Comparison of Risk Assessment Instruments in Juvenile Justice

• NCJ Number: 244477
• Author(s): Chris Baird; Theresa Healy; Kristen Johnson Ph.D.; Andrea Bogie; Erin Wicke Dankert; Chris Scharenbroch
• Date Published: December 2013
• Length: 492 pages
• Grant Number(s): 2010-JR-FX-0021
• Agencies: OJJDP-Sponsored
• Publication Type: Report (Study/Research); Report (Grant Sponsored)

Annotation

This study evaluated nine risk assessment instruments for juveniles in 10 jurisdictions by comparing their predictive validity, reliability, equity, and utility.

Abstract

The study's overall conclusion is that the proper use of valid, reliable risk assessment can improve decisionmaking regarding the risk level for recidivism; however, the study also found that the power of some risk assessment instruments to accurately classify juvenile offenders by risk level may have been overestimated. Only three of the nine risk instruments were capable of accurately separating cases into low, moderate, and high risk levels. Several risk instruments touted over the last decade have substantial shortcomings and fail to convey what is most important to correctional administrators, i.e., the difference in outcomes between risk levels and the distribution of cases across the risk continuum. Study findings indicate that simple, actuarial approaches to risk assessment can produce the strongest results. Adding factors with relatively weak statistical relations to recidivism - including dynamic factors and criminogenic needs - can result in reduced capacity to accurately identify high-, moderate- and low-risk offenders. The lack of standards for measuring instrument validity and reliability further complicates decisionmaking for administrators. Greater emphasis should be given to reliability testing and validation studies before and after risk assessment instruments are transferred to other jurisdictions. The establishment of national standards would assist in this effort. The nine risk instruments examined are the Positive Achievement Change Tool (PACT); the Youth Assessment and Screening Instrument (YASI); the Youth Level of Service/Case Management Inventory (YLS/CMI); the Comprehensive Risk and Needs Assessment (CRN); the Juvenile Sanction Center risk assessment instrument; the Girls Link risk assessment instrument; the Arizona Administrative Office of the Courts risk assessment instrument; the Arizona Department of Juvenile Correction Dynamic Risk Instrument (DRI); and the Oregon Juvenile Crime Prevention (JCP) assessment. 24 tables, 78 references, and appended risk/needs assessment systems, validation results by site, reliability results by site, expert scorer qualification, staff perceptions, and administrator advice