Alcohol outcome expectancies and alcohol use and problems

Numerous studies have demonstrated that alcohol outcome expectancies are strong correlates of various aspects of alcohol use. However, it has been suggested that forced-choice alcohol expectancy items may not measure the most salient anticipated effects of alcohol for an individual, and thus may create superficial responses. Additionally, research on attitude-behavior consistency has demonstrated that the nature of how an attitude is elicited can lead to attitudes that are more or less cognitive. The present study examined the applicability of this aspect of attitude theory to the study of alcohol outcome expectancies and the utility of self-generated outcome expectancies and subjective evaluations in predicting alcohol use and problems.

In an extension of methodology previously used in attitude research, alcohol-using college students (N = 165, 87 female) generated alcohol expectancies in response to two sentence stems designed to elicit "affectively" and "cognitively" based alcohol expectancies and made subjective evaluations (SE) ratings of each item generated on dimensions of positivity, likelihood and immediacy.

While no support was found for the generalizability of a particular aspect of attitude theory, several interesting findings emerged. SE ratings were significantly related to measures of alcohol use and problems and, for alcohol use measures, predicted comparably to a standardized measure of expectancies. For alcohol dependence symptoms, the number of expectancies generated predicted significant unique variance over that accounted for by subjective evaluations.

These results indicate that the sheer number of alcohol expectancies accessible to an individual may be an important correlate of alcohol dependence, independent of subjective evaluations.