Measurement Matters: Comparing Different Operationalizations of Juvenile Recidivism through Point Estimates and Multivariate Multilevel Models

• In Event: NIJ-Supported Research on Understanding and Measuring Juvenile Reoffending

Abstract

State juvenile justice systems vary considerably in how they measure juvenile recidivism. The proposed best practice entails assessing two-year re-adjudication/reconviction and following youth into the adult criminal justice system, but only one-third of U.S. states currently use that combination of follow-up period, marker event, and data source. It is unclear whether, and how much, adopting all of these measurement strategies would change states’ estimates of juvenile recidivism rates and the predictive validity of their recidivism risk assessment tools. Using an expansive recidivism database on a large sample of youth leaving juvenile justice supervision in Florida, this study compares the rates and predictors of two-year re-adjudication/reconviction in the juvenile or criminal justice systems with those of other recidivism measures that are commonly used in the U.S. The results reveal that both the rates and the predictors of juvenile recidivism vary widely depending on how that recidivism is operationalized. The implications for measurement and risk prediction are discussed.

Authors

• Sonja Siennick, Florida State University

• William Casey, Florida State University