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
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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.
TAF-Y100248_FARR-09-0801-C003.indd 52TAF-Y100248_FARR-09-0801-C003.indd 52 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
TAF-Y100248_FARR-09-0801-C003.indd 53TAF-Y100248_FARR-09-0801-C003.indd 53 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
54 Handbook of Employee Selection
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
TAF-Y100248_FARR-09-0801-C003.indd 54TAF-Y100248_FARR-09-0801-C003.indd 54 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Validation Strategies for Primary Studies 55
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
TAF-Y100248_FARR-09-0801-C003.indd 55TAF-Y100248_FARR-09-0801-C003.indd 55 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
56 Handbook of Employee Selection
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,
TAF-Y100248_FARR-09-0801-C003.indd 56TAF-Y100248_FARR-09-0801-C003.indd 56 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Validation Strategies for Primary Studies 57
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).
TAF-Y100248_FARR-09-0801-C003.indd 57TAF-Y100248_FARR-09-0801-C003.indd 57 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
58 Handbook of Employee Selection
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.
TAF-Y100248_FARR-09-0801-C003.indd 58TAF-Y100248_FARR-09-0801-C003.indd 58 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Validation Strategies for Primary Studies 59
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
TAF-Y100248_FARR-09-0801-C003.indd 59TAF-Y100248_FARR-09-0801-C003.indd 59 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
60 Handbook of Employee Selection
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.
TAF-Y100248_FARR-09-0801-C003.indd 60TAF-Y100248_FARR-09-0801-C003.indd 60 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Validation Strategies for Primary Studies 61
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.
TAF-Y100248_FARR-09-0801-C003.indd 61TAF-Y100248_FARR-09-0801-C003.indd 61 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
62 Handbook of Employee Selection
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
TAF-Y100248_FARR-09-0801-C003.indd 62TAF-Y100248_FARR-09-0801-C003.indd 62 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Validation Strategies for Primary Studies 63
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
TAF-Y100248_FARR-09-0801-C003.indd 63TAF-Y100248_FARR-09-0801-C003.indd 63 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
64 Handbook of Employee Selection
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
TAF-Y100248_FARR-09-0801-C003.indd 64TAF-Y100248_FARR-09-0801-C003.indd 64 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Validation Strategies for Primary Studies 65
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)
TAF-Y100248_FARR-09-0801-C003.indd 65TAF-Y100248_FARR-09-0801-C003.indd 65 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
66 Handbook of Employee Selection
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.
TAF-Y100248_FARR-09-0801-C003.indd 66TAF-Y100248_FARR-09-0801-C003.indd 66 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Validation Strategies for Primary Studies 67
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.
TAF-Y100248_FARR-09-0801-C003.indd 67TAF-Y100248_FARR-09-0801-C003.indd 67 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
68 Handbook of Employee Selection
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.
Personnel Psychology, 43, 313–334.
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.
TAF-Y100248_FARR-09-0801-C003.indd 68TAF-Y100248_FARR-09-0801-C003.indd 68 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Validation Strategies for Primary Studies 69
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.
Journal of Applied Psychology, 75, 297–300.
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.
TAF-Y100248_FARR-09-0801-C003.indd 69TAF-Y100248_FARR-09-0801-C003.indd 69 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
70 Handbook of Employee Selection
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.
Personnel Psychology, 40, 455–488.
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.
Personnel Psychology, 43, 231–378.
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.
TAF-Y100248_FARR-09-0801-C003.indd 70TAF-Y100248_FARR-09-0801-C003.indd 70 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Validation Strategies for Primary Studies 71
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.
Personnel Psychology, 53, 901–912.
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.
Journal of Applied Psychology, 71, 461–468.
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.
TAF-Y100248_FARR-09-0801-C003.indd 71TAF-Y100248_FARR-09-0801-C003.indd 71 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
TAF-Y100248_FARR-09-0801-C003.indd 72TAF-Y100248_FARR-09-0801-C003.indd 72 1/20/10 5:11:07 PM1/20/10 5:11:07 PM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
652 Handbook of Employee Selection
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
TAF-Y100248_FARR-09-0801-C030.indd 652TAF-Y100248_FARR-09-0801-C030.indd 652 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
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
TAF-Y100248_FARR-09-0801-C030.indd 653TAF-Y100248_FARR-09-0801-C030.indd 653 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
654 Handbook of Employee Selection
TABLE 30.1 (continued)Disadvantaged Groups Within Each Country
Country Group Percentage of Population
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
TAF-Y100248_FARR-09-0801-C030.indd 654TAF-Y100248_FARR-09-0801-C030.indd 654 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Perspectives From Twenty-Two Countries on the Legal Environment for Selection 655
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
Country Group Percentage of Population
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.
TAF-Y100248_FARR-09-0801-C030.indd 655TAF-Y100248_FARR-09-0801-C030.indd 655 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
656 Handbook of Employee Selection
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.
TAF-Y100248_FARR-09-0801-C030.indd 656TAF-Y100248_FARR-09-0801-C030.indd 656 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Perspectives From Twenty-Two Countries on the Legal Environment for Selection 657
(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,
TAF-Y100248_FARR-09-0801-C030.indd 657TAF-Y100248_FARR-09-0801-C030.indd 657 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
658 Handbook of Employee Selection
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
TAF-Y100248_FARR-09-0801-C030.indd 658TAF-Y100248_FARR-09-0801-C030.indd 658 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Perspectives From Twenty-Two Countries on the Legal Environment for Selection 659
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
TAF-Y100248_FARR-09-0801-C030.indd 659TAF-Y100248_FARR-09-0801-C030.indd 659 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
660 Handbook of Employee Selection
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.
TAF-Y100248_FARR-09-0801-C030.indd 660TAF-Y100248_FARR-09-0801-C030.indd 660 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Perspectives From Twenty-Two Countries on the Legal Environment for Selection 661
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
TAF-Y100248_FARR-09-0801-C030.indd 661TAF-Y100248_FARR-09-0801-C030.indd 661 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
662 Handbook of Employee Selection
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.
TAF-Y100248_FARR-09-0801-C030.indd 662TAF-Y100248_FARR-09-0801-C030.indd 662 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Perspectives From Twenty-Two Countries on the Legal Environment for Selection 663
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
TAF-Y100248_FARR-09-0801-C030.indd 663TAF-Y100248_FARR-09-0801-C030.indd 663 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
664 Handbook of Employee Selection
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.
TAF-Y100248_FARR-09-0801-C030.indd 664TAF-Y100248_FARR-09-0801-C030.indd 664 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Perspectives From Twenty-Two Countries on the Legal Environment for Selection 665
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
TAF-Y100248_FARR-09-0801-C030.indd 665TAF-Y100248_FARR-09-0801-C030.indd 665 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
666 Handbook of Employee Selection
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
TAF-Y100248_FARR-09-0801-C030.indd 666TAF-Y100248_FARR-09-0801-C030.indd 666 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Perspectives From Twenty-Two Countries on the Legal Environment for Selection 667
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
TAF-Y100248_FARR-09-0801-C030.indd 667TAF-Y100248_FARR-09-0801-C030.indd 667 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
668 Handbook of Employee Selection
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.
TAF-Y100248_FARR-09-0801-C030.indd 668TAF-Y100248_FARR-09-0801-C030.indd 668 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Perspectives From Twenty-Two Countries on the Legal Environment for Selection 669
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
TAF-Y100248_FARR-09-0801-C030.indd 669TAF-Y100248_FARR-09-0801-C030.indd 669 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
670 Handbook of Employee Selection
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.
TAF-Y100248_FARR-09-0801-C030.indd 670TAF-Y100248_FARR-09-0801-C030.indd 670 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Perspectives From Twenty-Two Countries on the Legal Environment for Selection 671
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.
TAF-Y100248_FARR-09-0801-C030.indd 671TAF-Y100248_FARR-09-0801-C030.indd 671 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
672 Handbook of Employee Selection
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
TAF-Y100248_FARR-09-0801-C030.indd 672TAF-Y100248_FARR-09-0801-C030.indd 672 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Perspectives From Twenty-Two Countries on the Legal Environment for Selection 673
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).
TAF-Y100248_FARR-09-0801-C030.indd 673TAF-Y100248_FARR-09-0801-C030.indd 673 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
674 Handbook of Employee Selection
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.
TAF-Y100248_FARR-09-0801-C030.indd 674TAF-Y100248_FARR-09-0801-C030.indd 674 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
Perspectives From Twenty-Two Countries on the Legal Environment for Selection 675
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.
TAF-Y100248_FARR-09-0801-C030.indd 675TAF-Y100248_FARR-09-0801-C030.indd 675 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373
676 Handbook of Employee Selection
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.
TAF-Y100248_FARR-09-0801-C030.indd 676TAF-Y100248_FARR-09-0801-C030.indd 676 1/22/10 10:58:54 AM1/22/10 10:58:54 AM
http://www.workpsychologyarena.com/handbook-of-employee-selection-9780805864373