Date of Award


Document Type


Degree Name


Organizational Unit

Morgridge College of Education, Research Methods and Information Science, Research Methods and Statistics

First Advisor

Elizabeth Anderson

Second Advisor

Julia Dmitrieva

Third Advisor

Jeffrey Lin

Fourth Advisor

Lisa Pasko


Adolescent, D.A.R.E., Positive peer support, Prevention, Program, Substance


To evaluate this study’s research question of ”Does the latent construct of Positive Peer Support (PPS) relate to the construct of Adolescent Substance Use (ASU) over time, controlling for neighborhood safety, race, and sex?”, Structural Equation (SEM) and Latent Growth Curve Modeling (LGCM) were used to investigate trajectories. Secondary longitudinal data from Zimmerman (2014) of 604 students enrolled for four consecutive years in public schools located in Flint, Michigan. In the secondary data resource, students who participated were declared “at risk” by GPA. Significant relationships were found in SEM: Positive Peer Support to Adolescent Substance Use, All Control Variables to Adolescent Substance Use including Positive Peer Support. Longitudinal direct effects found no significant relationships with any variables, but Race, concluding that Race was the most significant variable in predicting initial status and growth rate of Adolescent Substance Use. Although Positive Peer Support did not show significant longitudinal effects of initial status and growth rate of ASU, there were significant directional effects in the SEM models. This concludes that Positive Peer Support does share a relationship with Adolescent Substance Use and should be taken into account in curriculum development and implementation of in-school substance prevention programming.

Publication Statement

Copyright is held by the author. User is responsible for all copyright compliance.

Rights Holder

Kady Rost


Received from ProQuest

File Format




File Size

254 pgs


Statistics, Curriculum development, Social research