The Bright and Dark Sides of Personality: Implications for Personnel Selection in Individual and Team Contexts

Abstract

That personality has shown itself relevant to individual attitudes and behavior, and to team and organizational functioning, seems an incontrovertible statement. Barrick and Mount (2005: 361) flatly state: ‘Personality traits do matter at work’ and indeed the data appear to support their conclusion (Hogan, 2005). Barrick et al. (2001) analyzed extant meta-analyses on relationship between the ‘big five’ personality traits and job performance, finding a multiple correlation of R 0.47 when the five traits were used to predict overall job performance. Other large-scale reviews have linked personality to job satisfaction (Judge et al., 2002a), leadership (Judge et al., 2002b), workplace deviance (Salgado, 2002), well-being (DeNeve and Cooper, 1998), and organizational commitment (Erdheim et al., 2006). However, skeptics remain. One line of criticism argues that whilst personality has nonzero associations with important criteria, the effect sizes are small. In arguing that little has changed since Guion and Gottier’s (1965) influential (and pessimistic) review, Schmitt (2004: 348) observed, ‘The observed validity of personality measures, then and now, is quite low even though they can account for incrementally useful levels of variance in work-related criteria beyond that afforded by cognitive ability measures because personality and cognitive ability measures are usually minimally correlated’. Hogan (2005) takes issue with overall assessment, while also arguing that the validity of personality measures is often underestimated by failing to account for poor measures, the source of personality ratings (self versus observer), and the situationally specific nature of performance. He concludes, ‘The bottom line is, personality measures work pretty well, especially when compared with all the other measures’ (p. 340). Our own view is that whereas it is true that the validities for personality variables cannot be labeled as strong using the Cohen (1977) effect size conventions, the same is true of virtually any meaningful predictor of broad, complex criteria such as job performance. For example, there is perhaps no theory in organizational behavior more respected for its validity than goal-setting theory. Locke and Latham (2002: 714) concluded, ‘Goal-setting theory is among the most valid and practical theories of employee motivation in organizational psychology’ and Miner (2003) found that organizational behavior scholars ranked goal-setting theory as the most important of all (73 were rated) management theories. Yet meta-analyses have revealed that the overall validity of goal difficulty in predicting job performance is dc 0.577 (Wood et al., 1987), which translates into a correlation of Rc 0.277. This differs little from the overall validity of conscientiousness (Rc 0.23) or core self-evaluations (Rc 0.23) in predicting job performance. When one considers the constellation of traits, the validity is much higher (Rc 0.47, as noted above).