Multilevel Recidivism Prediction: Incorporating Neighborhood Socioeconomic Ecology in Juvenile Justice Risk Assessment

• NCJ Number: 235922
• Journal: Criminal Justice and Behavior Volume: 38 Issue: 8 Dated: August 2011 Pages: 840-853
• Author(s): Eyitayo Onifade; Jodi Petersen; Timothy S. Bynum; William S. Davidson, II
• Date Published: August 2011
• Length: 14 pages
• Grant Number(s): 0804515
• Sponsoring Agency: National Science Foundation
• Publication Format: Article
• Publication Type: Report (Study/Research)
• Language: English
• Country: United States of America

Annotation

This article examines juvenile justice risk assessment.

Abstract

Risk assessments such as the Youth Level of Service/Case Management Inventory (YLS/CMI) that predict delinquency outcomes based on proximal risk factors may benefit from an incorporation of distal risk factors in their prediction models. This study utilized a juvenile probationer sample and block group SES data in exploring the differential predictive validity of the YLS/CMI with youth of similar person-centered risk levels from different criminogenic neighborhood types. The study entailed an exploratory factor analysis of block group socioeconomic variables, which were used in a cluster analysis to create criminogenic neighborhood typology system. Hierarchical logistic regression was used to analyze the relationship among recidivism (Level 1), risk score (Level 1), neighborhood SES factors (Level 2), and neighborhood types (Level 2). Significant interactions were found across levels among variables, suggesting the riskrecidivism relationship was moderated by neighborhood socioeconomic ecology. Implications for practice and policy are discussed. (Published Abstract)