Examing the Factor Structure of the Ohio Youth Assessment System Residential Tool

• In Event: Posters I

Abstract

The Ohio Youth Assessment System (OYAS; 2010) was designed to determine the risk of re-offending for youths in the juvenile justice system. This assessment consists of five risk measures for critical points in the system (i.e., diversion, detention, disposition, admission to residential facilities [OYAS-RES], and community re-entry). All the above measures are lacking in specific psychometric support, namely construct validity evidence. As a first step, the 33-item dichotomously-scored (i.e., 0 = Not Present, 1 = Present) OYAS-RES was explored. The purpose of the OYAS-RES is to determine the appropriate length of stay in residential facilities. This self-report measure is completed after disposition for youths (i.e., ages 11-20) placed in residential facilities (i.e., ≥ 90 days). A random sample of 458 cases (i.e., 84.9% Male, 52.2% African-American) were analyzed using Categorical Principal Components Analysis (CATPCA) with nominal optimal scaling Varimax rotation. Fifty-three percent of the total variance is explained by the retained 23-item, six-component model with good internal consistency reliability (Kuder-Richardson-20 > .467 for all): (1) Substance Abuse, (2) Fighting/Gangs, (3) Decision-Making, (4) Family Perceptions, (5) Delinquency History, and (6) Authority Relationships. The results demonstrate construct validity evidence for the tool in making psychometrically-sound high-stakes decisions in the juvenile justice system.

Authors

• Linda M. Julian, Kent State University

• Aryn C. Karpinski, Kent State University