Skip to main content

Risk/Needs Assessment for Recidivism: Implementation and Effectiveness Study

Principal Investigator

Dr. Gina Vincent, Ph.D.

Center for Mental Health Services Research
Department of Psychiatry
University of Massachusetts Medical School



One consistent need expressed across sites in the Models for Change states has been for standardized tools that can assist the courts in making processing and placement decisions for young offenders. Ideally, tools will assist with decision-making in a manner that fosters rehabilitation through appropriate and targeted interventions and decreases risk for delinquency in the future (risk reduction). At the same time, another goal of Models for Change states is to reduce the number of youths in secure care facilities. However, it is important that while striving towards these goals, the "right" youths are diverted from incarceration. Arguably, these would be youth at low or only moderate risk for serious delinquent or violent acts in the future.

One promising approach for achieving these goals is the implementation of a tool that can identify adjudicated youths at the highest risk for engaging in serious delinquency (risk assessment) and also assist with case management (otherwise known as risk management) at a key point in the JJ system. We refer to these tools as risk/needs assessment tools. A number of states around the U.S. have introduced risk/needs assessment tools into their JJ systems. This study proposes to implement a standardized and valid risk/needs assessment tool in probation departments to guide disposition recommendations and service planning for adjudicated youths. 

Purpose for the Research

Although it is a promising practice, there are several outstanding issues with adopting a risk/needs assessment tool in probation departments. Mainly, there are no experimental or even quasi-experimental studies available to suggest that the implementation of standardized risk/needs assessment tools in correctional systems actually results in changes in the processing of offenders or in reducing recidivism. Additionally, the effectiveness of these instruments has not been studied. Effectiveness in this case means the degree to which an instrument that has been shown to have adequate reliability and validity in controlled studies maintains its inter-rater reliability and validity once it is put in place in less-controlled, real-world sites.  

Study Methodology

This is an implementation and effectiveness study of a validated, best practice, youth risk/needs assessment tool for recidivism as selected by the test sites (e.g., Structured Assessment of Violence Risk in Youth [SAVRY], Washington State Juvenile Court Assessment [WSJCA], Youth Level of Service Inventory [YLS]). The study uses a quasi-experimental, pre-post test design with a two-year follow-up, involving the collection of both qualitative and quantitative data from multiple sources. We will implement the risk/needs assessment tool at a minimum of two interested test sites (county or parish probation offices) per state (in Louisiana and Pennsylvania only) to be used by probation officers for post-adjudication/pre-disposition placement recommendations and service delivery planning. The study will examine the following outcomes:

  • Implementation-level outcomes including change in staff knowledge about the factors that contribute to a youth's risk for recidivism, and change in staff attitudes about rehabilitation. This will involve interviews with all probation officers and administrators at the test sites pre-post implementation.

  • Intervention-level outcomes including an increase in the proportion of adjudicated youths receiving community supervision, a decrease in the number of youths receiving custodial placements, an increase in rates of service usage, and a potential decrease in recidivism. This will involve gathering administrative data sets two-years pre-and post-implementation in the test sites within each state.

  • Effectiveness of the risk/needs assessment tool: The effectiveness of the tool (meaning, how well it works when actually used in practice) will be determined by testing the inter-rater reliability, predictive validity for re-offending, and presence of gender or racial bias. This will involve gathering case-level data from the probation offices over a minimum 6-month period (about 250 cases per site) and tracking records for approximately 15 months. 

What are the benefits to participating?

Sites would receive a number of benefits for their participation in this study, including training for their staff and supervisors on a risk/needs assessment tool and free technical assistance around implementation. Technical assistance includes assistance in setting up their procedures and decision-making process for pre-disposition recommendations, development of forms for case management and referral tracking, and assistance with designing a database for entering scores. In addition, they will get a half-time research associate. The project is intended to enhance the probation officers' existing assessment process, rather than adding superfluous work.

Supported by

Models for Change is supported by the John D. and Catherine T. MacArthur Foundation, website operated by Justice Policy Institute.