Date of Completion

5-7-2015

Embargo Period

5-7-2015

Keywords

Crossover youth, dually-involved youth, child welfare, juvenile justice, trajectories

Major Advisor

Preston A. Britner, Ph.D.

Associate Advisor

Anne F. Farrell, Ph.D.

Associate Advisor

Ronald M. Sabatelli, Ph.D.

Field of Study

Human Development and Family Studies

Degree

Doctor of Philosophy

Open Access

Open Access

Abstract

Crossover youth (COY) is a broad term that describes those who are served, at any point, by both the child welfare and juvenile justice systems (Herz & Ryan, 2008). In 2012, The Connecticut Department of Children and Families (DCF; the state child welfare agency) and the Court Support Services Divisions (CSSD; the state agency that handles juvenile court matters) entered into a data sharing agreement to better understand the COY population in the state. The current study uses this DCF-CSSD dataset to address two aims: (1) understand how COY differ from youth who are involved only in the child welfare system, and (2) describe the variance that exists within the COY with the goal of identifying distinct profiles of youth who are involved with both systems.

Using the sample of all youth born in 1996 who had DCF involvement (N= 7,268), latent class growth analysis was used to establish trajectories of maltreatment based on the number of substantiated maltreatment allegations experienced by each individual over the first 16 years of life. The analysis found five distinct trajectories of maltreatment. Consistent with previous variable-centered studies, a trajectory class with child welfare involvement beginning in late child hood or early adolescence was more likely to crossover than most others. However, the analyses also detected a small subgroup that had persistent child welfare contact over the 16-year period and that was most likely to crossover.

A second set of analyses was conducted only on the youth who did crossover (n= 1312). Previous research has largely treated COY as a single group ignoring the heterogeneity that is likely to exist in the population. To extend the understanding of COY the current study used latent class analysis to identify subgroups within the COY sample. Analyses supported a four-class solution; classes differed from each other on the extent of involvement in both the child welfare and the juvenile justice system. Taken together, these analyses provide a greater understanding of COY and how individual patterns of involvement contribute to youth experiences.

COinS