Development and Initial Validation of the Alcohol Expectancy Task.

BACKGROUND: Although studies have shown that alcohol expectancies are prominent predictors of alcohol initiation and subsequent drinking levels, the questionnaires used to assess these expectancies among young adolescents have been criticized as being time-intensive, biased, and inappropriate.

METHODS: In response, we developed the Alcohol Expectancy Task (AET), in which 8 scenarios featuring adults in everyday situations and in different emotional states, accompanied by photographs of a range of beverages (4 alcoholic, 8 nonalcoholic), are displayed on a tablet screen, and participants are then asked to tap on the beverage they think the given person had been drinking.

RESULTS: In a first study among 184 adults (75.1% women; mean age = 37.8, SD = 12.2), results from a repeated-measures analysis of variance revealed a strong correspondence between the emotions depicted in the scenarios and how the participants interpreted them. In a second study, this time among 283 third and fourth graders (50.2% girls; mean age = 10.6, SD = 0.69), a confirmatory factor analysis confirmed the 4-factor structure of the AET. The results from a logistic regression analysis showed that the more often young adolescents assigned alcohol to the adults in an arousal-positive mood than to those in a sedation-negative mood, the more likely they were to have already consumed alcohol more than twice. Questionnaire-assessed expectancies were unrelated to adolescents' drinking and did not affect the associations of the AET.

CONCLUSIONS: The AET has the advantage of being time-efficient and convenient and could overcome certain limitations associated with questionnaire-based assessments of alcohol expectancy.