The Predictive Utility Of Mental Health Symptoms On Criminal Behaviors Among Female Offenders

• In Event: Posters I

Abstract

Among female offenders, evidence about the predictive utility of mental health on criminal behaviors is mixed. It may be critical to discern this relationship to design gender responsive prison diversion programs. However, studies examining broad indicators of mental health symptoms may neglect identifying specific symptoms most predictive of future criminal behavior. Thus, the present study examined if broad and specific mental health symptoms predicted criminal behavior among 110 women enrolled in a prison diversion program. Standardized measures of mental health symptoms (anxiety, depression, hostility, interpersonal sensitivity, paranoia, posttraumatic stress disorder, and psychosis) were used to predict self-reported criminal behavior one year post treatment. Logistic regressions indicated that broad indictors of mental health symptoms were not predictive of criminal behavior. Further, logistic regressions indicated that specific mental health symptoms towards the end of treatment minimally improved the prediction of criminal behavior one year post treatment, above and beyond prior mental health and criminal history. Specifically, interpersonal sensitivity predicted self-reported criminal behavior and hostility predicted criminal behavior that came to the attention of authorities. Among female offenders, results suggest that specific symptoms, including a sense of inadequacy or anger, were more useful in predicting future criminal behavior than broad mental health indicators.

Authors

• Merdijana Kovacevic, University of Tulsa

• Chelsea Shotwell Tabke, University of Tulsa

• Madisen Fae Dorand, University of Tulsa

• Ethan Rex, University of Tulsa

• Elana Newman, University of Tulsa