The predictive validity of cognitive ability and personality tests toward police on-the-job training

Abstract

This study investigates the predictive validity of psychometric tests included in the Norwegian Police University College selection process for 106 accepted trainees. Predictor variables were from the Scales (lst, verbal admin, lt-no) cognitive ability tests and the Shapes (basic) personality test, distributed by the contractor Aon Assessment Norway AS (formerly cut-e). The criterion variable was from the on-the-job training phase of the police bachelor program, in which instructors rated trainees’ work behaviors two years after initial selection. The results revealed a lack of predictive validity, as witnessed in both correlational analysis and a regression approach controlling for police academy performance. Contrary to expectations, cognitive ability test scores negatively predicted trainee work behaviors in the regression model. These results are compared with those of other relevant studies and are evaluated in light of this study's main strengths and weaknesses. The results of this study deserve an exploratory interpretation, given the limited sample size and a lack of hitherto peer-reviewed psychometric documentation for the tests. This study substantiates the need for further research with respect to the use of psychometric tests in law enforcement selection, particularly contractor-distributed tests, which are rarely studied.