Effects of Evaluative Conditioning on Implicit Evaluation of Alcohol and Drinking Behaviors: A Direct Replication.

Abstract

Recent research suggests that evaluative conditioning (EC) can change implicit evaluations of alcohol and reduce drinking behaviors among college students (Houben et al., 2010a). This research has been conceptually replicated in two previous studies. To date, however, no direct and independent replication of the original study has been performed. In this paper, we report a high-powered direct replication of Houben et al.'s (2010a) study. About 168 French college students took part in this preregistered study. Drinking behavior was assessed before and 2 weeks after the intervention. The intervention consisted of 120 trials of words related to alcoholic beverages or soft drinks paired with neutral, positive or negative pictures. The two conditions were factually equivalent and differed only in the repeated pairing between alcohol-related words and negative pictures; in the EC condition, but not in the control condition, alcohol-related words were systematically paired with negative pictures. EC did not change participants' implicit evaluations of alcohol and drinking behaviors. However, EC reduced drinking behaviors among hazardous drinkers. Yet, further non-preregistered Bayesian analysis did not provide much support for this hypothesis. This high-powered preregistered direct replication of Houben et al.'s (2010a) study suggests that the original effects are more fragile than initially thought. The effect of EC on drinking behaviors may be restricted to heavy drinkers, and we found no evidence that this effect is mediated by a change in implicit attitudes. It is necessary to perform further studies to test the original effects in clinical populations.