Application of expectancy-value theory to drug use and drug use intentions

The primary purpose of this study was to evaluate the use of two expectancy-value models in predicting intentions to use marijuana and tranquilizers. One model was developed by Martin Fishbein; the other, by Harry Triandis. Both have been found to predict behavioral intentions in several areas. A review of the drug use literature revealed few studies which have applied a well-researched model or theory of behavior to drug use.

Attitudes, beliefs, and behavioral intentions toward use of marijuana and tranquilizers were measured by administering a questionnaire to 484 college students and entering freshmen at The University of Tennessee, Knoxville. Attitudes toward the use of these drugs were measured by four semantic differential items. Expectancies and values were measured with respect to nine possible consequences of drug use. Two hundred and forty-five participants returned the questionnaire.

Multiple regression analyses were conducted separately for each model for each drug. All analyses yielded significant multiple correlations, although the Triandis model accounted for more of the variance in intentions to use both marijuana and tranquilizers than did the Fishbein model. It was hypothesized that the relative importance of

vi normative and affective components would differ, depending on the drug. This was not borne out by the results; in both cases, a stepwise multiple regression analysis revealed that the personal normative belief made the largest contribution to the variance in drug use intentions, the attitude toward the behavior made the. second largest, and the expectancyvalue component and social normative beliefs did not add uniquely to the variance.

The results indicated that intentions to use drugs can be best predicted using the Triandis model of interpersonal behavior. The particular significance of personal normative beliefs and attitudes may explain the failure of traditional drug education and drug abuse prevention efforts. The results are also consistent with a cognitive social learning approach to drug use.