What Works in Juvenile Justice Outcome Measurement: A Comparison of Predicted Success to Observed Performance

• NCJ Number: 197814
• Journal: Federal Probation Volume: 66 Issue: 2 Dated: September 2002 Pages: 50-56
• Author(s): Kristin Parsons Winokur Ph.D.; Ted Tollett M.S.; Sherry Jackson M.S.W
• Date Published: September 2002
• Length: 7 pages
• Sale Source: NCJRS Photocopy Services
• Publication Format: Article
• Publication Type: Program/Project Evaluation
• Language: English
• Country: United States of America

Annotation

The authors discuss their roles in the development of "Program Accountability Measures" (PAM) analysis, which is an outcome-based model that has been used to evaluate juvenile day treatment and commitment programs in Florida.

Abstract

The PAM model calculates how well a program is expected to do based on the program youths' risk of reoffending (expected success) and compares this to how well the program youths actually performed (observed success). This ensures that programs which serve more difficult youth are not held to inequitable standards due to the higher recidivism risk of the youth they serve. The authors also incorporated into the model a mean cost differential factor that compares the program's average cost per successful completion to the statewide average cost. The PAM analyses presented in this article include effectiveness comparisons for all day treatment and residential programs that were serving youth in Florida during the 2-year period between July 1, 1998, and June 30, 2000. A total of 17,762 juveniles were released from 186 programs during this period. A PAM score was calculated for each program to provide a program rank based on its effectiveness and cost relative to other programs. The score was derived from a formula based on program youths' recidivism and average cost per youth who completed the program. Program effectiveness was defined as the difference between a program's predicted success and its actual success. The authors present findings from the analysis of all 186 programs according to program model, gender composition, and security level. This presentation is intended by the authors to demonstrate the type of analysis permitted by the PAM model. The authors conclude that the statistical approach of the PAM model offers evaluators, policymakers, and funding sources an important new option to measure and reward the intangible factors that contribute to successful outcomes. 4 tables, 9 references, and appended description of the steps for calculating the PAM score