On the use of personality measures in personnel selection.

P. GREGORY IRVINGThe University of Western OntarioAbstractAlthough the use of personality as a predictor in personnel selection has not met with substantial success in the past, recent evidence has suggested that personality measures are related to performance criteria which are unrelated to cognitive ability when the traits measured are conceptually related to these criteria. It seems that personality measures may predict job performance dimensions which cannot be predicted by cognitive ability measures. The use of personality measures in personnel selection may be warranted when a careful job analysis is undertaken to determine which performance dimensions may be related to personality traits. As early as 1923, Freyd (cited in Guion, 1983) recognized that certain steps must be undertaken to ensure that the personnel selection procedures used by organizations are valid. These steps included conducting a job analysis to determine the characteristics which led to success or failure on the job, designating a single - measure criterion of success, developing an exhaustive list of abilities required for success, finding or devising a measure of these abilities, and statistically comparing the test scores with the criterion scores. Freyd's steps continue to represent sound personnel selection practises in major Industrial/Organizational psychology textbooks (e.g., Cascio, 1987; Landy, 1989).Despite the fact that Freyd outlined these steps some 70 years ago, researchers investigating the relationship between personality and job performance have tended to ignore his advice. Early attempts to use personality traits to predict various job criteria have generally used a shotgun approach in which a large number of scales were correlated with a large number of criteria. Such an approach has been used to predict accident rates of truck drivers (Parker, 1953) and job satisfaction of farmers (Brayfield & Marsh, 1957). The predictor in both studies was the clinical scales of the Minnesota Multiphasic Personality Inventory (MMPI). The problem with this empirical approach to predicting job - related criteria is that one may expect at least some relationships to be significant by chance alone and attempts at cross - validation would likely result in substantial shrinkage in the validity coefficients. The assessment of personality has been a controversial topic in personnel selection. Over the past several decades, a number of literature reviews have been conducted resulting in conflicting viewpoints regarding the predictability of job performance based on personality traits (Ghiselli & Barthol, 1953; Guion & Gottier, 1965; Schmitt, Gooding, Noe & Kirsch, 1984; Tett, Jackson & Rothstein, 1991). Although Guion and Gottier (1965) stated that "there is no generalizable evidence that personality measures can be recommended as good or practical tools for employee selection" (p. 159), they observed the importance of predicting job criteria which are unrelated to cognitive ability. Without exception, these reviews have cautioned against the shotgun approach to prediction which has plagued previous attempts to validate personality measures in personnel selection. Hollenbeck and Whitener (1988) have suggested that one of the reasons for the poor predictive ability of personality variables in previous studies is that many of the validation studies lacked statistical power (see Schmidt, Hunter, & Urry, 1976, for a discussion of the lack of power in validation studies). As evidence, they point to the fact that in Guion and Gottier's (1965) review, 62 of 100 validation studies had sample sizes of less than 84.Personality assessments have long been used as a part of the selection process for several professions including police officers (Burbeck & Furnham, 1985; Inwald & Shusman, 1984), flight attendants (Ferris, Bergin & Gilmore, 1986), and firefighters (Johnson, 1983). Much of the recent research involving personality measures in personnel selection, however, has employed such measures in screening for psychological problems (e. …