Predicting Violent and Any Recidivism Among Individuals With a Sexual Offence History: A Comparison of Nine Risk Assessment Scales

Risk assessments for individuals with a history of sexual offending should consider the risk for violence and any recidivism, given that these reoffending outcomes are more likely to occur compared to sexual recidivism. The current study examined the ability of nine risk assessment scales (Static-99R, Static-2002R, BARR-2002R, PCL-R, PCL:SV, SORAG, VRAG, VRAG-R, and SIR) to predict violent and any recidivism among individuals with a history of sexual offending over 5-, 10-, and 15-year follow-up periods ( N =  368). We also examined the incremental predictive validity of key underlying constructs contributing to recidivism risk in this population: general criminality, sexual criminality, and age. Results indicated that all nine scales significantly predicted violent and any recidivism with moderate-to-large effect sizes (Area Under the Curves ranging from 0.643 to 0.782). Both general criminality and age added incrementally to the prediction of both outcomes; sexual criminality was unrelated to violent and any recidivism after controlling for the other constructs. The scales predicted similarly well across those with adult or child victims (or both). When assessing individuals with a history of sexual offending for the risk of violent or any recidivism, evaluators therefore have numerous defensible options to choose from. Further research is needed to understand the underlying constructs contributing to risk and the application of these tools to more diverse samples.