Examining psychopathy as a predictor of recidivism in a sample of women who have sexually offended

Purpose: As there are no validated risk assessment tools to predict recidivism for females who have sexually offended, the current study examines the utility of PCL-R to predict recidivism for this population. Methods: Psychopathy scores were collected from a sample of 244 women that received risk assessments before they were released from prison for a sexual offense.Results: Results indicate that only one woman in the sample sexually recidivated during the follow-up time. However, 40% of the sample recidivated with any type of arrest, including general, violent, and breach offenses. Harrell's C and Cox Regression analyses found significant predictions for the PCL-R total score, factors, and items on overall recidivism and criminal recidivism for the sample, with most of the relationships small to moderate in effect. Additionally, some of the PCL-R items were significant predictors, offering more detailed information about connections to recidivism for this population. However, when control variables were added to multivariate prediction, age and number of total prior arrests negated many significant predictors for the PCL-R.Conclusions: The construct of psychopathy may be part of a risk assessment scheme when assessing risk of women who have sexually offended, especially in relation to items on the PCL-R.