An Evaluation of a Family Preservation Juvenile Justice Program with a Cox Regression Model

Abstract

This article illustrates the use of Cox regression to analyze recidivism among felony offending juveniles who were assigned to one of two criminal youth programs. The program that employed an intensive home-based family preservation model was identified as the experimental group. The other program, which did not utilize an intensive home-based family preservation model, was labeled the control group. The study used a quasi-experimental design and Cox regression analysis to compare recidivism outcomes of juvenile offenders treated in the Partners Program with a control group of juvenile offenders who were not given the treatment. The Cox regression analysis revealed that for the juveniles treated in the Partners Program their length of time to recidivism was longer and risk of being re-arrested was lower than for the juveniles in the control group, adjusting for the various covariates.