Risk Factors for Adolescent Sex Offender Recidivism: Evaluation of Predictive Factors and Comparison of Three Groups Based Upon Victim Type

Abstract

This study investigated differences in recidivism risk factors and traits associated with psychopathy among 3 groups of male adolescent sexual offenders (N=156): offenders against children, offenders against peers or adults, and mixed type offenders. Furthermore, those same variables were examined for their association with sexual and nonsexual recidivism and the 3 groups were compared for differences in rates of recidivism. Based upon both juvenile and adult recidivism data, 6.4% of the sample reoffended sexually and 30.1% reoffended nonsexually. Retrospective risk assessments were completed using the Juvenile Sex Offender Assessment Protocol-II (JSOAP-II) and the Psychopathy Checklist:Youth Version (PCL:YV). Comparisons of the 3 preexisting groups for differences on scale and factor scores were conducted using analyses of variance (ANOVAs). Differences among groups for recidivism were measured using survival curve analysis. Associations between risk scales and recidivism were measured using Cox regression analyses. Results suggest significant differences among the 3 offender groups on multiple scales of the JSOAP-II and PCL:YV, with mixed type offenders consistently producing higher risk scores as compared to those who exclusively offend against children or peers/adults. The Impulsive/Antisocial Behavior scale of the JSOAP-II and the Interpersonal and Antisocial factors of the PCL:YV were significant predictors of sexual recidivism. The Behavioral and Antisocial factors of the PCL:YV were significant predictors of nonsexual recidivism. Results supported previous research indicating that most adolescents who sexually offend do not continue offending into adulthood. Such results can lead to improved treatment by targeting specific risk factors for intervention and better use of risk management resources in the community, while preserving the most restrictive treatment options for the highest risk offenders.

An examination of the Sex Offender Registration and Notification Act as applied to juveniles: Evaluating the ability to predict sexual recidivism.

“…Similarly, in a study of 169 male juvenile sex offenders, Viljoen and her colleagues (2008) reported that the J-SOAP-II scale scores did not predict sexual aggression during treatment or during a 6.6-year follow-up. In contrast, in a study of 156 sex-offending adolescent males, Parks and Bard (2006) found that one of the two components of the Static Summary scale (the Impulsive/Antisocial Behavior scale) significantly predicted sexual recidivism but the Dynamic Summary scale components were unrelated to sex offending.…”

Predictive validity of adult risk assessment tools with juveniles who offended sexually.

“…In some studies, one or more subscales/scales have been shown to exceed chance-level prediction for general and sexual recidivism (e.g., Martinez, Rosenfield, & Flores, 2007;Parks & Bard, 2006;Prentky et al, 2000;Prentky et al, 2010), while in other studies, they have not (e.g., Caldwell et al, 2008;Hecker, Scoular, Righthand, & Nagle, 2002;Waite et al, 2005;Viljoen et al, 2008). Possibly the reason for this mixed evidence has to do with the nature of the samples used to assess the predictive accuracy, namely, the use of small and unrepresentative samples in some studies.…”

Childhood Sexual Abuse in Adolescents Adjudicated for Sexual Offenses: Mental Health Consequences and Sexual Offending Behaviors

“…An advanced graduate research assistant with training in risk assessment and a staff psychologist coded the interview and records for the purposes of this study. Internal consistency on the subscales was between .64 and .95, and the measure has been shown to demonstrate concurrent validity with parallel measures and sexual offense information (Parks & Bard, 2006;Righthand et al, 2005). Although the extant literature on the utility of JSOAP-II scores in predicting sexual recidivism is inconsistent, a meta-analysis by Viljoen, Mordell, and Beneteau (2012) found that aggregated correlations for total scores on the JSOAP-II significantly predicted sexual reoffending.…”