Finding appropriate assessment tools to predict recidivism is a difficult aim, which may lead to actions with unintended consequences. Aims don't have consequences. At times, the research has been used to justify penalising reoffenders with punitive measures rather than treating them with effective psychological interventions. This study aims to contribute to untangling and assessing the potential predictors of reoffender drivers. In this study, 296 drivers: 86 reoffenders (7 women and 79 men) and 206 non-reoffenders (105 women and 101 men) responded to a battery of assessment questionnaires in which they were asked for demographic data (i.e. gender and age), alcohol consumption habits, driving styles, general estimation of risk in everyday life, sensitivity to reward and punishment and anger while driving. The results provided a logistical regression model capable of predicting reoffending and explaining 34% of variability, successfully classifying 77.6% of participants. In this model, the best predictor of reoffending is higher consumption of alcohol (Alcohol Use Disorders, AUD), followed by incautious driving (since cautious driving style correlates negatively with reoffending) and to a lesser extent, infraestimation of recreational risk and a greater sensitivity to reward. Relying on results to predict recidivism could be important to plan better interventions to prevent it.
Keywords: Anger; Driving under the influence (DUI); Driving while intoxicated (DWI); Recidivist; Risk perception; Road safety; Sensitivity to punishment; Sensitivity to reward.
Copyright © 2018 Elsevier Ltd. All rights reserved.