Social influence processes and college student drinking: the mediational role of alcohol outcome expectancies.

OBJECTIVE: Social influences are among the most robust predictors of adolescent substance use and misuse. Studies with early adolescent samples have supported the need to distinguish among various types of social influences to better delineate relations between social factors and alcohol use and problems.
METHOD: The first major goal of the present study (N = 399, 263 women) was to examine unique relations between particular facets of social influence and alcohol use and problems in a relatively heavy-drinking population (i.e., college students). We hypothesized that active social influences (offers to drink alcohol) and passive social influences (social modeling and perceived norms) would demonstrate positive associations with measures of alcohol use and problems. We also tested the hypothesis that alcohol outcome expectancies would mediate associations between social influences and drinking behaviors.
RESULTS: Structural equation modeling analyses provided strong support for the first hypothesis. Social modeling demonstrated the strongest association with alcohol use and problems, and active social influences demonstrated significant associations with both use and problems. Perceived norms were related to alcohol use, but not directly with alcohol problems. Support for the second hypothesis was positive but limited to one type of social influence. Strong evidence for a mediational role of outcome expectancies was found for relations between social modeling and alcohol use and problems.
CONCLUSIONS: Together, these findings demonstrate the unique and relative contribution of active and passive social influences and provide limited support for a hypothesized process by which social factors influence cognitions and alcohol-related behaviors.