Defense Date

2014

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Epidemiology

First Advisor

Resa Jones

Abstract

This study focuses on the social networks of probationers, who comprise the largest segment of the criminal justice population in the United States, but about whom there are few studies of network processes. It provides information on how elements of a probationer’s social network change over time and can affect drug use. This study employs longitudinal analysis of the social networks for 251 substance abusers on probation to examine how these networks are influenced by an intervention designed to increase pro-social behaviors and how network changes impact drug use. Baseline drug use of the probationers was examined according to the number of substances used in the last 6 months. Blacks were less likely to be polydrug users (aOR: 0.34, 95% CI: 0.14 to 0.84), while those using cocaine or heroin as their primary drug of choice were more likely to be polydrug users (aOR: 3.02, 95% CI: 1.32 to 6.94). Age at first illicit drug use was also significant, with those initiating drug use younger than 18 more likely to be polydrug users (aOR: 2.12, 95% CI: 1.01 to 4.46).The majority of probationers had drug user networks with the same number of persons in them over the 12-month follow-up period (82.5%), and perceived social support that also did not change (76.1%). Men were less likely to change their drug user networks over time and older persons were less likely to have decreasing social support over time. Those with low drug use that have increasing (aOR= 5.08, 95% CI: 1.09 to 23.75) and decreasing (aOR= 6.45, 95% CI: 1.35 to 30.85) drug user networks over time were more likely to be in the lowest drug using group compared to those with stable larger drug user networks. Older persons were less likely to be in the drug use trajectory (aOR=0.96, 95% CI: 0.92 to 0.99), whereas those with high criminal risk were more likely to be in a stable drug use class compared to an increasing drug use class (aOR=2.52, 95% CI: 1.03 to 2.64). The findings of this study indicate that changing the drug using networks of probationers may be difficult, given that most are stable over time and effective interventions to decrease substance use may need to target individual and structural factors, rather than social support and network composition. The finding that smaller networks that do change over time were associated with lower rates of drug use indicates that programs could also focus on mechanisms that determine how and why probationers choose drug using network members. Reducing drug using peers for corrections-involved populations may be difficult, but can lead to lower drug use rates which can also reduce recidivism.

Rights

© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

May 2014

Included in

Epidemiology Commons

Share

COinS