Influence of dose and beverage type instructions on alcohol outcome expectancies of DUI offenders

Objective: To test the hypothesis that self-reported alcohol outcome expectancies are situationally specific beliefs that vary depending on contextual variables, such as the quantity and type of beverage one is instructed to imagine consuming. Method: One hundred thirty DUI offenders attending a 72-hour weekend program were randomly assigned to either a "small dose" or "large dose" instruction condition and then completed five versions of the Comprehensive Effects of Alcohol scale, a separate version for each of five beverage types ("alcohol," "beer," "wine," "mixed drinks" and "straight liquor"). Results: Two separate 2 (dose) X 5 (beverage type) ANOVAs revealed that positive and negative expectancies varied significantly as a function of both dose and beverage type. Specifically participants endorsed more positive outcomes and more negative outcomes when they imagined drinking a large amount versus a small amount. In addition, participants endorsed significantly more positive expectancies for beer and mixed drinks than for wine. They endorsed the largest number of negative expectancies for straight liquor and the fewest for wine. Conclusions: Expectancies are to some degree context-bound and researchers and clinicians should consider providing explicit dose and beverage type instructions to research participants and clients when they use self-report questionnaires to assess alcohol outcome expectancies.