Document Type



Clinicians are often called upon to estimate the level of risk of a juvenile sex offender to sexually reoffend. Risk assessment evaluations often employ measures utilizing structured professional judgment to weigh empirically supported risk factors associated with recidivism in order to determine an overall level of risk. However, there is a lack of empirically validated risk assessment measures specifically developed for adolescent sex offenders. This study will examine the ability of the Estimate of Risk of Adolescent Sexual Offender Recidivism (ERASOR) to predict sexual and nonsexual reoffending. The archived records of 100 juvenile sex offenders, who had previously been the subject of a risk assessment by a forensic psychologist, were rated on the ERASOR and were coded for other relevant historical and clinical information. The predictive validity of the ERASOR was examined using Receiver Operating Characteristic (ROC) and Cox regression analyses. Results indicate that the ERASOR did not accurately predict sexual recidivism and had limited ability to accurately predict nonsexual or general recidivism. Implications on the future use of sexual recidivism specific risk assessment tools are discussed.