Retention in the Canadian Forces: The Predictive Validity of Three Selection Measures

Abstract

This study examined the predictive validity of the Canadian Forces Aptitude Test (CFAT), educational levels, and the Military Potential (MP) score in predicting retention in the Canadian Armed Forces (CAF). This set of selection measures is fundamental to CAF selection process. The CFAT is a cognitive test that assesses verbal skills, spatial ability, and problem solving ability. The MP score is derived from candidates' performance at a semi-structured interview and their application materials (biographical information, CFAT scores, and educational achievement). The criterion of retention in this research was basic training completion on first attempt. Analyses were performed on data from 14,381 recruits enrolled between fiscal years 04/05 and 08/09. A series of unadjusted logistic regressions were conducted to examine the effects of each measure individually. Then adjusted logistic regressions were conducted to examine the impact of each measure after controlling for the others and for demographic variables (gender and age). All analyses were conducted separately for officers and Non-Commissioned Member (NCM) recruits. Overall, for officer recruits, results from unadjusted logistic regression indicated that high CFAT scores were positively associated with completion of basic training but higher education levels were negatively associated with completion of basic training. However, these relationships did not hold in adjusted logistic regression where only gender and age emerged as significant. For NCM recruits, results of unadjusted logistic regression showed that CFAT scores, and having a high school diploma or higher degree was positively associated with basic training completion and these relationships remained significant in adjusted logistic regression. A positive relation was also found between MP score and basic training completion for NCM recruits, but these relationships were no longer significant in adjusted logistic regression. More studies are needed to fully understand the relative role of all CAF selection measures to the prediction of other retention and performance-related criteria.