Knowledge of important etiological factors in the development of problem drinking and alcoholism should usefully inform the design of prevention efforts. Few strategies for prevention have been based on such information, however, primarily because the majority of identified etiological factors are not malleable. Research has provided an extensive construct validation network that shows that these other factors in part influence drinking via cognitive processes (expectancies) that are potentially malleable. Expectancies are memory templates of previous experiences that guide future behavior. Recent investigations have shown that drinkers can be provided with experiences in which drinking and its usual behavioral outcomes can be disconnected in a highly salient manner, and that this """"""""expectancy challenge"""""""" may result in short-term decreases in consumption in college students. These investigations provide the foundation for the first extended tests of the prevention capabilities of the expectancy challenge model. The proposed research will: replicate and extend preliminary findings regarding the efficacy of the expectancy challenge procedure with emphasis on explicating the underlying processes that prevent and reduce alcohol misuse; compare the efficacy of the challenge procedure to alcohol skills training and to a combined challenge/skills program; and investigate various risk factors as possible moderators of the challenge effect. College students will be the target of these efforts because of the recent recognition of the serious consequences of excessive alcohol use in this population. A 3 (expectancy challenge manipulation) x 2 (alcohol skills training manipulation) factorial design will be used. This design will be replicated across gender; all participants, assessment personnel, and intervention personnel, will be either male or female within each replication. Slopes of individual drinking, expectancy, and other variables, followed over two years, will be subjected to recently developed statistical analyses (including hierarchical linear modeling and possibly latent growth curve modeling). Results will provide a foundation for large scale trials to prevent alcohol misuse in college students, and possibly for the extension of these methods to other risk groups.
Showing the most recent 10 out of 13 publications
