Predicting job performance in correctional officers with pre-employment psychological screening

There is substantial cost in the hiring and training of a correctional officer, with a high rate of turnover compounding these costs. While pre-employment psychological screening is suggested as one method to prevent these losses, mandates to screen are not as common in corrections as they are in law enforcement. Further, minimal research has examined the validity of psychological testing in correctional officers. This dissertation examined pre-employment psychological screening for 421 correctional officers hired by one of three upstate New York sheriff's departments. Assessments were conducted by Public Safety Psychology, PLLC from March, 1997 to June, 2012. T scores and risk estimates from the California Psychological Inventory (CPI) and Personality Assessment Inventory (PAI), DQ admission and problem points from the Personal History Questionnaire (PHQ) and Psychological History Questionnaire (PsyQ) and the psychologist's recommendation were used as predictors of supervisor rating and job status. Utilizing logistic regression and controlling for agency of hire, high ratings by the psychologist, high scores on PAR-H and low scores on BOR-S from the PAI were associated with satisfactory supervisor ratings. Multinomial logistic regression revealed that being non-White, having a lower rating by the psychologist, higher To and Ai scores and lower So scores from the CPI, and more General problem points on the PsyQ were predictive of officers who were fired compared to being currently employed. Furthermore, previous law enforcement experience, being younger, lower Gi, So and Wo scores on the CPI, higher To and Sc scores on the CPI, and lower probability of substance abuse issues as based on the PAI and PHQ were predictive of officers quitting rather than staying on the job. Limitations and future directions are discussed.