The New Jersey Department of Corrections (NJDOC) has been awarded a $400,000 grant to assess the effectiveness of sex offender classification categories in predicting sex offender recidivism over time and over sub-samples across six states: Florida, Massachusetts, Minnesota, New Jersey, North Carolina, and South Carolina.
Specifically, the study will examine the accuracy of sex offender categories/classifications, as defined within the 2006 Sex Offender Notification and Registration Act (SORNA), in predicting sex offense recidivism compared to an actuarial risk assessment instrument (STATIC-99) created for the same purpose.
This multi-state study will be among the first to investigate procedures commonly used to classify risk in American sex offenders. The project was developed by NJDOC researcher Dr. Kristen Zgoba in collaboration with University of Minnesota Medical School, Lynn University, Brandeis University, Sand Ridge Secure Treatment Center and Medical University of South Carolina. In addition, a consultant from the Association for the Treatment of Sexual Abusers with specialization in SORNA’s sex offender tiering and the Adam Walsh classification system will assist with the project.
Using a quasi-experimental research design, designated state tier levels will be recorded for sex offenders, and all offenders will be classified using SORNA and STATIC 99 criteria relative to sex offender recidivism risk. Comparisons between the predictive accuracy of the SORNA levels and the STATIC-99 will be completed both within and collapsed across states.
Specific research areas that will be addressed in this study include:
- comparisons of STATIC-99 with extant tier levels in the state and the determination of level by SORNA criteria;
- effectiveness of two states’ tier levels and resulting community notification levels in predicting recidivism;
- the frequency of STATIC-99 and STATIC-2002 risk levels in U.S. samples, and the recidivism rates associated with the risk levels;
- the contribution of offender age (during various points in the criminal justice system process) to the predictive accuracy of recidivism;
- the overall reliability of Static-99 and Static-2002 risk assessment instruments; and
- denial, among other variables, as potential risk predictors of recidivism.
This project is expected to begin in the upcoming weeks.
