(PS1-10) College Alcohol Beliefs or General Alcohol Expectancies: Which Is More Important in the Prediction of Hazardous Alcohol Use Among College Students?

Abstract Text: Alcohol use is prominent on college campuses. National data from 2022 indicated that lifetime alcohol use by college students was 71%, with 38.5% reporting weekly use (American College Health Association, 2022). Many students drink heavily and binge drinking is common. However, not all students who drink heavily experience negative outcomes and substance use scholars have advanced the construct hazardous drinking to delineate a type of drinking that resides between non-problematic and formally disordered consumption. Vicary and Karshin (2002) argued that hazardous drinking is a more representative construct than binge-drinking as it captures both frequency/quantity and alcohol-related negative consequences. Understanding the predictors of hazardous alcohol use on college campuses is of high importance. A well-studied, robust predictor of overall alcohol consumption across ages and populations is alcohol expectancies (AE). AE refers to individuals’ beliefs about the outcomes of drinking (Black et al., 2012). Within the larger universe of expectancies that operate across all ages are a subset specific to college students and which have been linked to heavy drinking: college alcohol beliefs. College alcohol beliefs (CABs) are an individuals’ beliefs about the role and importance of alcohol to the college experience (Lemoine et al., 2020; Osberg & Boyer, 2018). Higher CABs—meaning seeing alcohol use as more central to the college experience—have been linked to both higher negative alcohol-related consequences and higher alcohol consumption. As part of a larger ongoing project on college alcohol use (final N = 200), we examined the relations between college alcohol beliefs and general positive alcohol expectancies in the ability to predict hazardous alcohol use. Specifically, we tested whether CABs carried unique predictive power over and beyond general AE. We hypothesized that both CABs and AE would show significant positive relations with hazardous use and each would account for unique variance beyond the other. We tested this by building two prediction models and varying the order of entry. At present, 76 participants (M age = 20.3, 70.5% White, 61.5% women) have completed the study. Preliminary results indicate that, as predicted, both CABs (β = .465, p < .001) and AE (β = .255, p = .029 < .05) predict hazardous use in isolation. However, the models indicated that CABs explained 21.6% of the variance in hazardous alcohol use (R2 = .216) while AE explained only 6.5% of the variance of hazardous alcohol use (R2 = .065) at the first step. When AE were added to CABs, they did not increment the variance and only explained an additional .2% (R2 = .218; change in R2 = .002). However, when CABs were added to the model containing AE, an additional 15.3% of variance was explained (R2 = .218; change in R2β = .153). Additionally, when tested together, CABs retained its unique predictive power (β = .443, p < .001) while AE did not (β = .045, p = .709 > .05). The results suggest that college alcohol beliefs are prepotent in the prediction of college hazardous alcohol use and completely subsume general alcohol expectancies. This result is key for the design of college alcohol use interventions, which should strive to challenge and modify these cognitive structures.