)?

Table 30.3 presents summary information addressing the above questions for each country.

Several fi ndings emerge. First, there is some basis for legal protections for members of specifi ed

groups in all countries. The bases for these protections vary widely. In many cases the national

constitution provides general, or at times specifi c, protections. This may be seen as analogous to the

5th and 14th Amendments to the U.S. Constitution, which respectively state that “no person shall …

be deprived of life, liberty, or property without due process of law,” and that “no state shall …

deny to any person within its protection the equal protection of the laws.” However, in virtually all

cases there are also specifi c laws defi ning specifi ed protected classes, specifying which employment

practices are covered and which employers are required to comply. The intent here is to identify the





TAB

LE 3

0.3

Inte

rnat

iona

l Law

s an

d Pr

acti

ces

Cou

ntry

Law

Empl

oyer

s C

over

edEm

ploy

men

t Pr

acti

ces

Cov

ered

Aust

rali

aT

he

Cri

mes

Act

1914

Rac

ial

Dis

crim

inat

ion A

ct 1

975

Sex

Dis

crim

inat

ion 1

984

Hum

an R

ights

and E

qual

Opport

unit

y C

om

mis

sion

Act

1986

Dis

abil

ity D

iscr

imin

atio

n A

ct 1

992

Work

pla

ce R

elat

ions

Act

1996

Equal

Opport

unit

y f

or

Wom

en i

n t

he

Work

pla

ce A

ct

1999

Age

Dis

crim

inat

ion A

ct 2

004

All

em

plo

yer

s; E

OW

W o

f 1999 r

efer

s to

org

aniz

atio

ns

of

100+

All

sta

ges

of

the

emplo

ym

ent

rela

tionsh

ip i

ncl

udin

g b

ut

not

lim

ited

to r

ecru

itm

ent,

sel

ecti

on, te

rmin

atio

n, tr

ainin

g, an

d

pro

moti

on.

Bel

giu

mB

elgia

n C

onst

ituti

on o

f 1994 A

rtic

le 1

0, 11, 191

Law

Equal

ity o

f M

en-W

om

en o

f 1978

Anti

dis

crim

inat

ion l

aw o

f 2003

All

em

plo

yer

sM

ost

em

plo

ym

ent

pra

ctic

es i

ncl

udin

g s

elec

tion a

nd

appoin

tmen

t, p

rom

oti

ons,

em

plo

ym

ent

opport

unit

ies,

labor

condit

ions,

dis

mis

sal,

and w

ages

.

Can

ada

Can

adia

n H

um

an R

ights

Code

of

1985

Sec

tion 1

5 o

f th

e C

har

ter

of

Rig

hts

and F

reed

om

s

(1982)

Fed

eral

Em

plo

ym

ent

Equit

y A

ct (

2004)

Fed

eral

Contr

acto

rs P

rogra

m

Pay

equit

y l

egis

lati

on (

feder

al a

nd s

om

e pro

vin

ces)

Fed

eral

gover

nm

ent

dep

artm

ents

, cr

ow

n

corp

ora

tions,

and o

ther

fed

eral

ly r

egula

ted

agen

cies

and o

rgan

izat

ions

Most

em

plo

ym

ent

pra

ctic

es i

ncl

udin

g s

elec

tion,

per

form

ance

appra

isal

, te

rmin

atio

n, an

d c

om

pen

sati

on.

Chil

eC

onst

ituti

on, C

hap

ter

3 (

Rig

hts

and D

uti

es),

art

icle

19

Nº

16 (

Fre

edom

of

Work

and i

ts p

rote

ctio

n)

and

Work

Code,

Art

icle

2º

(2002)

All

em

plo

yer

sT

he

Const

ituti

on e

stab

lish

es t

he

gen

eral

nondis

crim

inat

ion

pri

nci

ple

on t

he

bas

is o

f ra

ce, co

lor,

sex

, ag

e, m

arit

al

stat

us,

unio

n m

ember

ship

sta

tus,

rel

igio

n, poli

tica

l

opin

ions,

nat

ional

ity,

and n

atio

nal

or

soci

al o

rigin

. In

Mar

ch 2

008, a

new

law

wen

t in

to t

ake

effe

ct (

law

#

20,0

87).

This

new

law

defi

nes

dis

crim

inat

ion a

s an

y a

ctio

n

that

is

agai

nst

the

equal

opport

unit

y f

or

all

work

ers.

A

new

reg

ula

tion w

ill

spec

ify t

he

pra

ctic

es t

hat

are

cover

ed

by t

he

law

.

cont

inue

d





TAB

LE 3

0.3

(con

tinu

ed)

Inte

rnat

iona

l Law

s an

d Pr

acti

ces

Cou

ntry

Law

Empl

oyer

s C

over

edEm

ploy

men

t Pr

acti

ces

Cov

ered

Fra

nce

Fre

nch

Const

ituti

on o

f 1958

Inte

rnat

ional

conven

tion o

f th

e U

nit

ed

Nat

ions

(1965)

rati

fi ed

in 1

971

Inte

rnat

ional

conven

tion o

f th

e In

tern

atio

nal

Lab

or

Org

aniz

atio

n (

1958)

rati

fi ed

in 1

981

“The

law

conce

rnin

g t

he

fi ght

agai

nst

raci

sm”

of

1972

“The

law

conce

rnin

g w

ork

er’s

lib

erti

es i

n

org

aniz

atio

ns”

of

1982

Tre

aty o

f A

mst

erdam

of

1997

L. 122-4

5 f

rom

Lab

or

Law

225-1

and 2

25-2

fro

m t

he

Pen

al C

ode

All

em

plo

yer

sM

any e

mplo

ym

ent

pra

ctic

es i

ncl

udin

g s

elec

tion, ac

cess

to

trai

nin

g, pay

, la

yoff

s, t

ransf

ers,

and j

ob c

lass

ifi c

atio

n.

Ger

man

yA

llgem

eines

Gle

ichbeh

andlu

ngsg

eset

z: G

ener

al E

qual

Opport

unit

y L

aw

All

em

plo

yer

s, e

xce

pt

tenden

cy o

rgan

izat

ions

(e.g

. re

ligio

us

org

aniz

atio

ns)

All

sta

ges

of

the

emplo

ym

ent

rela

tionsh

ip i

ncl

udin

g p

laci

ng

a jo

b a

d, hir

ing a

nd s

elec

tion, defi

nit

ion o

f pay

men

t,

per

form

ance

appra

isal

and p

rom

oti

on, jo

b-r

elat

ed t

rain

ing

and j

ob c

ounse

ling, co

rpora

te h

ealt

h s

ervic

es, des

ign o

f

work

ing c

ondit

ions,

soci

al s

ervic

es, an

d d

ism

issa

l.

Gre

ece

Gre

ek L

aw 3

304 o

f 2005, eq

ual

tre

atm

ent

Gre

ek L

aw 3

488 o

f 2006, on e

qual

tre

atm

ent

bet

wee

n

peo

ple

in t

he

labor

mar

ket

All

em

plo

yer

sC

ondit

ions

for

acce

ss to e

mplo

ym

ent, to s

elf-

emplo

ym

ent, o

r to

occ

upat

ion, in

cludin

g s

elec

tion c

rite

ria

and r

ecru

itm

ent

condit

ions;

pro

moti

on; ac

cess

to a

ll types

and to a

ll lev

els

of

voca

tional

guid

ance

, voca

tional

tra

inin

g, ad

van

ced v

oca

tional

trai

nin

g a

nd r

etra

inin

g, in

cludin

g p

ract

ical

work

exper

ience

,

emplo

ym

ent an

d w

ork

ing c

ondit

ions;

dis

mis

sals

, pay

,

mem

ber

ship

, an

d inv

olv

emen

t in

an o

rgan

izat

ion o

f w

ork

ers

or

emplo

yer

s, o

r an

y o

rgan

izat

ion w

hose

mem

ber

s ca

rry o

n a

par

ticu

lar

pro

fess

ion, in

cludin

g the

ben

efi t

s pro

vid

ed f

or

by

such

org

aniz

atio

ns;

soci

al p

rote

ctio

n, in

cludin

g s

oci

al

insu

rance

and s

anit

ary r

elie

f; s

oci

al p

rovis

ions;

educa

tion; an

d

acce

ss to d

isposa

l an

d to p

rovis

ion o

f ben

efi t

s, w

hic

h a

re

pro

vid

ed to the

publi

c, incl

udin

g h

ousi

ng.





India

India

n C

onst

ituti

on

Art

icle

15. P

rohib

itio

n o

f dis

crim

inat

ion o

n g

rounds

of

reli

gio

n, ra

ce, ca

ste,

sex

, or

pla

ce o

f bir

th

Art

icle

16. E

qual

ity o

f opport

unit

y i

n m

atte

rs o

f

publi

c em

plo

ym

ent

Art

icle

39

Art

icle

46

Art

icle

335

Gover

nm

ent

enti

ties

, publi

c se

ctor

org

aniz

atio

ns,

and o

rgan

izat

ions

rece

ivin

g g

over

nm

ent

fundin

g

Sel

ecti

on;

pre

vio

usl

y p

rom

oti

on.

Isra

elB

asic

Law

on H

um

an D

ignit

y a

nd L

iber

ty

Bas

ic L

aw o

n t

he

Fre

edom

of

Occ

upat

ion

Wom

en’s

Equal

Rig

hts

Law

of

1951

Equal

Pay

Law

of

1996

Equal

Em

plo

ym

ent

Opport

unit

y o

f 1988

All

em

plo

yer

s

All

em

plo

yer

s 6+

Com

pen

sati

on, st

affi

ng, co

ndit

ions

of

emplo

ym

ent,

pro

moti

on, tr

ainin

g a

nd d

evel

opm

ent,

dis

mis

sal,

sev

eran

ce

pay

, re

tire

men

t ben

efi t

s.

Ital

yIt

alia

n C

onst

ituti

on o

f 1948

Art

icle

3

Leg

isla

tive

dec

ree

216 o

f 2003

All

em

plo

yer

sR

ecru

itm

ent,

sel

ecti

on, pro

moti

on, em

plo

ym

ent

agen

cies

,

outp

lace

men

t pro

cedure

s, t

rain

ing, w

ork

ing c

ondit

ions.

Japan

Lab

our

Sta

ndar

ds

Law

of

1947

Law

on S

ecuri

ng E

qual

Opport

unit

y a

nd T

reat

men

t

bet

wee

n M

en a

nd W

om

en i

n E

mplo

ym

ent

of

1972

Law

for

Em

plo

ym

ent

Pro

moti

on

, et

c. o

f th

e D

isab

led

of

1960

Law

Conce

rnin

g S

tabil

izat

ion o

f E

mplo

ym

ent

of

Old

er P

erso

ns

of

1971

All

em

plo

yer

s

All

em

plo

yer

s

All

em

plo

yer

s

All

em

plo

yer

s

Wag

es, w

ork

ing h

ours

, oth

er w

ork

ing c

ondit

ions.

Rec

ruit

men

t an

d h

irin

g, as

signm

ent,

pro

moti

on, dem

oti

on,

trai

nin

g, fr

inge

ben

efi t

s, c

han

ge

in j

ob t

ype

and

emplo

ym

ent

stat

us,

enco

ura

gem

ent

of

reti

rem

ent,

man

dat

ory

ret

irem

ent

age,

dis

mis

sal

and r

enew

al o

f

emplo

ym

ent

contr

act.

Rec

ruit

men

t an

d h

irin

g.

Man

dat

ory

ret

irem

ent.

Ken

ya

Ken

yan

Const

ituti

on C

hap

ter

5,

Sec

tion 8

2

HIV

and A

IDS

Pre

ven

tion a

nd C

ontr

ol A

ct 1

4

The

Per

sons

wit

h D

isab

ilit

ies

Act

14 o

f 2003

All

em

plo

ym

ent

pra

ctic

es.

cont

inue

d





TAB

LE 3

0.3

(con

tinu

ed)

Inte

rnat

iona

l Law

s an

d Pr

acti

ces

Cou

ntry

Law

Empl

oyer

s C

over

edEm

ploy

men

t Pr

acti

ces

Cov

ered

Kore

aN

atio

nal

Hum

an R

ights

Com

mis

sion A

ct o

f 2001

Equal

Em

plo

ym

ent A

ct o

f 1987

The

Act

of

Em

plo

ym

ent

Pro

moti

on a

nd V

oca

tional

Reh

abil

itat

ion f

or

the

Dis

able

d o

f 1990

The

Aged

Em

plo

ym

ent

Pro

moti

on A

ct o

f 1991

The

Bas

ic E

mplo

ym

ent

Poli

cy A

ct o

f 1993

Not

spec

ifi e

d

All

em

plo

yer

s

Em

plo

yer

s of

500+

work

ers

for

affi

rmat

ive

acti

on

clau

se

Em

plo

yer

s w

ith 5

0+

work

ers

Gover

nm

ent

emplo

yee

s

Em

plo

yer

s w

ith 3

00+

em

plo

yer

s

Not

spec

ifi e

d

Rec

ruit

men

t, h

irin

g, tr

ainin

g, pla

cem

ent,

pro

moti

on,

com

pen

sati

on, lo

ans,

man

dat

ory

ret

irem

ent

age,

reti

rem

ent,

and d

ism

issa

l.

Rec

ruit

men

t, s

elec

tion, co

mpen

sati

on, ed

uca

tion, tr

ainin

g,

job p

lace

men

t, p

rom

oti

ons,

set

ting a

man

dat

ory

ret

irem

ent

age,

ret

irem

ent,

and d

ism

issa

l.

Hir

ing, pro

moti

on, tr

ansf

er, ed

uca

tion, an

d t

rain

ing.

Rec

ruit

men

t, h

irin

g, an

d d

ism

issa

l.

Rec

ruit

men

t an

d h

irin

g.

The

Net

her

lands

Const

ituti

on, A

rtic

le 1

of

2003

Gen

eral

Law

Equal

Tre

atm

ent

of

1994

All

em

plo

yer

s (e

xce

pt

reli

gio

us,

phil

oso

phic

al,

or

poli

tica

l org

aniz

atio

ns)

Rec

ruit

men

t, s

elec

tion, em

plo

ym

ent

agen

cies

, dis

mis

sal,

labor

agre

emen

ts, ed

uca

tion b

efore

and d

uri

ng

emplo

ym

ent,

pro

moti

on, an

d w

ork

ing c

ondit

ions.

New

Zea

land

Hum

an R

ights

Act

of

1993

All

em

plo

yer

sR

efusa

l of

emplo

ym

ent,

les

s fa

vora

ble

em

plo

ym

ent,

condit

ions

of

work

, su

per

annuat

ion, fr

inge

ben

efi t

s,

trai

nin

g, pro

moti

on, tr

ansf

er, te

rmin

atio

n, re

tire

men

t,

and r

esig

nat

ion.

South

Afr

ica

Const

ituti

on o

f th

e R

epubli

c of

South

Afr

ica

of

1996

Lab

our

Rel

atio

ns

Act

, Act

66, of

1995

Em

plo

ym

ent

Equit

y A

ct, N

o. 55

, of

1998

All

em

plo

yer

s ex

cept

the

Nat

ional

Def

ense

Forc

e, N

atio

nal

Inte

llig

ence

Agen

cy, an

d S

outh

Afr

ican

Sec

ret

Ser

vic

e

Inclu

des,

bu

t is

no

t li

mit

ed

to

, re

cru

itm

en

t p

roced

ure

s,

ad

vert

isin

g,

sele

cti

on

cri

teri

a,

ap

po

intm

en

t an

d

ap

po

intm

en

t p

rocess

, jo

b c

lass

ifi c

ati

on

an

d g

rad

ing

,

rem

un

era

tio

n,

em

plo

ym

en

t b

en

efi

ts,

term

s an

d

co

nd

itio

ns

of

em

plo

ym

en

t, j

ob

ass

ign

men

ts,

wo

rkin

g

env

iro

nm

en

t an

d f

acil

itie

s, t

rain

ing

an

d d

evelo

pm

en

t,

perf

orm

an

ce e

valu

ati

on

sy

stem

s, p

rom

oti

on

, tr

an

sfer,

dem

oti

on

, d

iscip

lin

ary

measu

re o

ther

than

dis

mis

sal,

an

d d

ism

issa

l.





Spai

nS

pan

ish C

onst

ituti

on, A

rtic

le 1

4 o

f 1978

Law

of W

ork

er’s

Sta

tute

of

1980, 2

005, A

rtic

le 4

.2 y

17

Org

anic

Law

for

Eff

ecti

ve

Equal

ity b

etw

een W

om

en

and M

en o

f 2007, A

rtic

le 1

, 3, 4, 5, 6

Law

of

Bas

ic S

tatu

te o

f P

ubli

c E

mplo

yee

of

2005,

Art

icle

14.i

All

em

plo

yer

sR

ecru

itm

ent,

sel

ecti

on, pro

moti

on, co

mpen

sati

on, tr

ainin

g,

tem

pora

l em

plo

ym

ent

com

pan

ies,

em

plo

ym

ent

agen

cies

,

dis

mis

sal,

lab

or

agre

emen

ts, co

llec

tive

bar

gai

nin

g,

educa

tion b

efore

and d

uri

ng e

mplo

ym

ent,

hea

lth

pro

gra

ms,

and w

ork

ing c

ondit

ions.

Sw

itze

rlan

dB

undes

ver

fass

ung o

f 1999 (

Sw

iss

Fed

eral

Const

ituti

on)

Bundes

ges

etz

über

die

Bes

eiti

gung v

on

Ben

achte

ilig

ungen

von M

ensc

hen

mit

Beh

inder

ungen

of

2002 (

Fed

eral

Law

for

the

Equal

Tre

atm

ent

of

Peo

ple

wit

h D

isab

ilit

ies)

Bundes

ges

etz

über

die

Gle

ichst

ellu

ng v

on M

ann u

nd

Fra

u o

f 1995 (

Fed

eral

Law

for

the

Equal

Tre

atm

ent

of

Men

and W

om

en)

Sch

wei

zeri

sches

Ziv

ilges

etzb

uch

of

1907 (

Sw

iss

Civ

il

Code)

Bundes

ges

etz

bet

reff

end d

ie E

rgän

zung d

es

Sch

wei

zeri

schen

Ziv

ilges

etzb

uch

es —

Obli

gat

ionen

rech

t of

1912 (

Sw

iss

Code

of

Obli

gat

ions)

Publi

c em

plo

yer

s

All

em

plo

yer

s

All

em

plo

yer

s

Incl

udes

pre

- (p

arti

cula

rly),

duri

ng, an

d p

ost

emplo

ym

ent

pra

ctic

es.

Incl

udes

pre

-, d

uri

ng, an

d p

ost

emplo

ym

ent

pra

ctic

es (

i.e.

,

recr

uit

men

t, s

exual

har

assm

ent,

ear

nin

gs,

pro

moti

ons,

etc

.).

Pro

tect

ion o

f em

plo

yee

per

sonal

ity a

nd p

erso

nal

dat

a

thro

ughout

all

stag

es o

f th

e em

plo

ym

ent

pro

cess

.

Tai

wan

Art

icle

5 o

f th

e E

mplo

ym

ent

Ser

vic

es A

ct o

f 1992

Gen

der

Equal

ity i

n E

mplo

ym

ent

Law

of

2002

Equal

Em

plo

ym

ent

Opport

unit

y f

or

Abori

gin

es A

ct

of

2001

All

em

plo

yer

s

All

em

plo

yer

s

Publi

c em

plo

yer

s an

d p

rivat

e em

plo

yer

s w

ho

are

gover

nm

ent

contr

acto

rs w

ith d

om

esti

c

emplo

yee

of

100+

Sta

ffi n

g.

Rec

ruitm

ent, s

elec

tion, p

rom

otion, j

ob a

lloca

tion, p

erfo

rman

ce

eval

uat

ion, p

rom

otion, t

rain

ing, c

om

pen

sation, b

enefi

ts,

retire

men

t, d

ism

issa

l, a

nd q

uit.

Sta

ffi n

g.

cont

inue

d





TAB

LE 3

0.3

(con

tinu

ed)

Inte

rnat

iona

l Law

s an

d Pr

acti

ces

Cou

ntry

Law

Empl

oyer

s C

over

edEm

ploy

men

t Pr

acti

ces

Cov

ered

Turk

eyR

epubli

c of

Turk

ey C

onst

ituti

on

of

1982

Art

icle

10

Art

icle

49

Art

icle

50

Art

icle

70

Lab

or

Law

, Art

icle

5 o

f 2003

UN

’s C

onven

tion o

n t

he

Eli

min

atio

n o

f A

ll S

ort

s of

Dis

crim

inat

ion A

gai

nst

Wom

en A

rtic

le 1

1

Pri

me

Min

iste

r’s

offi

ce

circ

ula

r of

2004

All

em

plo

yer

s

All

em

plo

yer

s (e

xce

pt

sea

tran

sport

atio

n, ai

r

tran

sport

, ag

ricu

ltura

l an

d f

ore

stry

wit

h l

ess

than

50 e

mplo

yee

s, h

om

e se

rvic

es, in

tern

ship

s,

pro

fess

ional

ath

lete

s, r

ehab

ilit

atio

n w

ork

ers,

busi

nes

ses

wit

h 3

work

ers,

han

dm

ade

art

jobs

done

at h

om

e, j

ourn

alis

ts)

All

em

plo

yer

s

Publi

c em

plo

yer

s

Art

icle

70 s

pec

ifi c

ally

cover

s se

lect

ion f

or

publi

c

inst

ituti

ons;

oth

er p

ract

ices

are

im

pli

citl

y c

over

ed

incl

udin

g p

ay, pro

moti

on, an

d d

ism

issa

l in

oth

er a

rtic

les.

Per

form

ance

appra

isal

, pay

, pro

moti

on, an

d t

erm

inat

ion

pra

ctic

es a

re i

mpli

citl

y c

over

ed;

sele

ctio

n i

s not

cover

ed

bec

ause

the

law

only

cover

s pri

vat

e se

ctor

emplo

yee

s w

ho

are

alre

ady e

mplo

yed

.

Gen

eral

ly a

ll e

mplo

ym

ent

pra

ctic

es. in

cludin

g s

elec

tion,

pro

moti

on, te

rmin

atio

n, pay

, per

form

ance

appra

isal

,

acce

ss t

o t

rain

ing, an

d t

reat

men

t.

Sel

ecti

on.

Unit

ed K

ingdom

Rac

e R

elat

ions

Act

of

1976

Sex

Dis

crim

inat

ion A

ct o

f 1975

Em

plo

ym

ent

Equal

ity (

Age)

Reg

ula

tions

2006

Equal

Pay

Act

of

1970

Dis

abil

ity D

iscr

imin

atio

n A

ct 1

995

Euro

pea

n C

om

munit

y D

irec

tives

All

em

plo

yer

s, t

rade

unio

ns,

pro

fess

ional

bodie

s,

and e

mplo

ym

ent

agen

cies

All

em

plo

yer

s, t

rade

unio

ns,

pro

fess

ional

bodie

s,

and e

mplo

ym

ent

agen

cies

All

ages

, young a

nd o

ld

Gen

eral

ly a

ll e

mplo

ym

ent

pra

ctic

es:

sele

ctio

n, pro

moti

on,

term

inat

ion, pay

, per

form

ance

appra

isal

, ac

cess

to t

rain

ing,

and t

reat

men

t.

Unit

ed S

tate

sC

ivil

Rig

hts

Act

of

1964, T

itle

VII

(am

ended

1972,

1991)

Age

Dis

crim

inat

ion A

ct 1

967

Am

eric

ans

wit

h D

isab

ilit

ies

Act

1990 a

nd

Reh

abil

itat

ion A

ct 1

973

Equal

Pay

Act

1963

All

publi

c em

plo

yer

s an

d p

rivat

e em

plo

yer

s w

ith

15+

em

plo

yee

s

Pri

vat

e em

plo

yer

s w

ith 2

0+

em

plo

yee

s, s

tate

and

loca

l gover

nm

ents

AD

A c

over

s pri

vat

e em

plo

yer

s, s

tate

and l

oca

l

gover

nm

ents

; R

ehab

ilit

atio

n A

ct c

over

s fe

der

al

gover

nm

ent;

Vir

tual

ly a

ll e

mplo

yer

s

Ran

ge

of

emplo

ym

ent

dec

isio

ns

incl

udin

g h

irin

g, co

m-

pen

sati

on, te

rms,

condit

ions,

and p

rivil

eges

of

emplo

ym

ent.

Pro

hib

its

dis

crim

inat

ion a

gai

nst

indiv

idual

s ag

e 40 o

r old

er.

Pro

hib

its

dis

crim

inat

ion a

gai

nst

indiv

idual

s w

ith d

isab

ilit

ies

in t

he

full

ran

ge

of

emplo

ym

ent

dec

isio

ns.

Pro

hib

its

dis

crim

inat

ion a

gai

nst

wom

en i

n p

ay d

ecis

ions.





major contemporary federal laws and government decrees, and as such it is not a complete record

of all historical employment regulations. For example, in the United States a specialist can rightly

note that the Civil Rights Acts of 1866 and 1871 are still relied upon on occasion, although these are

not listed in the table. Also, several states and cities have additional statutes, offering protection to

groups beyond those covered by federal law.

Second, the protections offered are generally quite sweeping in terms of the types of employers

covered. In most cases all employers are covered. Some laws are restricted to government employ-

ees, and in some cases, coverage is restricted to larger employers, with the coverage threshold

varying quite widely for some statutes (e.g., more than 6 employees in Israel, 15 in the U.S., 100 in

Taiwan, and 300 in Korea).

Third, it is typical for a broad range of employment practices to be included. Employee selection

is specifi cally included in all countries except Chile, which has the least developed set of employ-

ment rights regulations of the countries examined here, and which has yet to specify a set of cov-

ered em ployment practices. However, Chile does prohibit discrimination based on race, color, sex,

age, marital status, union membership, status, religion, political opinions, nationality, and national

or social origin in its Constitution but does not specify which specifi c employment practices are

covered.

Fourth, there is considerable commonality and variation in the classes that receive protection in

each country. Table 30.4 identifi es the most common protected classes and indicates whether those

classes are covered in each of the contributing countries. The classes covered in U.S. Civil Rights

law emerge as widely commonly covered across countries: race, color, religion, gender, national ori-

gin, age, and disability status. Three categories not protected by federal statute in the United States

are protected in most countries: political opinion, sexual orientation, and marital/family status.

Several protected classes are covered in only a few counties or are unique to a few countries; Table

30.5 identifi es these less common protected classes. Examples include language, physical appear-

ance, union membership, socioeconomic status, and HIV status.

Question 5: What is required as prima facie evidence of discrimination? What is required to refute a claim of discrimination?

In most countries, direct (e.g., differential treatment) and indirect (e.g., disparate impact) prima

facie evidence of discrimination are acknowledged. In India, disparate impact is necessary but

not suffi cient to prove a case of discrimination; underrepresentation must be shown to be due to

historical social or religious discrimination toward a particular group. Only two countries require

evidence of the intent to discriminate, Taiwan and Turkey, thus ruling out a disparate impact theory

of discrimination.

However, although disparate impact evidence can be used as evidence in most countries, highly

specifi c evidentiary rules used in the United States (e.g., the four-fi fths rule and tests of the statisti-

cal signifi cance of the difference between passing rates for various groups) are generally not in use

(Canada, is an exception, because cases using the four-fi fths rule in the United States have been

used to make a case for a similar standard). Commentators note that in most cases there are few or

no cases involving disparate treatment challenges to predictors commonly used by psychologists,

and thus, there is not the extensive case law that has developed in the United States. Recall that the

four-fi fths rule in the United States derives from guidelines issued by enforcement agencies, and

the use of signifi cance testing derives from case law; neither the concept of disparate impact nor

the mechanisms for identifying its presence are contained in statute. Absent a history of challenges

resulting in case law, it is not surprising to see the lack of specifi city as to evidentiary standards.

A similar lack of specifi city applies to the question of what is required to refute a claim of dis-

crimination. Table 30.6 summarizes information across countries. In general, there is some version

of the shifting burden of proof model in countries where disparate impact evidence is permissible.

After a prima facie showing, the burden to justify the use of the employment practice shifts to the

employer in all countries except Switzerland, where the burden of showing that the practice is not





TAB

LE 3

0.4

Mos

t C

omm

on P

rote

cted

Cla

sses

Cou

ntry

Com

mon

Pro

tect

ed C

lass

es

Rac

eSe

xN

atio

nal/

Ethn

ic O

rigi

nC

olor

Age

Rel

igio

nD

isab

ility

Polit

ical

Opi

nion

Sexu

alO

rien

tati

onM

arit

al/

Fam

ily S

tatu

s

Aust

rali

aX

XX

XX

XX

Bel

giu

mX

XX

XX

XX

XX

X

Can

ada

XX

XX

XX

XX

X

Chil

eX

XX

XX

XX

X

Fra

nce

XX

XX

XX

XX

X

Ger

man

yX

XX

XX

X

Gre

ece

XX

XX

X

India

XX

Isra

elX

XX

XX

XX

X

Ital

yX

XX

XX

XX

Japan

XX

XX

XX

Ken

ya

XX

XX

XX

X

Kore

aX

XX

XX

XX

XX

X

The

Net

her

lands

XX

XX

XX

XX

X

New

Zea

land

XX

XX

XX

XX

XX

South

Afr

ica

XX

XX

XX

XX

XX

Spai

nX

XX

XX

XX

XX

Sw

itze

rlan

dX

XX

XX

XX

Tai

wan

XX

XX

XX

X

Turk

eyX

XX

XX

Unit

ed K

ingdom

XX

XX

XX

X

Unit

ed S

tate

sX

XX

XX

XX





job-related is only partially reduced or remains with the plaintiff. There is a general notion that

the employer should present evidence to support the job relatedness of the employment practice in

question, but rarely is the required form of such evidence specifi ed. The identifi cation of validity

evidence as a mechanism for establishing job relatedness is rare.

Question 6: What are the consequences of violation of the laws?Table 30.6 also summarizes possible consequences of violation in each participating country.

There is considerable variation in the array of possible remedies. As a point of reference, note that in

the United States the focus is on compensatory or “make-whole” remedies, with punitive damages

reserved for instances of intentional discrimination. Similarly, make-whole remedies are part of the

landscape in all countries for which information could be obtained. Several countries also provide

fi nes and punitive damages (e.g., Switzerland and Turkey), and several include imprisonment as a

possible consequence (e.g., Belgium, France, and Greece).

Question 7: Are particular selection methods limited or banned as a result of legislation or court rulings?

There are relatively few restrictions on specifi c selection methods. As a point of reference, U.S.

law regulates the use of the polygraph, prohibiting its use for most private employers; several other

countries restrict polygraph use as well (e.g., Germany, Israel, and Turkey). The only selection

method specifi cally mentioned in U.S. law is the reference in the Tower amendment to Title VII

of the Civil Rights Act of 1964 (U.S. Code, 1964) to the permissibility of professionally developed

TABLE 30.5Other Protected Classes by Country

Country Other Protected Classes

Australia Breastfeeding, family or career responsibilities, irrelevant criminal record, physical features,

potential pregnancy, trade union or employer association activity, sexual harassment, pregnancy and

transgender status

Belgium Union membership, membership of other organizations, health, and any other personal characteristic

Chile Union membership status

France Moral principles, genetic characteristics, union activities or activities in a “mutuelle,” physical

appearance, family name, and health

Germany Philosophy of life, sexual harassment

India Scheduled castes, scheduled tribes, and other backward classes

Israel Personal status and military service

Italy Personal and social conditions and language

Japan Social status

Kenya Tribe, local connection, and HIV/AIDS status

Korea Social status, region of birth, appearance, criminal record after punishment has been served,

academic background, medical history, pregnancy, and physical conditions (e.g. appearance, height,

weight)

The Netherlands Philosophy of life, chronic disease, full-time/part-time work, and type of contract

New Zealand Ethical belief, employment status, and sexual and racial harassment

South Africa HIV status, conscience, belief, culture, birth, pregnancy, and language

Spain Social condition and membership to a labor union

Switzerland Socioeconomic status, way of life, and language

Taiwan Thought, provincial origin, appearance, facial features, union membership, status, and language

Turkey Philosophical belief, sect, and language

United Kingdom Persons who have undergone gender reassignment or intend to

United States Pregnancy





TAB

LE 3

0.6

Evid

ence

Nee

ded

to R

efut

e a

Dis

crim

inat

ion

Cla

im, C

onse

quen

ces

of V

iola

tion

, and

Per

mis

sibi

lity

of P

refe

rent

ial T

reat

men

t by

Cou

ntry

Cou

ntry

Evid

ence

Nee

ded

to R

efut

e a

Cla

imC

onse

quen

ces

of V

iola

tion

Perm

issi

bilit

y of

Pre

fere

ntia

l Tre

atm

ent

Aust

rali

a In

her

ent

requir

emen

ts o

f th

e jo

b, ex

iste

nce

of

spec

ial

mea

sure

s to

eli

min

ate

dis

crim

inat

ion, occ

upat

ional

requir

emen

ts, ac

tions

requir

ed b

y l

aw, em

plo

ym

ent

wit

hin

sm

all

org

aniz

atio

ns,

consi

sten

t bel

iefs

(e.

g.

reli

gio

us

org

aniz

atio

ns

or

educa

tional

inst

itute

s). T

he

stat

ute

s m

ake

no r

efer

ence

to t

he

psy

cholo

gic

al

conce

pt

of

val

idit

y n

or

has

it

aris

en i

n c

ase

law

.

Inju

nct

ion t

o s

top t

he

act,

aw

ard o

f dam

ages

, ord

er t

o

the

org

aniz

atio

n t

o r

edre

ss t

he

situ

atio

n, var

iati

on, or

cance

llat

ion o

f a

contr

act

or

agre

emen

t th

at v

iola

tes

the

law

.

Wit

hin

-gro

up n

orm

ing i

s not

ban

ned

and i

s use

d b

y

som

e psy

cholo

gic

al t

este

rs a

s a

mea

ns

of

com

ply

ing

wit

h l

egis

lati

on (

Myors

, 2003).

Tar

get

s m

ay b

e use

d

in s

om

e E

EO

pla

ns,

but

expli

cit

quota

s ar

e av

oid

ed.

Bel

giu

mS

tati

stic

al d

ata

or

pra

ctic

al t

ests

can

be

use

d a

s

evid

ence

.

Med

iati

on o

r bin

din

g j

udgm

ent

from

civ

il c

ourt

.

Impri

sonm

ent

and/o

r fi

nes

.

Pre

fere

nti

al t

reat

men

t is

per

mit

ted t

o r

emed

y a

his

tori

cal

dis

crim

inat

ion a

gai

nst

a g

roup. Q

uota

s ar

e

per

mit

ted, but

seld

om

uti

lize

d. S

om

e org

aniz

atio

ns

also

uti

lize

tar

get

num

ber

s.

Can

ada

The

emplo

yer

must

dem

onst

rate

th

at t

he

emplo

ym

ent

poli

cy, pra

ctic

e, o

r pro

cedure

that

is

chal

lenged

is

a

bona

fi de

occ

upat

ional

req

uir

emen

t. T

ribunal

s an

d

court

s ar

e quit

e li

ber

al i

n t

he

evid

ence

that

they

wil

l

acce

pt

from

em

plo

yer

s in

def

ense

of

thei

r

emplo

ym

ent

pra

ctic

es. E

mpir

ical

and s

tati

stic

al

evid

ence

gen

erat

ed b

y I

-O p

sych

olo

gis

ts (

e.g., l

oca

l

val

idat

ion s

tudie

s) m

ay b

e use

ful

in d

efen

din

g

emplo

ym

ent

pra

ctic

es, but

court

s an

d t

ribunal

s oft

en

lack

the

sophis

tica

tion t

o m

ake

full

use

of

such

det

aile

d a

nd c

om

ple

x t

echnic

al i

nfo

rmat

ion.

Fin

es, pay

men

t fo

r lo

st w

ages

, re

inst

atem

ent,

and

ord

erin

g o

f sp

ecia

l pro

gra

ms.

Pre

fere

nti

al t

reat

men

t per

mit

ted (

mai

nly

in t

he

publi

c

sect

or)

.

Chil

eU

ncl

ear,

unle

ss f

or

sexual

har

assm

ent

or

unio

niz

atio

n

suit

s. E

mpir

ical

evid

ence

not

requir

ed.

Unknow

n. C

urr

entl

y, s

exual

har

assm

ent

suit

s m

ay

resu

lt i

n m

onet

ary c

om

pen

sati

on a

nd u

p t

o 3

yea

rs

impri

sonm

ent.

Gover

nm

ent

has

enac

ted a

n i

nfo

rmal

quota

for

wom

en

in m

inis

ter

posi

tions;

how

ever

, th

is h

as n

ot

cross

ed

over

into

the

pri

vat

e se

ctor.





Fra

nce

Vag

ue.

Em

plo

yer

should

pre

sent

any i

nfo

rmat

ion

show

ing t

he

dec

isio

n i

s le

git

imat

e, n

ondis

crim

inat

ory

,

and b

ased

on o

bje

ctiv

e in

form

atio

n.

Thre

e yea

rs i

mpri

sonm

ent

and/o

r a

fi ne

for

convic

tion

in a

cri

min

al c

ourt

. D

iscr

imin

atory

act

is

annull

ed i

n a

civil

court

and p

oss

ibly

fi n

anci

al c

om

pen

sati

on.

Consi

der

able

dis

cuss

ion a

bout

this

; poli

tica

lly,

pre

fere

nti

al t

reat

men

t is

see

n a

s undes

irab

le. H

ow

ever

,

ther

e ar

e se

ttin

gs

wher

e it

is

use

d.

When

par

ties

pre

sent li

sts

of

candid

ates

for

regio

nal

and

senat

ori

al e

lect

ions

they

are

req

uir

ed to h

ave

an e

qual

num

ber

of

men

and w

om

en. A

lso, th

ere

are

quota

s in

one

sett

ing: at

lea

st 6

% o

f w

ork

forc

e nee

ds

to b

e han

dic

apped

for

org

aniz

atio

ns

wit

h m

ore

than

20 e

mplo

yee

s.

Ger

man

yN

eeds

to b

e bas

ed o

n j

ob r

equir

emen

ts.

Em

plo

yee

has

rig

ht

to r

efuse

to w

ork

whil

e on

pay

roll

and s

ue

emplo

yer

s fo

r dam

ages

.

No f

orm

aliz

atio

n, but

publi

c au

thori

ties

are

to g

ive

pre

fere

nce

to w

om

en a

nd h

andic

apped

per

sons.

Gre

ece

Em

plo

yer

must

show

that

ther

e has

bee

n n

o

bre

ach o

f th

e pri

nci

ple

of

equal

tre

atm

ent.

The

emplo

yer

who i

nfr

inges

the

law

s ab

out

equal

trea

tmen

t on t

he

gro

unds

of

raci

al o

r et

hnic

ori

gin

,

reli

gio

n o

r bel

ief,

dis

abil

ity,

age

or

sex i

s punis

hed

by

impri

sonm

ent

of

6 m

onth

s up t

o 3

yea

rs a

nd t

oget

her

wit

h a

pen

alty

of

1,0

00 u

p t

o 5

,000 e

uro

s.

Pre

fere

nti

al t

reat

men

t to

pre

ven

t or

com

pen

sate

for

dis

advan

tages

lin

ked

to a

ny o

f th

e pro

tect

ed c

lass

es.

India

At

the

dis

cret

ion o

f th

e ju

dge.

P

refe

renti

al t

reat

men

t in

the

form

of

a re

laxat

ion o

f

qual

ifyin

g s

core

s fo

r pro

tect

ed g

roups

in e

xte

rnal

recr

uit

men

t is

per

mit

ted;

how

ever

, a

com

mon s

tandar

d

is r

equir

ed f

or

pro

moti

on. N

ot

all

mem

ber

s of

pro

tect

ed

gro

ups

are

equal

ly e

ligib

le, al

so d

epen

den

t on s

oci

al/

econom

ic s

tatu

s. G

over

nm

ent

posi

tions

also

use

quota

s.

Isra

elE

vid

ence

of

test

rel

iabil

ity a

nd v

alid

ity,

whic

h c

an b

e

bas

ed o

n v

alid

ity g

ener

aliz

atio

n. In

addit

ion, th

e

Nat

ional

Lab

or

Court

rec

entl

y r

ule

d t

hat

em

plo

yer

s

seek

ing t

o p

rove

thei

r in

noce

nce

wil

l be

subje

ct t

o

less

sev

ere

test

s of

sele

ctio

n v

alid

ity t

o t

he

exte

nt

that

they

are

acc

use

d o

f dis

crim

inat

ing a

gai

nst

inte

rnal

as

oppose

d t

o e

xte

rnal

can

did

ates

; th

e lo

gic

bei

ng t

hat

emplo

yer

s ty

pic

ally

hav

e fa

r gre

ater

info

rmat

ion u

pon

whic

h t

o b

ase

a se

lect

ion d

ecis

ion w

hen

choosi

ng

among i

nte

rnal

can

did

ates

.

Sm

all

fi nes

. H

irin

g, re

inst

atem

ent,

or

care

er

advan

cem

ent

of

pla

inti

ff, pay

men

t of

bac

k w

ages

.

Pre

fere

nti

al t

reat

men

t is

req

uir

ed b

y p

ubli

c

org

aniz

atio

ns

and s

tate

-ow

ned

ente

rpri

ses

for

wom

en

and m

inori

ties

. P

refe

renti

al t

reat

men

t is

per

mit

ted i

n

the

pri

vat

e se

ctor.

Ital

yV

alid

ity e

vid

ence

not

reques

ted. E

vid

ence

to r

efute

a

clai

m i

s cu

rren

tly u

ncl

ear.

Unknow

n.

Pre

fere

nti

al t

reat

men

t per

mit

ted f

or

wom

en.

cont

inue

d





TAB

LE 3

0.6

(con

tinu

ed)

Evid

ence

Nee

ded

to R

efut

e a

Dis

crim

inat

ion

Cla

im, C

onse

quen

ces

of V

iola

tion

, and

Per

mis

sibi

lity

of P

refe

rent

ial T

reat

men

t by

Cou

ntry

Cou

ntry

Evid

ence

Nee

ded

to R

efut

e a

Cla

imC

onse

quen

ces

of V

iola

tion

Perm

issi

bilit

y of

Pre

fere

ntia

l Tre

atm

ent

Japan

Adm

inis

trat

ive

advic

e.P

refe

renti

al t

reat

men

t per

mit

ted a

nd s

upport

ed b

y t

he

gover

nm

ent.

Quota

s re

quir

ed f

or

dis

able

d.

Ken

ya

Must

show

that

dec

isio

ns

wer

e bas

ed o

n a

ppli

cant

apti

tudes

and a

bil

itie

s. E

mpir

ical

val

idit

y e

vid

ence

not

requir

ed.

Rem

edy b

y f

oll

ow

ing r

ecom

men

dat

ions

of

Min

istr

y o

f

Hea

lth, L

abour,

& W

elfa

re.

Poss

ible

publi

c an

nounce

men

t of

vio

lati

on. C

ivil

fi n

e

of

max

imum

200,0

00 y

en (

$2,4

00 U

.S.)

.

Dif

fere

nt

cut-

off

sco

res

are

set

for

mem

ber

s fr

om

dif

fere

nt

ethnic

gro

ups

to e

nsu

re t

hat

som

e m

ember

s

from

eac

h g

roup w

ill

be

sele

cted

. T

her

e ar

e re

quir

ed

quota

s of

5%

in t

he

pri

vat

e an

d p

ubli

c se

ctor

for

dis

able

d i

ndiv

idual

s.

Kore

aS

how

job r

elat

ednes

s, b

ut

spec

ifi c

met

hod u

ncl

ear.

N

atio

nal

Hum

ans

Rig

ht

Com

mis

sion w

ill

mak

e a

bin

din

g c

onci

liat

ion r

esolu

tion. F

ines

.

Quota

s re

quir

ed f

or

dis

able

d. P

refe

renti

al t

reat

men

t fo

r

wom

en, al

though fi

rm

s w

ith o

ver

50%

wom

en i

n

work

forc

e ar

e ex

empt.

The

Net

her

lands

Gen

eral

ly n

o v

alid

ity e

vid

ence

is

reques

ted b

ecau

se t

he

val

idit

y o

f co

mm

on p

sych

olo

gic

al t

ests

, su

ch a

s te

sts

for

cognit

ive

abil

itie

s, p

erso

nal

ity i

nven

tori

es a

nd

asse

ssm

ent

cente

r ex

erci

ses,

is

taken

for

gra

nte

d.

Most

cla

ims

conce

rn d

irec

t dis

crim

inat

ion o

r

trea

tmen

t dis

crim

inat

ion (

Com

mis

sie

Gel

ijke

Beh

andel

ing, 2006).

Exce

pti

ons

are

clea

r-cu

t ca

ses

of

indir

ect

dis

crim

inat

ion i

n w

hic

h i

nap

pro

pri

ate

job

requir

emen

ts w

ere

set.

Nonbin

din

g j

udgm

ent

by t

he

Com

mis

sion o

f E

qual

Tre

atm

ent

and p

oss

ibly

judgm

ent

refe

rral

to a

civ

il

court

.

Pre

fere

nti

al t

reat

men

t is

per

mit

ted f

or

wom

en a

nd

ethnic

min

ori

ties

(does

not

hav

e to

be

equal

ly

qual

ifi e

d).

New

Zea

land

Uncl

ear,

bec

ause

few

cas

es m

ake

it t

o c

ourt

. G

enuin

e

Occ

upat

ional

char

acte

rist

ics

(GO

Q).

Apolo

gy,

pay

men

t or

com

pen

sati

on, as

sura

nce

that

the

dis

crim

inat

ory

act

wil

l not

be

repea

ted, or

refe

rral

to a

Hum

an R

ights

Tri

bunal

for

furt

her

judgm

ent.

This

is

curr

entl

y b

eing e

xplo

red. P

refe

renti

al t

reat

men

t

appea

rs t

o b

e per

mit

ted (

and m

ay b

e so

on a

ppli

ed t

o

the

Mao

ri p

opula

tion).

South

Afr

ica

Qual

itat

ive

and e

mpir

ical

dat

a ca

n b

e bro

ught

to b

ear

to s

upport

val

idit

y.

Fin

es. P

oss

ible

can

cell

atio

n o

f gover

nm

ent

contr

acts

. P

refe

renti

al t

reat

men

t is

per

mit

ted a

nd a

ppli

ed. R

acia

l

quota

s ar

e le

gal

and p

ract

iced

by m

any l

arge

emplo

yer

s. T

he

pra

ctic

al i

mpli

cati

on f

or

this

is

that

it

is l

egal

in t

he

South

Afr

ican

conte

xt

to u

se r

ace

norm

ing, or

wit

hin

-gro

up t

op-d

ow

n s

elec

tion

stra

tegie

s, t

o a

ddre

ss a

ffi r

mat

ive

acti

on n

eeds

of

org

aniz

atio

ns.





Spai

nR

ecen

t la

ws

may

lea

d t

o g

reat

er f

ocu

s on e

mpir

ical

evid

ence

; up u

nti

l now

, val

idit

y o

f te

sts

was

tak

en

for

gra

nte

d.

Com

pen

sati

on, re

ject

ion o

f th

e dec

isio

n, an

d

subse

quen

t ap

pli

cati

on o

f th

e co

urt

dec

isio

n,

repet

itio

n o

f th

e se

lect

ion p

roce

ss w

ith n

ew

pro

cedure

s.

Pre

fere

nti

al t

reat

men

t fo

r w

om

en i

n s

om

e ca

ses.

Sw

itze

rlan

dE

mpir

ical

evid

ence

not

gen

eral

ly p

rese

nte

d o

r

requir

ed.

Court

s ca

n a

war

d d

amag

es i

ncl

udin

g p

aym

ent

of

ow

ed e

arnin

gs

and p

aym

ent

of

com

pen

sati

on a

nd

sati

sfac

tion.

Pre

fere

nce

is

per

mit

ted b

ut

not

requir

ed.

Tai

wan

Pro

vid

e ev

iden

ce o

f jo

b r

elat

ednes

s.

Fin

es.

Quota

s re

quir

ed f

or

abori

gin

es (

at l

east

1%

of

pri

vat

e

org

aniz

atio

ns’

work

forc

e).

Turk

eyR

einst

atem

ent,

bac

k p

ay, an

d/o

r m

onet

ary d

amag

es.

Pre

fere

nti

al t

reat

men

t is

not

requir

ed o

r per

mit

ted a

nd

is a

ctual

ly f

orb

idden

.

Unit

ed

Kin

gdom

Show

that

req

uir

emen

t is

just

ifi e

d. T

he

emplo

yer

can

show

that

they

took a

ll “

reas

onab

le”

step

s to

pre

ven

t

dis

crim

inat

ion. N

o i

mpac

t ca

ses

involv

ing t

ests

hav

e

reac

hed

the

stag

e of

a co

urt

dec

isio

n, so

ther

e is

as

yet

no r

equir

emen

t of

val

idit

y e

vid

ence

.

Court

has

dis

cret

ion. C

om

pen

sati

on t

o t

he

pla

inti

ff.

Form

al i

nves

tigat

ion b

y g

over

nin

g b

odie

s th

at c

an

reco

mm

end c

han

ges

in p

roce

dure

s.

Pre

fere

nti

al t

reat

men

t is

not

per

mit

ted, but

“posi

tive

acti

on”

such

as

enco

ura

gin

g c

erta

in g

roups

to a

pply

or

off

erin

g t

rain

ing t

o t

hes

e gro

ups.

Unit

ed S

tate

sE

vid

ence

that

the

chal

lenged

pra

ctic

e is

job-r

elat

ed f

or

the

posi

tion i

n q

ues

tion a

nd c

onsi

sten

t w

ith b

usi

nes

s

nec

essi

ty (

larg

ely t

hro

ugh v

alid

ity s

tudie

s).

Upon a

fi n

din

g o

f dis

crim

inat

ion, a

judge

can s

pec

ify

“mak

e w

hole

” re

med

ies,

such

as

bac

k p

ay, hir

ing, or

rein

stat

emen

t. T

her

e ar

e no p

unit

ive

dam

ages

abse

nt

a

fi ndin

g o

f in

tenti

onal

dis

crim

inat

ion.

1991 a

men

dm

ents

to T

itle

VII

of

Civ

il R

ights

Act

pro

hib

it p

refe

renti

al t

reat

men

t, s

pec

ifi c

ally

in t

he

form

of

adju

stin

g s

core

s or

usi

ng s

epar

ate

norm

s fo

r

min

ori

ty g

roup m

ember

s. P

refe

renti

al t

reat

men

t is

per

mit

ted a

fter

a fi

ndin

g o

f dis

crim

inat

ion a

s par

t of

a

judic

iall

y o

rder

ed r

emed

y.





ability tests, provided that such tests are not designed, intended, or used to discriminate. Additional

instances reported of restrictions on specifi c selection methods in participating countries include a

prohibition against comprehensive personality assessment in Switzerland and a restriction on the

use of certain Minnesota Multiphasic Personality Inventory (MMPI) and California Psychological

Inventory (CPI) items in Spain.

The most strikingly different approach to regulating selection practices is found in South

Africa. Rather than the common approach of a presumptive right of an employer to use a particu-

lar method absent a successful challenge by a plaintiff, South African law puts the burden imme-

diately on the employer. According to the Employment Equity Act of 1998 (Government Gazette,

1999), psychological testing and other similar assessments are prohibited unless the test is proven

to be scientifi cally valid and reliable, can be applied fairly to all employees, and is not biased

against any employee or group. The Society for Industrial and Organizational Psychology (SIOP)

in South Africa published “Guidelines for the Validation and Use of Assessment Procedures for

the Workplace” during 2005 to provide guidelines for practitioners in the fi eld of I-O psychology to

ensure that their assessment instruments and practices comply with the scientifi c requirements

and international best practices. These guidelines were largely based on the American SIOP

guidelines.

Question 8: What is the legal status of preferential treatment of members of minority groups (e.g., quotas or softer forms of preference)?

To set the stage, note that the term “affi rmative action” is used in various contexts, only some of

which involve preferential treatment for protected groups. Some forms of affi rmative action involve

outreach efforts to publicize openings and to encourage applications from members of protected

groups. However, there is no preferential treatment given once an individual is in the applicant pool.

Approaches involving preferential treatment fall into two main classes: (a) those that set differing

standards for protected and nonprotected groups without setting aside a specifi ed number or propor-

tion of openings for members of protected groups (e.g., using different cut-off scores, using within-

group norming) and (b) quota approaches that set aside a fi xed number or proportion of openings

for members of protected groups.

Table 30.6 summarizes the status of preferential treatment in the participating countries. Preferential

treatment is a domain in which the United States emerges as a clear outlier. Preferential treatment in

terms of differing score cutoffs or separate norming of tests within group is prohibited by the U.S.

Civil Rights Act of 1991 (U.S. Code, 1991), and the use of quotas is restricted to very limited set-

tings, such as a court-ordered remedy following a fi nding of discrimination. In contrast, in only two

countries do commentators report a prohibition against minority preference (Turkey and the United

Kingdom). The types of preference permitted and the settings in which it is used do vary widely. The

status of quotas varies, from prohibited (Australia), to permitted but rarely used (Belgium), to permit-

ted and widely used (South Africa), to used in government sectors (backward classes in India and

women in Chile), to required for certain groups (e.g., aborigines in Taiwan, individuals with disabili-

ties in France, Japan, Kenya, and Korea). Several commentators note that applying lower standards to

protected groups (e.g., different cutoffs or within-group norming) is used (Australia, India, and South

Africa). In India, lower qualifying scores for protected groups are permitted for external selection,

but not for promotion.

Question 9: How have laws and the legal environment affected the practice of science-based employee selection in this country?

In only a few countries (Canada, South Africa, and the United States) is the legal environment

seen as having a large effect on science-based employee selection. In general, this can partially be

attributed to the much more amorphous legal standards and consequences with regards to employ-

ment discrimination in most countries surveyed. The reciprocal relationship between science-based

selection and the legal environment will need to be continually monitored because many countries





are still in the process of developing legal statutes and requirements or establishing guidelines for

the prosecution and rulings on employment discrimination.

Overall, most employers in the countries surveyed have great latitude in choosing what selec-

tion procedures to utilize. However, most employers are aware of the social and political nature of

selection procedures and seem to err on the side of mainstream, popular, and usually well-validated

selection methods. The most common type of selection procedures do vary by country. It is com-

mon to see reports of increased use of the tools and techniques of science-based selection, but the

driving forces are more commonly the presence of multinational fi rms and consulting fi rms that

import these techniques into the country.

DISCUSSION

In this section we offer 35 broad summary statements about the patterns emerging from the narra-

tives from the various countries.

DISADVANTAGED GROUPS

1. Disadvantaged groups could be divided into four main groups: immigrants or foreign resi-

dents, religious minorities, racial/ethnic minorities, and language group minorities (speak

different primary language).

2. Many European (especially European Union) nations have disadvantaged groups who are

immigrants or foreign workers. The groups that are disadvantaged are usually Eastern

European or African.

3. Many Asian countries also have disadvantaged groups who are immigrants or foreign

workers.

4. Many of the racial/ethnic minorities are indigenous people (e.g., Australia, Canada, New

Zealand, Taiwan, and the United States).

5. Most disadvantaged groups are a relatively small proportion of the population, most below

the 20% “breaking point” specifi ed in research on tokenism (Kanter, 1977).

6. Disadvantaged groups can constitute the majority of the population (e.g., South Africa).

WOMEN IN THE WORKPLACE

7. Women are now well represented in the workforce, and between one quarter to approxi-

mately one half of the workforce are women in most countries.

8. Women have generally substantially increased their participation rate in the workforce in

the last decade. However, men’s rate of participation in the workforce continues to greatly

outstrip that of women.

9. Women are still underrepresented in management and professional positions. However,

European nations and the United States have a sizeable representation of women in lower

and middle-management positions. However, all countries have very few women in top-

and senior-management positions.

10. Wage differentials are still sizeable between men and women; women generally earn

60–80 cents to the dollar compared with men.

11. Considerable occupational segregation remains for women, such that women tend to be

heavily concentrated in lower-income-segment occupations. These include clerical/secre-

tarial jobs, service jobs, nursing and childcare services, and primary education.

12. Women tend to engage in more part-time work (partly because of childcare

responsibilities).





SUBGROUP MEAN DIFFERENCES

13. Very few countries have research exploring potential mean differences in cognitive ability,

personality, or job performance. In terms of cognitive ability, fi ndings usually favor the

advantaged group and/or men.

14. Mean differences between local and immigrant populations are affected by immigration

policies. Targeting either high- or low-skill immigrants can affect the magnitude and direc-

tion of mean differences.

DISCRIMINATION LAWS

15. Every country has a law or directive that prevents discrimination on the basis of sex or

race/ethnic origin and many other personal characteristics and beliefs.

16. Most discrimination cases seem to be settled by special commissions and/or courts rather

than by juries (which do not exist in several countries).

17. In many countries, few actual cases are actually fi led and/or brought to trial, not because

discrimination does not occur, but because workers do not understand their rights, are not

used to protecting these rights (e.g., collectivistic orientation, etc.), or do not see much

benefi t in going to court.

18. Punishment is generally usually rather light (e.g., minimal to moderate fi ne or reinstate-

ment, payment of back wages).

19. Concerns about privacy are very prominent in Europe. Many European countries are so

concerned that data on race or gender are not collected.

MAKING AND REFUTING A CLAIM OF DISCRIMINATION

20. For many countries, although there are laws in place, there is very little clarity about how

to establish discrimination and/or what kind of evidence required.

21. Intent to discriminate is not required in most countries (exceptions are Taiwan and

Turkey).

22. Most discrimination cases are handled on a case-by-case basis and are based on treating

people differently on the basis of group membership (direct discrimination) rather than a

procedure or test that systematically disadvantages a group (indirect discrimination). In

most countries surveyed, both are illegal.

23. Few actual cases outside of the United States challenging the adverse impact or discrimi-

natory nature of formal tests (cognitive ability or personality) exist, and therefore most

countries do not really use validity evidence to refute discrimination.

24. Most countries do not require validity evidence. In many places the empirical validity of

formal tests (e.g., cognitive ability, personality) is implicitly assumed.

25. Most countries do not use relevant workforce comparisons as a basis for discrimination

although this information is sometimes taken under consideration in certain countries.

26. The evidence to refute a claim of discrimination is usually some qualitative evidence of

job-relatedness or bona fi de occupational requirement.

MINORITY PREFERENCE

27. Minority preference is permitted (and even recommended) in most countries. This is more

likely to be true for women or those with disabilities than for racial groups.

28. It is more common for government entities than for private-sector fi rms to engage in prac-

tices involving preferential treatment.





29. Forms of affi rmative action vary, ranging from active recruitment and training of women

or racial groups that have been traditionally disadvantaged to lower standards for these

groups.

30. Quotas are relatively rare but are present in several countries; for example, India (lower

castes), Taiwan (aborigines), Korea and France (disabled), and South Africa (race and

gender).

31. Explicitly forbidding preferential treatment is rare (e.g., Turkey).

SPECIFIC SCIENCE-BASED SELECTION TOOLS

32. Generally, science-based tools are not explicitly referenced in laws or in common legal

practices (exceptions include South Africa, Switzerland, and the United Kingdom).

33. Generally, although fi rms are free to use whatever selection methods they desire, large

fi rms tend to be aware of social and business pressures for effective selection.

34. The selection method that is most limited/banned is the polygraph.

35. Selection practice tends to be infl uenced more by the presence of multinational corpora-

tions and consulting fi rms than by legal pressures (with the exception of the United States,

Canada, and South Africa).

We fully anticipate that some readers may question the value of knowing the legal environment

of countries other than their own, because they are inevitably bound by the legal constraints of the

country they operate in. We have several responses. First, in today’s global world, more and more

fi rms engage in business that extends across national boundaries. Second, there is value in extend-

ing one’s framework beyond the national setting with which one is most familiar. Discovering that

the same issue is treated differently elsewhere breaks the mold of viewing a certain set of circum-

stances as inevitable. Third, documenting these differences sets the stages for comparative research

asking questions about why certain variations are found. For example, why is preferential treatment

not generally permitted and held in such negative popular opinion in the United States and not in

many other countries? Why are some groups protected in some countries but not others?

In conclusion, we hope this compilation of information about perspectives from a wide range of

countries is useful to students, researchers, and practitioners around the globe. We encourage inter-

national collaborations on other workplace issues, and hope this project provides a useful model.

AUTHORS’ NOTE

This research was conducted while Antonio Mladinic was on leave from the Pontifi cia Universidad

Católica de Chile and holding a visiting appointment at the University of Texas at El Paso and while

Herman Aguinis was on sabbatical leave from the University of Colorado Denver and holding a

visiting appointment at the University of Salamanca (Spain). Oleksandr Chernyshenko is now at the

Nanyang Business School, Nanyang Technological University (Singapore).

REFERENCES

Attal-Toubert, K., & Lavergne, H. (2006). Premiers résultats de l’enquête sur l’emploi 2005. [Initial results

from the 2005 employment survey]. INSEE Première, number 1070. Paris: INSEE. Retrieved April 15,

2007, from http://www.insee.fr/fr/ffc/ipweb/ip1070/ip1070.pdf

Cascio, W. F., & Aguinis, H. (2008). Applied psychology in human resource management (6th ed.). Upper

Saddle River, NJ: Pearson Education.

Chernyshenko, O. S. (2005). Report on psychometric evaluation of the general reasoning test (GRT2) for the New Zealand Police: Measurement equivalence across ethnic and gender groups. Auckland, New

Zealand: OPRA Consulting Group.



http://www.insee.fr/fr/ffc/ipweb/ip1070/ip1070.pdf



Council of Indigenous Peoples. (2002). Yearbook of Taiwanese aborigines statistics. Taipei, Taiwan: Executive

Yuan.

Equal Opportunities Commission. (2004). Sex and power: Who runs Britain. Manchester, England: Author.

Fontaine, J. R. J., Schittekatte, M., Groenvynck, H., & De Clercq, S. (2006). Acculturation and intelligence among Turkish and Moroccan adolescents in Belgium. Unpublished manuscript, Ghent University,

Ghent, Belgium.

Government Gazette. (1999). Employment Equity Act, 1998 (Act No. 55 of 1998), R 1360.

Guenole, N., Englert, P., & Taylor, P. (2003). Ethnic group differences in cognitive ability test scores within a

New Zealand applicant sample. New Zealand Journal of Psychology, 23, 39–54.

Guion, R. M. (1998). Assessment, measurement, and prediction for personnel decisions. Mahwah, NJ: Lawrence

Erlbaum.

Japan Institute of Labor Policy and Training. (2007). Labor situation in Japan and analysis 2006/2007. Retrieved

June 5, 2007, from http://www.jil.go.jp/english/laborinfo/library/documents/Labor2006_2007.pdf

Kanter, R. M. (1977). Men and women of the corporation. New York, NY: Basic Books.

Kinyungu, C. (2006). KCPE: Public schools feel the heat. Retrieved January 31, 2007, from http://www.

eastandard.net/archives/cl/hm_news/news.php?articleid=1143963072

Kriek, H. J. (2006). Personality assessment: Group differences, language profi ciency and fairness. Presented at

the Society of Industrial and Organizational Psychology Conference, May 2006, Dallas, TX.

Myors, B. (2003). Within-group norming: Just because it’s illegal in America, doesn’t mean we can’t do it here.

Paper presented at the 5th Australian Conference on Industrial/Organisational Psychology, Melbourne,

Australia.

Myors, B., Lievens, F., Schollaert, E., Van Hoye, G., Cronshaw, S. F., Mladinic, A., et al. (2008). International

perspectives on the legal environment for selection. Industrial and Organizational Psychology: Perspectives on Science and Practice, 1, 206–256.

Roth, P. L., Bevier, C. A., Bobko, P., Switzer, F. S., & Tyler, P. (2001). Ethnic group differences in cognitive

ability in employment and educational settings: A meta-analysis. Personnel Psychology, 54, 297–330.

SHL. (2006). Validity study V036. Retrieved on June 7, 2007, from http://www.shl.com/globallocations/pages/

southafrica.aspx

Society for Industrial and Organisational Psychology in South Africa. (2005). Guidelines for the validation

and use of assessment procedures for the workplace. Retrieved on June 7, 2007, from http://www.siopsa.

org.za

State Institute of Statistics. (2006). Census of population: Social and economic characteristics. Ankara, Turkey:

Author.

Statistics South Africa. (2001). Census 2001: Primary tables South Africa, Census 1996 and 2001 compared.

Johannesburg, South Africa: Author.

te Nijenhuis, J., de Jong, M., Evers, A., & van der Flier, H. (2004). Are cognitive differences between immi-

grant and majority groups diminishing? European Journal of Personality, 18, 405–434.

te Nijenhuis, J., van der Flier, H., & van Leeuwen, L. (1997). Comparability of personality test scores for

immigrants and majority group members: Some Dutch fi ndings. Personality and Individual Differences, 23, 849–859.

te Nijenhuis, J., van der Flier, H., & van Leeuwen, L. (2003). The use of a test for neuroticism, extraver-

sion, and rigidity for Dutch immigrant job-applicants. Applied Psychology: An International Review, 52,

630–647.

U.S. Code. (1964). Pub. L. 88-352.

U.S. Code. (1991). Pub. L. 102-166.

U.S. Equal Employment Opportunity Commission. (1978). Uniform guidelines on employee selection proce-dure. 29 CFR 1607.1. Washington, DC: Author.

U.S. Offi ce of General Accounting. (2003). Women’s earnings: Work patterns explain difference between men and women’s earning. Retrieved on June 20, 2008, from http://usgovinfo.about.com/gi/dynamic/offsite.

htm?zi=1/XJ&sdn=usgovinfo&cdn=newsissues&tm=110&gps=64_261_1276_825&f=00&tt=2&bt=

0&bts=0&zu=http%3A//www.gao.gov/new.items/d0435.pdf

van Leest, P.F. (1997). Persoonlijkheidsmeting bij allochtonen [Assessment of personality for ethnic minori-

ties]. Lisse, The Netherlands: Swets & Zeitlinger.

Zeidner, M. (1986). Are scholastic aptitude tests in Israel biased toward Arab student candidates? Higher Education, 15, 507–522.



http://www.jil.go.jp/english/laborinfo/library/documents/Labor2006_2007.pdf

http://www

http://www.shl.com/globallocations/pages/

http://www.siopsa

http://usgovinfo.about.com/gi/dynamic/offsite

http://www.gao.gov/new.items/d0435.pdf


Taylor & Francis Group 270 Madison Avenue · PDF fileTaylor & Francis Group 270 Madison Avenue...

Documents

Transcript of Taylor & Francis Group 270 Madison Avenue · PDF fileTaylor & Francis Group 270 Madison Avenue...