The influence of drinking consequences on alcohol expectancy likelihoods and valences: an item level multilevel approach

Abstract

Theories of risk evaluation propose that experiencing an alcohol consequence should alter youth’s expectancies of that consequence occurring in the future. Prior studies have found this to be true for positive, but not negative consequences. However, previous research has aggregated across positive and negative consequences. We re-analyzed data from a sample of college students who drank in the past 12-months (n = 378), using Bayesian cross-classified multilevel ordinal regressions to estimate the associations between the experience of 39 different alcohol related consequences, participants’ expectancies for future drinking episodes, and the differences in these associations across participants and consequences. Our results align with expectancy theory suggesting that people do update their expectancy likelihoods and valences based on experiences. Implications of this work and avenues for future research are discussed.