This study aimed to prospectively identify ecological risk factors for juvenile arrest in a sample of youth at risk for maltreatment. Chi-Squared Automatic Interaction Detector analysis was performed with data from 592 youth from the Longitudinal Studies of Child Abuse and Neglect to identify the optimal combination of age 14 predictors of past-year arrest at age 16. Results extended previous research, which has identified being male, having more conduct disorder symptoms, suspension from school, perceived school importance, witnessing family violence, and having a jailed family member as key risk and protective factors for arrest by identifying important interactions among these risk factors. These interactions differentiate youth at the greatest risk of arrest, which, in this sample, were males with greater than two symptoms of conduct disorder who witnessed family violence. These findings suggest that longitudinal and multi-informant data could inform the refinement of actuarial risk assessments for juvenile arrest.
Keywords: ACEs; actuarial approach; juvenile justice involvement; school engagement.