RESUMO
BACKGROUND: Attempts to predict prison recidivism based on the personality have not been very successful. This study aims to provide data on recidivism prediction based on the scores on a personality questionnaire. For this purpose, a predictive model combining the actuarial procedure with a posteriori probability was developed, consisting of the probabilistic calculation of the effective verification of the event once it has already occurred. METHOD: Cuestionario de Personalidad Situacional (CPS; Fernández, Seisdedos, & Mielgo, 1998) was applied to 978 male inmates classified as recidivists or non-recidivists. RESULTS: High predictive power was achieved, with the area under the curve (AUC) of 0.85 (p <.001; Se = 0.012; 95% CI [0.826, 0.873]. The answers to the CPS items made it possible to properly discriminate 77.3% of the participants. CONCLUSIONS: These data indicate the important role of the personality as a key factor in understanding delinquency and predicting recidivism.