Alcohol Use in College Students : Testing the Self-Medication Hypothesis Applying Pooled Time Series Cross-Section Analysis

by William J. Davis

Institution: Central Connecticut State University
Department: Department of Psychology
Year: 2015
Keywords: College students – Alcohol use.; Self medication.
Record ID: 2062143
Full text PDF: http://content.library.ccsu.edu/u?/ccsutheses,2062


The co-occurrence of mood disorders such as depression or anxiety with alcohol abuse is well-documented. However, attempts to explain this relation have been controversial. The self-medication hypothesis (SMH) posits people use alcohol in order to attenuate or cope with distressing emotional states. Two of the predictions of the SMH were tested: (1) depression and/or anxiety are related to alcohol use and (2) symptoms of depression and/or anxiety precede alcohol use. A third prediction of the SMH, that the relief of psychiatric symptoms negatively reinforces alcohol use leading to continued use, was indirectly examined by testing the hypothesis that drinking motivation moderates the relation between depressive and anxiety symptoms and alcohol use. A total of 1308 participants from the Brain and Alcohol Research in College Students study who completed at least 2 monthly surveys over the course of a year were selected for analysis. A pooled time series cross-sectional design was used due to its ability to detect both change over time and the effect of predictors on individual outcomes. Granger non-causality testing was used to empirically establish the cause and effect relation between alcohol use and symptoms of depression and/or anxiety. The relation between depression and alcohol use was supported but no evidence was found supporting the relation between anxiety and alcohol. The hypothesis that depression precedes alcohol use was also supported. The drinking motivation factor of conformity was found to partially moderate the relation between depression and overall monthly alcohol use while the drinking motivation factor of enhancement was found to partially moderate the relation between anxiety and binge drinking. These findings support the self medication hypothesis for depression and alcohol use, but not for anxiety and alcohol use. The results of this study provide encouragement that the use of pooled time series cross-section analysis can be useful in psychological research. "Submitted in Partial Fulfillment of the Requirement for the Degree of Master of Arts in General Psychology."; Thesis advisor: Carol Shaw Austad.; M.A.,Central Connecticut State University,,2015.;