The major aim of this dissertation was to assess the utility of general social strain theory (Agnew 1992) in explaining substance use among Black, Hispanic, and White adolescents. In terms of my outcome variable, I examined a cumulative measure of substance use, marijuana use, and alcohol use separately. Data were analyzed from wave three of the Project on Human Development in Chicago Neighborhoods (2000-2002). This study explored how under-researched race/ethnic-related social strains such as racial discrimination and general forms of strain such as witnessing violence, physical and verbal victimization affected adolescent substance use. In addition, I examined two important mediating negative affective states, depression and fear. Finally, and in accordance with social strain theory, the moderating effects of parental social support and religion were also included in my statistical analysis. Statistical analysis was primarily composed of path analysis using full information parameter estimates.
With respect to fear as a negative affective state, results indicated that reporting fear decreased substance use among adolescents while depression had no significant effect. Other findings indicate that racial discrimination, witnessing violence, and being the victim of verbal bullying indirectly affected alcohol and marijuana use as separate constructs through fear. Physical victimization was the only social strain that had a direct positive effect on marijuana and alcohol use. Social support and religion had a direct negative effect on alcohol and marijuana use. As respondents report higher rates of social support and religious participation, respondents reported less substance use. However, social support and religion did not moderate the relationship between social strains and substance use as theoretically expected. Finally, the sample was separated into three race/ethnic categories: African American, Hispanic, and White. The results revealed that social strains may be operating differently for Hispanics and African Americans compared to Whites. In the Hispanic model fear mediated the relationship between discrimination and verbal bullying but not in the African American and White models. Hispanic respondents reported more depression, engaged in more marijuana use but not more alcohol use. These findings reveal that under-researched ethnic/race based social strains and processes should be considered in future studies.