NCJ Number
244872
Date Published
May 2011
Length
67 pages
Annotation
This study examined the effectiveness of the Structured Assessment of Violence Risk in Youth (SAVRY) in predicting sexual and non-sexual recidivism in a sample of 100 male juvenile sex offenders.
Abstract
The study found that the SAVRY total score and overall SAVRY risk rating, along with several of the subscales, significantly predicted general and non-sexual recidivism for this sample. Significance was not found for sexual reoffending, except for scores on the SAVRY historical risk factors subscale. These findings suggest that juvenile sex offenders be considered as a smaller subgroup of a larger delinquent population, rather than as a unique population of juvenile offenders. Archival case information, which contains forensic reports, was used to score the SAVRY. The Area Under the Receiver Operating Characteristic Curve (ROC) and Cox regression were used to analyze the predictive validity of SAVRY risk ratings. Appendixes contain the SAVRY Scoring Form and the Forensic Evaluation Data Sheet. 5 tables and 45 references