Personality is important to job performance; meta-analyses published over the years repeatedly showed that self-rated personality traits can significantly predict overall job performance (Barrick & Mount, 1991; Barrick, Mount, & Judge, 2001). Despite their significance, these same meta-analyses, generally showed personality only had a small effect on overall job performance. The exception was conscientiousness, which had a less than medium effect. However, there is also a growing body of evidence suggesting that other-ratings of personality can show higher concurrent validities than self-ratings. Meta-analytic results showed that personality can have a large effect on overall job performance, if the personality traits are rated by others (Connelly & Ones, 2010). Moreover, concurrent validities increased when utilising narrow measures of both personality (Judge, Rodell, Klinger, Simon, & Crawford, 2013) and job performance (Bartram, 2005).
In this study, the author examined the suggestion from meta-analyses that observer-ratings, rather than self-ratings, provide greater explanatory power when predicting job performance. Further, the concurrent validities of using narrow personality traits (facets) as predictors of narrow measures of job performance were investigated.
This study comprised 1,041 participants, of which 92% were employed in a UK police organisation. Employees provided self-ratings and identified two co-workers and a manager who could provide other-ratings of personality and job performance. Online questionnaires measured 71 personality facets of the 11+ Factor Model (Irwing & Booth, 2013) and Bartram’s (2005) Great Eight factors of job performance. Arguably the most comprehensive measure of personality, the 11+ Factor Model is comprised of 11 factors and 74 facets. Items from the International Personality Item Pool (IPIP; Goldberg, 1999) were utilised to create scales for each of the 74 personality facets. A planned missing data design was implemented to improve response rates (Graham, Taylor, Olchowski, & Cumsille, 2006). Measurement models were estimated first, followed by testing of the structural models (J. C. Anderson & Gerbing, 1988) to estimate the combined effects of personality facets on each of the job performance outcomes. Since cross-validation is a powerful approach for evaluating models (Millsap & Meredith, 2007), all models were cross-validated on two datasets.
Fifty-two personality facets were identified and cross-validated. Some of these facets provided superior prediction over factors, when predicting narrow measures of job performance. The facets of integrity, leadership, harm avoidance and empathy explained much of the variance in the Great Eight job competencies. In some cases, self-ratings of personality provided superior prediction over other-ratings.