International Journal of Selection and Assessment

This paper explores the information-seeking behaviors newcomers engage in relating to their psychological contract and addresses the impact of work values (Autonomy, Advancement, Group Orientation and Economic Rewards) and Work Locus of Control. We propose that these individual characteristics could explain differences in the frequency with which newcomers search for information about the promises their employer has made to them. A two-wave longitudinal study was conducted in which 527 newcomers from eight organizations (representing 3 sectors) participated. The results largely support the proposed relationships between work values and contract-related information seeking, while the relation between Work Locus of Control and contract-related information seeking is rather weak. Implications for psychological contract formation are discussed.

One of the most theoretically developed models of organizational socialization is Van Maanen and Schein's (1979) typology of six tactics. Jones' (1986) operationalization of these tactics has been used in most survey studies of the Van Maanen and Schein model. However, questions remain concerning the dimensionality of the operationalization. Self-report data from business school graduates after 4 months (n = 295) and 10 months (n = 222) on the job indicate that: (1) a 6-factor model better approximates the covariance matrix of the socialization items than do competing 1- and 3-factor models, (2) the 6-factor model better predicts certain work adjustment variables than the 3-factor model, and the 3-factor model better predicts certain variables than the 1-factor model, and (3) the collective, formal, sequential, fixed, serial, and investiture tactics are positively interrelated, suggesting a structured program of early work experience.

James, Demaree, Muliak, and Ladd (1992) outlined a procedure for estimating the mean (M) and variance (V) of true validities (ρ's). This procedure was designed to take into account the potential nonzero intercorrelations among the three artifacts (predictor reliability, criterion reliability, and range restriction) and ρ. The accuracy of this new validity generalization procedure was compared with the accuracy of the Model 2-based (in which correlations are individually corrected for artifacts) procedure because this latter procedure does not require the assumption of uncorrelated artifacts. The current study included two different ρ distributions, three sample sizes, and six different levels of intercorrelations among the three artifacts and ρ. Both procedures yielded relatively accurate estimates of M and V even when the intercorrelations among the three artifacts and ρ were nonzero. The Model 2-based estimates were slightly more accurate than the James et al. estimates, and the accuracy of the Model 2-based estimates was much more stable across sample sizes and different levels of intercorrelatedness. Sample-based artifact data were used in this investigation.

Swailes and Senior’s (1999) examination of the psychometric properties and factor structure of the Learning Styles Questionnaire (LSQ) raised a number of questions regarding the instrument’s construct validity and the relationship between learning style and learning process. Swailes and Senior argued that there may be three learning styles as opposed to the four posited by Honey and Mumford (1986), however, they did not include a three-factor model in their confirmatory factor analysis and hence it is unclear if a three-factor solution represents an improvement on a four-factor solution. Furthermore, the present study draws a distinction between style and process and an argument is presented which contends that the three-factor and four-factor solutions represent process models, as opposed to the style models embodied in Kolb’s theory. This reply attempts to examine two- three- and four-factor solutions for the LSQ. The evidence appears to favour the conventional four-factor model, which may indicate that the LSQ measures individuals’ preferences for each of four stages of an experiential learning process and raises the question of its relationship with style per se.

Amongst the purported advantages of multi-rater feedback systems is their greater accuracy and objectivity. It is argued here that these benefits may be more imagined than real, and that there is no reason to believe such systems will avoid many of the rating errors and distortions found in traditional top-down appraisal. The first study reported in this paper investigates the psychometric soundness and concurrent validity of a pilot 360 degree feedback scheme operated by a multi-national oil company. The second study describes how analysis of the data provided by the first study was used to re-design the rating form, and demonstrates the resulting improvement achieved in the psychometric properties of the 360 degree scheme. It also examined some of the variables that influence rater’s assessment of the target managers. The conclusion is drawn that unless such feedback systems — irrespective of whether they are used for development or for appraisal — are constructed and evaluated along the lines associated with psychometric tests, they may produce assessments that are seriously misleading.

This report highlights some of the challenges faced when using multi-rater feedback around the world. It is based on the observations of coaches who have worked with a variety of managers from different cultures using a 360 degree feedback instrument — the PROFILOR®. Reference is made to some of the research on cultural differences in order to clarify, or otherwise, some of the issues raised.

This article proposes a new theory called the Applicant Attribution-Reaction Theory (AART) to better understand attributional processes in the formation of applicant reactions. The theory proposes that applicants' affective, behavioral, and cognitive reactions, such as fairness, test perceptions, test performance, and motivation, are fundamentally driven by an attributional process. A key implication of the theory is that perceptions such as fairness and test attitudes carry little explanatory power; instead they are consequences of attributional processing. We provide a brief review of dominant applicant reactions frameworks, review the social psychological literature on attributions, and present the theory. We then contrast the theory to existing conceptualizations, and finally describe its potential for better understanding several key topics in applicant reactions, including the justice judgment process, test performance, and racial subgroup differences. The theory has the potential to integrate many diverse perspectives on applicant reactions, and provides numerous directions for future research and practice.

The purpose of this paper is to examine the influence of a job applicant's verbal statements on interviewer ratings of an applicant's interpersonal characteristics. Applicant verbal statements were coded as containing affective, behavioral, and cognitive components. Verbal statements were analyzed using regression analysis and findings indicated that behavioral and cognitive speech components significantly impacted interviewer ratings of an applicant's self-confidence. In turn, interpersonal characteristics of enthusiasm, self-confidence and effectiveness had a significant influence on an interviewer's rating of an applicant's overall quality and the ultimate hiring decision.

Decision-making capabilities are absolutely crucial to a manager. Unfortunately, existing methods of assessing managers in this area for selection and development purposes (in-tray exercises, situational interviews, ability tests, etc.) leave a lot to be desired. This article focuses on an alternative for assessing managerial decision making – the situational inventory – and presents research findings and information on practical applications. Detailed findings are also presented for ‘Scenarios’, the UK’s first published situational measure of managerial judgement. Situational inventories work by presenting participants with realistic but difficult real-life management scenarios. Each scenario is accompanied by a number of possible responses which participants rate for effectiveness in dealing with the scenario. Participants’ ratings are then scored against a set of ideal answers, producing an assessment of current decision-making ability. Decision-making ability can be developed in individuals, making feedback invaluable to participants. Evidence that has been accumulating in the United States and the UK for at least 15 years is presented to support the general situational inventory approach. Additionally, specific evidence is presented for ‘Scenarios’. It was found to correlate significantly with a number of managerial performance and responsibility indicators while appearing to be largely separate from existing psychometric (ability and personality) tests.

A study was conducted to expand the nexus of cognitive and psychomotor abilities. A cognitive aptitude battery and a psychomotor battery were administered to 429 military recruits. A confirmatory factor analysis yielded higher-order factors of general cognitive ability (g) and psychomotor/technical knowledge (PM/TK). PM/TK was interpreted as Vernon's (1969) practical factor (k:m). In the joint analysis of these batteries, g and PM/TK each accounted for about 31% of the common variance. No residualized lower-order factor accounted for more than 7% PM/TK influenced a broad range of lower-order psychomotor factors. The first practical implication of these findings is that psychomotor tests are expected to be at least generally interchangeable. A second implication is that the incremental validity of psychomotor tests beyond cognitive tests is expected to be small. These findings should help guide test developers and inform personnel selecting agencies regarding the expected utility of psychomotor tests.

Use of ability tests in personnel selection is addressed beginning with methodological issues. Studies are reviewed that show that almost all the validity of cognitive tests comes from general cognitive ability, g. Psychomotor ability is reviewed and found to have both higher- and lower-order factors, contrary to long-held beliefs. It was found that the higher-order general psychomotor factor was one source of validity of psychomotor tests. Additionally, psychomotor tests were shown to contain measures of g and to increment the validity of g-based measures very little.

Modern computer technology permits efficient evaluation of test scores in terms of basic orthogonal factors of ability. A three-level hierarchical model of cognitive abilities was used as the theoretical basis of the computerized Swedish Enlistment Battery (CAT-SEB). Structural analysis of ten ability tests on a sample of 1,436 conscripts by confirmatory factor analysis (tested by the LISREL system) revealed a general, a verbal and a spatial factor -- although the determinacy of the latter was weak. A nested factor model was used, with direct influences of the latent variables on the tests. This result is a construct validity evaluation of the testing system. Unrelated factor scores of the three latent variables comprise the output of the testing system. Future research should evaluate the efficiency of the prediction from the latent variables.

The choice of performance rating format may influence employees' fairness perceptions. Participants in two studies, one consisting of 208 participants and the other of 393 participants, evaluated the fairness of common relative and absolute rating formats. The participants in the second study also evaluated the fairness of two rating formats, one absolute and one relative, presented in organizational contexts of varying procedural and distributive justice. Results indicate that not only are absolute formats perceived as more fair than relative formats, but differences in fairness perceptions also occur among relative and absolute formats. Furthermore, it appears that rating format influences procedural justice, especially when outcomes are perceived as fair. Implications for organizations' appraisal practices are discussed.

The purpose of this study was to increase our understanding of applicant perceptions of feedback by drawing upon feedback process models. In Study 1, participants (N=125) completed a personality questionnaire as a first stage of a selection simulation. Results showed that the effect of feedback on attitudes toward the organization was mediated by feedback acceptance. In Study 2, participants (N=252) completed two parallel versions of an in-basket exercise and received informative feedback between the two versions. Results showed that the effect of feedback on subsequent test performance was partially mediated by feedback acceptance. Together, these results highlight the important role of feedback acceptance in selection and suggest new strategies to enhance applicant perceptions in selection.

The selection interview remains the most popular method by which organizations select employees, yet it is also widely criticized for being vulnerable to bias and unfair discrimination. This paper draws upon two specific areas of attribution research: cross-cultural studies of attributional processes and studies of intergroup attributional bias in order to discuss how attributional processes can contribute to unfair discrimination in selection interviews. It is argued that unfair discrimination can arise in two ways: first, as a consequence of ethnocentric attributional biases on the part of interviewers when explaining the behaviour of candidates associated with in-group or out-group status; second, as a result of different patterns of attributions manifest by candidates from diverse cultural groups.

This research describes the personality characteristics of a sample of Canadian public accountants using the 16 Personality Factors Questionnaire. Comparisons are made with the general population, between male and female accountants, and between firms of different sizes. Since success in a public accounting firm is indicated by becoming a partner, the personalities of successful public accountants are determined by comparing partners and staff at other levels. Findings indicate that accountants, as a group, have personality profiles that are significantly different from the general population. Significant differences were found between male and female accountants, and between accountants in large and small firms. Personality appears to be a significant factor in promotion as partners have personality profiles that are significantly different from those of accountants at other levels.

We investigated the predictability of rating level and two measures of rating accuracy from rater Agreeableness (A) and Conscientiousness (C) scores of the Five Factor Model. One hundred and twenty-six students made peer ratings after participating in several group exercises under conditions designed to emulate the modal peer rating system in which raters had low accountability for their ratings. Scores were correlated with average rating level (r=.18, p<.05) and both measures of rating accuracy (p<.05) and C scores were correlated with rating level (−.20, p<.05) and both measures of rating accuracy (p<.05). As suggested by Bernardin, Villanova, and Cooke (Journal of Applied Psychology, 85, 232–236, 2000), raters who were both more agreeable and less conscientious made the most lenient and least accurate ratings. Contrary to Yun, Donahue, Dudley, and McFarland (International Journal of Selection and Assessment, 13, 97–107, 2005), more agreeable raters also tended to rate the least effective performers more leniently than did other raters.

The role of seven variables in predicting supervisor rating accuracy was examined. The study was conducted in natural settings, and the criterion for rating accuracy was the rated salesperson’s productivity in terms of sales. The predictors examined were the supervisor’s gender, the amount of time the supervisor spent working with the subordinate, the length of the supervisor–subordinate relationship, depth of acquaintance, the subordinate’s sensitivity to expressive behavior in others, the ability to modify self-presentation and the subordinate’s age. The participants were 208 supervisors and 268 female salespersons. The results of a discriminant analysis showed that the variables predicting accuracy–inaccuracy were different from those predicting overestimation–underestimation. Accuracy was best predicted by the length of the supervisor–subordinate relationship and by the supervisor’s female gender. The direction of the inaccurate ratings was best predicted by high acquaintance, which pulled the ratings in a positively-biased direction.

A first line of defense for an organization against counterproductive behavior in the workplace is the accurate detection during the interview process of a job candidate’s counterproductive traits. This article discusses aspects of the interview format and other considerations relevant to increasing the accuracy of the assessment of a job candidate’s personality. Recent research suggests that the unstructured interview may be of high value in comparison to the traditional structured interview format, when accurate personality prediction is the criterion (Blackman, in press–a; Ickes, Snyder and Garcia 1997). Interviewers can also become more sophisticated in evaluating the probable accuracy of their judgments by learning about four important moderators of accuracy: properties of the judge, of the target, of the trait being judged, and of the information upon which the judgment is based (Funder 1995).

Research on organizational socialization has usually focused on what organizations do to socialize newcomers or on what newcomers do to socialize themselves. The purpose of this study was to integrate these two perspectives of socialization by investigating the relationships between socialization tactics, newcomers' information acquisition (i.e. feedback and observation) and socialization outcomes. Consistent with previous research, socialization tactics and information acquisition were related to socialization outcomes. Furthermore, institutionalized socialization was positively related to the frequency of newcomers' feedback and observation and the frequency of newcomers' feedback and observation was found to mediate several of the relationships between socialization tactics and outcomes. These results suggest that what newcomers can do to socialize themselves (e.g. acquire information through feedback and observation) is partly a function of what organizations do to socialize newcomers (e.g. the use of specific socialization tactics). The implications of these findings are discussed for the development of a more complete and integrated theory of organizational socialization.

Job search theories traditionally assume that job seekers have open access to information about employment opportunities. This may be moderated by the degree that labor markets rely on external recruitment to convey information about employment opportunities. In particular, external recruitment may be more widely used in some countries than other countries. Accordingly, this paper hypothesizes that job search practices that rely on external recruitment and information sharing with potential recruits are more likely to be widely used in the United States than in China and that job search practices are more likely to influence starting wages in the United States than in China. Using samples of graduating undergraduate college students in both the United States and China, this study suggests that job seekers gather more information about employment opportunities in the United States than in China, and that job search effort is more likely to influence starting wages in the United States than in China. The research suggests that job search theories need to consider job seeker access to information as a significant moderating variable in job search theory.

Standardized tests are commonly used to select individuals in both pre-employment and educational settings. Nevertheless, research has yet to sufficiently explore the relationship between test anxiety (TA) and test performance in selection contexts. The goal of this study was to assess the dimensionality of TA, the relations between TA and test performance, and the impact of gender on TA within the realm of personnel selection. Test anxiety was found to be bi-dimensional and negatively related to performance on a cognitively based selection instrument. Gender moderated the relation between TA and test performance, with stronger associations for males. This finding is consistent with the sex-linked anxiety coping theory that is advanced here. Implications of this study and avenues for future research are discussed.

This research examined how the racial prototypicality of minority job applicants' faces influenced hiring decisions under different affirmative action (AA) policies: no AA, soft AA (recruitment of minorities with merit-based hiring), and hard AA (race as a tie-break factor in hiring). Participants (N=252) evaluated resume/photograph pairs, each containing a Caucasian and a Black applicant, with minority applicants representing three levels of racial prototypicality. The number of jobs awarded to minorities increased as Black racial prototypicality increased. Each level of AA policy strength increased the number of minority hires, but these increases came with a price: AA directives decreased the percentage of minority hires attributed to higher qualifications and increased perceptions that hires were due to AA more than was actually the case.

This study examines the effects that extracurricular activities have on the attributions a résumé reader makes about an applicant. Three characteristics of extracurricular activities (number of activities, holding positions of leadership, and relevance of the activities) were manipulated across 24 résumés of fictitious college graduates. Some 219 raters read the résumés and made judgments about the level of the applicant’s quality. The results showed a main effect for the number of activities, a main effect for holding leadership positions, and a main effect for relevance of the activity. Furthermore, a three–way interaction revealed the differential effects that the relevance of the activities had on the attributions at different levels of leadership and number of activities. A mix of career–related and social activities garnered higher ratings for those who held leadership positions in five activities. However, for those who were leaders in only two activities or were not leaders in five, having only career–related activities earned them higher ratings. For applicants who were not leaders in two activities, relevance of the activities played no role.

The reliance of traditional job analysis on job incumbents as the primary source of work-analytic data is critically examined. It is argued that the sole use of incumbents is practically and theoretically unjustified. The incorporation of non-incumbents to the work analysis process is advocated, especially when abstract human attributes and strategic requirements are evaluated. The time and resource savings afforded by the use of mechanical estimation of work dimensions are also discussed. A revision of traditional formats of data collection in job analysis such as paper-and-pencil surveys and face-to-face interviews is proposed. Instead, the potential work-analytic uses of electronic records of work information nowadays available in electronic performance monitoring systems are outlined.

The purpose of the study was to examine how specific aspects of adaptive testing influence test-takers pre-test reactions. Specifically, three different psychological reactions were examined in the study: perceived fairness of the test, attitude toward the test, and expectations about the test. Fifty-three undergraduates were presented with descriptions of hypothetical selection tests that were manipulated to reflect characteristics of adaptive tests that differ from traditional paper-and-pencil tests. The results indicate that certain features of adaptive tests, such as the inability to skip questions, may adversely impact test-takers' reactions. Implications for test designers are discussed.

887 respondents completed ipsative and normative versions of the PAL-TOPAS personality questionnaire. Data were analysed to test for (1) systematic bias in scores associated with the two response formats and (2) predictors of the magnitude of the discrepancy in the individual's ipsative and normative scores. Discrepancy was assessed for both item responses and scale scores. Sources of biases investigated included ipsative scaling artifact, extremeness of scores on the normative scales and response variability. Results showed that systematic bias in scale scores and magnitude of discrepancy were predicted by different factors. One source of systematic bias was associated with ipsative scaling artifact: the ipsative scales measure both the scale itself and rejection of other alternatives. A second source of systematic bias was acquiescence in response to normative items. A confirmatory factor analysis showed that a good but imperfect fit to the data may be obtained by constructing a structural model of the inter-relationship between normative and ipsative scores which accommodates both sources of bias. The strongest influence on discrepancy in scale scores was extremeness of normative scoring, associated with a bias towards either general acceptance or rejection of trait adjectives. It is concluded that both normative and ipsative response formats have limitations, and it may often be desirable to assess both.

This expaloratory study examines the relationship between personality characteristics (extraversion, core self evaluations), social tie characteristics (number, breadth, depth), and three types of expatriate adjustment (general, interaction, and work). Data was collected at two points in time from 75 expatriate employees from one organization on international assignments around the world. Results indicate that core self-evaluations, but not extraversion, are positively related to the number of ties formed with other expatriates and host country nationals. Social ties with other expatriates were found to provide greater social support, but similar access to information, than those with host country nationals (HCNs). In general, depth and breadth of relationships with other expatriates predicted general and work adjustment; whereas, breadth and total number of relationships with HCNs predicted all three types of adjustment. Overall, these results provide initial support for the importance of social ties in facilitating expatriate adjustment.

This study examined the coachability of two situational judgment tests, the College Student Questionnaire (CSQ) and the Situational Judgment Inventory (SJI), developed for consideration as selection instruments in the college admission process. Strategies for raising scores on each test were generated, and undergraduates were trained in the use of the strategies using a video-based training program. Results indicated that the CSQ was susceptible to coaching. In addition, the scoring format of the CSQ was found to be easily exploited, such that trainees could increase their scores by greater than 1 SD simply by avoiding extreme responses on that test. The results as a whole sounded a note of caution for the potential use of the CSQ in the college admission process.

The Graduate Management Admission Test (GMAT) has been shown to be a valid predictor of Masters of Business Administration (MBA) performance in the USA, but no UK validity studies have been published. This study uses a large sample of UK MBA students to examine the validity and fairness of GMAT. It is found that GMAT-Verbal is a good predictor of MBA examination performance but GMAT-Quantitative is not. It is also found that both components are unfair to native English speakers. The reasons for these findings are to be found in the nature of the criterion employed. Some observations are made regarding the consequences for best practice of the competitive and political context of MBA selection.

The gap between science and practice in personnel selection is an ongoing concern of human resource management. This paper takes Oliver's framework of organizations' strategic responses to institutional pressures as a basis for outlining the diverse economic and social demands that facilitate or inhibit the application of scientifically recommended selection procedures. Faced with a complex network of multiple requirements, practitioners make more diverse choices in response to any of these pressures than has previously been acknowledged in the scientific literature. Implications for the science-practitioner gap are discussed.

Past research has revealed that individuals' job mobility is affected by factors such as job satisfaction, specific career enhancing attributes and job availability. This study examined personality factors predicting voluntary internal and external job mobility. Three types of voluntary job mobility measures were studied: dissatisfaction changes, job improvement changes and job rotations within companies. These mobility measures were related to the Big Five personality factors, sensation seeking and adult attachment. Results showed that demographic variables and sensation seeking contributed to the variance in external job changes. Internal job rotations were not related to any of the demographic and personality variables.

We test the hypothesis that the Raven's Advanced Progressive Matrices has the same construct validity in African university students as it does in non-African students by examining data from 306 highly select 17- to 23-year olds in the Faculties of Engineering and the Built Environment at the University of the Witwatersrand (177 Africans, 57 East Indians, 72 Whites; 54 women, 252 men). Analyses were made of the Matrices scores, an English Comprehension test, the Similarities subscale from the South African Wechsler Adult Intelligence Scale, end-of-year university grades, and high-school grade point average. Out of the 36 Matrices problems, the African students solved an average of 23; East Indian students, 26; and White students, 29 (p<.001), placing them at the 60th, 71st, and 86th percentiles, respectively, and yielding IQ equivalents of 103, 108, and 118 on the 1993 US norms. The same pattern of group differences was found on the Comprehension Test, the Similarities subscale, university course grades, and high-school grade-point average. The items on the Matrices ‘behaved’ in the same way for the African students as they did for the non-African students, thereby indicating the test's internal validity. Item analyses, including a confirmatory factor analysis, showed that the African/non-African difference was most pronounced on the general factor of intelligence. Concurrent validity was demonstrated by correlating the Matrices with the other measures, both individually and in composite. For the African group, the mean r=.28, p<.05, and for the non-African group, the mean r=.27, p<.05. Although the intercepts of the regression lines for the two groups were significantly different, their slopes were not. The results imply that scores on the Raven's Matrices are as valid for Africans as they are for non-Africans.

This paper reviews some of the literature on women in management in the UK, with particular emphasis on the biographical data collected in major studies. In view of the small proportions of women in senior management it seeks to identify the potential barriers that assessment procedures create for women wanting to progress to top positions. As organizations are encouraged to increase the‘fairness’and 'sophistication’of these procedures, they may in fact be reinforcing the very nature of the discriminations they are supposedly attempting to reduce. This paper outlines how this may be happening.

In our rejoinder to the comments of Kehoe (this issue) and Potosky, Bobko and Roth (this issue) we emphasize that our proposal on Pareto-optimal predictor composite formation is a complementary and not a competitive alternative for reducing the tension between selection quality and adverse impact. Our work addresses the decisions to be made once one has decided to use a predictor composite. We also further clarify the basic features of Pareto-optimal tradeoffs and Pareto-optimal composites within the context of personnel selection. In particular, we indicate that Pareto-optimal tradeoffs between validity and adverse impact emerge because these goals are different and not because of any dualism between them.

Previous research on the validity and adverse impact (AI) of predictor composite formation focused on the merits of regression-based or ad hoc composites. We argue for a broader focus. Ad hoc chosen composites are usually not Pareto-optimal, whereas the regression-based composite represents only one element from the total set of Pareto-optimal composites and can, therefore, provide only limited information on the potential for validity and AI reduction of forming predictor composites when both validity and AI are of concern. In that case, other Pareto-optimal composites may provide a better benchmark to decide on the merits of the predictor composite formation. We summarize a method to determine the set of Pareto-optimal composites and apply the method to a representative collection of selection predictors. The application shows that the assessment of the AI and validity of predictor composite formation can differ substantially from the one arrived at when considering only regression-based composites.

This article is focused on the evidence related to the criterion and construct validity of interview, its adverse impact on minority groups and the applicant reactions. Based on the content of the questions included in personnel interview, two types of interview made by found: conventional structured interview and behavioural structured interview. With regard to criterion validity, evidence shows that, in general, interviews may be used to predict job performance and, specifically, behavioural structured interviews show the highest validity coefficients. Although construct validity was less well investigated, there is currently a number of studies carried out in order to clarify what constructs are assessed by interviews. It was shown that conventional structured interviews and behavioural structured interviews clearly measure different constructs. With regard to group differences, interviews show only a small adverse impact, but this impact decreases if behavioural structured interviews are used. In connection with applicant reactions, more negative applicant reactions appear with behavioural structured interviews than conventional structured interviews. Practical implications and future lines of research are suggested.

This paper offers an organizational perspective about the Pareto-optimality of trade-offs between selection quality and adverse impact reduction described by De Corte, Lievens and Sackett. Based on considerations of culture and human resource strategy, the need to understand the impact of any trade-off, and a desire to compare this approach to alternatives, it is concluded that organizations may be strategically disinclined to implement such trade-offs and, if interested, organizations would desire more information than is available from the accumulated research in this domain. In particular, validity differences between Pareto-optimal composites are not likely to be good indicators of selection quality differences and the comparative effectiveness of alternative approaches is not clear.

Mean subgroup (gender, ethnic/cultural, and age) differences are summarized across studies for several predictor domains – cognitive ability, personality and physical ability – at both broadly and more narrowly defined construct levels, with some surprising results. Research clearly indicates that the setting, the sample, the construct and the level of construct specificity can all, either individually or in combination, moderate the magnitude of differences between groups. Employers using tests in employment settings need to assess accurately the requirements of work. When the exact nature of the work is specified, the appropriate predictors may or may not have adverse impact against some groups. The possible causes and remedies for adverse impact (measurement method, culture, test coaching, test-taker perceptions, stereotype threat and criterion conceptualization) are also summarized. Each of these factors can contribute to subgroup differences, and some appear to contribute significantly to subgroup differences on cognitive ability tests, where Black–White mean differences are most pronounced. Statistical methods for detecting differential prediction, test fairness and construct equivalence are described and evaluated, as are statistical/mathematical strategies for reducing adverse impact (test-score banding and predictor/criterion weighting strategies).

The Flemish Admission Exam ‘Medical and Dental Studies’ is comprised of four cognitive ability tests and four situational tests, namely two work samples (i.e., a lecture and a medical text) and two video-based situational judgement tests (i.e., a physician–patient interaction and a medical expert discussion). On the basis of the Admission Exam scores of 941 candidates (359 men, 582 women) this study shows that situational tests significantly can predict better than cognitive ability tests, with lecture and text emerging as significant predictors. When situational tests are combined with cognitive ability tests, there are no mean gender differences. Situational tests also enable us to measure a broader range of constructs. For example, in this study, the personality factor Openness is related to better situational test performance. Overall, this study demonstrates that situational tests may be a useful complement to traditional student selection procedures.

Although there has been empirical attention paid to the criterion-related validity of predictor composites, there has been much less attention paid to the standardized ethnic group differences associated with these composites. One important area of inquiry in predictor composite research is the influence of adding predictors to a test of general mental ability. The limited empirical literature on this practice is mixed, but the prevailing expectation is that there is likely to be higher validity and less adverse impact. Unfortunately, much of the previous work is limited by the presence of inaccurate validity and standardized ethnic group difference values. In this analysis we formed meta-analytic matrices to more accurately estimate the validity and standardized ethnic group differences of several composites that combine a measure of cognitive ability with measures of conscientiousness, a structured interview, or biodata. While results were somewhat complex, we found that adding alternative predictors does not result in a situation in which validity automatically goes up and adverse impact potential automatically goes down. In fact, the reductions in adverse impact (if any) from adding “non-cognitive” predictors were more modest than much of the literature suggests.

The paper proposes a procedure for designing Pareto-optimal selection systems considering validity, adverse impact and constraints on the number of predictors from a larger subset that can be included in an operational selection system. The procedure determines Pareto-optimal composites of a given maximum size thereby solving the dual task of identifying the predictors that will be included in the reduced set and determining the weights with which the retained predictors will be combined to the composite predictor. Compared with earlier proposals, the simultaneous consideration of both tasks makes it possible to combine several strategies for reducing adverse impact in a single procedure. In particular, the present approach allows integrating (a) investigating a large number of possible predictors (such as multitest battery of ability tests, or a collection of ability and nonability measures); (b) explicit predictor weighting within feasible test procedures of a given limited size.

There is a wealth of advice offered by ‘experts’ in HRM concerning the content of the effective job advertisement. This study considers whether following such advice would provide the potential job applicant with the information they want and questions the extent to which employers actually heed this advice. It firstly compares the normative guidelines with what potential applicants' for managerial posts consider to be important. A second strand to the analysis examines the extent to which advertising practice mirrors the requirements of the job-seeker market. The results give some solace to recruiters but also demonstrates that there is a long way to go before job advertisements achieve the content that job-seekers require.

Two measures of job performance of 301 employees of a manufacturing enterprise in Québec (the criteria) were used to assess the stability of predictive validity over time of two specific intelligence tests, OTIS and BETA, whilst the moderating effects of two other variables, tenure and job complexity, were also examined. The predictive validity of the tests was not time-based but when job complexity was considered, the validity increased in relation to the complexity of the job.

Performance on the Abbreviated Reid Report, a pre-employment integrity attitude scale designed to predict counterproductivity, was examined in three countries. Some 889 job applicants and current employees in Argentina, Mexico and South Africa completed the scale, with appropriate language translations. Supervisors provided performance evaluations for the majority of employees on the dimensions of counterproductive behavior, general work performance, social interaction and positive employee traits. Comparisons of mean scores and reliability coefficients indicated comparable responses to the scale across cultures and with US samples. Reid scale scores significantly predicted supervisor evaluations in all three countries. Comparisons were also made between employee’s admissions of acts of counterproductivity and supervisor perceptions of such acts. Implications for cross-cultural integrity testing and future research are discussed.