Abstract:
The current research investigates the psychometric properties of three personality inventories: two five-factor measures (the Trait Self-Descriptive Inventory; TSD; Christal, 1994, and the NEO-Five Factor Inventory; NEO-FFI; Costa & McCrae, 1992) and one six-factor measure (the HEXACO Personality Inventory; Lee & Ashton, 2004). The data from a sample of Canadian Forces (CF) recruits (N=515) and officer candidates (N=124) attending initial military training were analyzed via confirnlatory factor analysis and hierarchical regression to assess model fit and the ability of the personality measures to explain variance in several military job performance criteria (i.e., job performance, including leadership for officers; contextual performance; and counterproductive workplace behaviours) beyond the variance accounted for by cognitive ability. Results of confirmatory factor analyses were inconsistent for the TSD versions and NEO-FFI, and the HEXACO was generally a poor fitting model. The factors of all personality inventories demonstrated good internal consistency and convergent validity with their respective factors on the other measures. In hierarchical regression analyses, the Conscientiousness factors generally predicted significant incremental variance in job performance criteria for recruits, but not for officers. For officers, Extraversion predicted supervisor assessments of leadership. Recommendations for the implementation of personality testing in the CF are discussed, as are limitations and directions for future research.