Reliability and Predictive Validity of the Juvenile Sex Offender Assessment Protocol-II in an Australian Context

This study investigated the reliability and predictive validity of the Juvenile Sex Offender Assessment Protocol-II (J-SOAP-II), in an Australian context, and for Australian Indigenous and non-Indigenous youth. Participants were 123 (n = 91 non-Indigenous; n = 32 Indigenous) youth assessed using the J-SOAP-II and followed up over an average period of 4 years. The reliability of the J-SOAP-II was assessed using measures of internal consistency (Cronbach's alphas and inter-item correlations). Receiver operating characteristic (ROC) analyses were conducted to assess the J-SOAP-II's validity in predicting sexual and non-sexual (violent and non-violent) recidivism. Overall, the J-SOAP-II demonstrated moderate to high levels of reliability and weak to moderate levels of validity for predicting sexual and non-sexual (violent and non-violent) recidivism for the whole sample. While the J-SOAP-II total score was significantly associated with non-sexual (violent and non-violent) recidivism for non-Indigenous youth, it was not significantly associated with any of the outcome variables for Indigenous youth. Across all analyses, the dynamic scales of the J-SOAP-II predicted recidivism with greater accuracy than the static scales. However, the static scales had greater accuracy in predicting the risk of sexual recidivism for Indigenous youth. While a reliable measure, these preliminary findings raise questions about the validity of the J-SOAP-II for predicting sexual recidivism for Indigenous and non-Indigenous youth in an Australian context, and warrant further investigation.