An assessment of the relationship between attitudes and behavior: Attitudes of adolescents, normative values, and marihuana-use behavior

Abstract

The relationship between attitudes related to marihuana use and marihuana-use behaviors was examined within the context of impression-management theory of attitudes (Tedeschi, Schlenker, & Bonama, 1971). Marihuana use was treated as problem behavior that elicited negative sanctions from established social forces. Normative values in this country related to marihuana use were estimated from the attitudes expressed by parent-age adults (PARENT, N=494). Attitudes expressed by adolescents were expected. to vary from the norms with reported marihuana-use behaviors. Attitudes expressed by a grandparent-age group (GPARENT, N=590) were examined relative to the other groups as a control for generation-gap effects. Adolescents who had no opportunity to use marihuana (NONE, N=529) were expected to reflect the attitudes expressed by the PARENT group. Nonusers who had an opportunity to use (NOOP, N=185), those who used at the first opportunity and later stopped (FQUIT, N=36), and those who resisted the first opportunity, later used, and then stopped using (LQUIT, N=17) were expected to express attitudes toward marihuana which were more negative than the norms. Adolescents who used at the first opportunity and continued using (FNOW, N=49) were expected to express positive attitudes toward marihuana. Data were collected by interview from a national probability sample by Abelson, Cohen, and Schrayer (1972) for the National Commission on Marihuana and Drug Abuse and were obtained for the present study from the Drug Abuse Epidemiology Data Center, Institute of Behavioral Research, Texas Christian University. The reliability and validity of the attitude items selected were estimated by subjective rating, scaling analysis, factor analyses, and a comparison study of factor structures produced by PARENT and NONE scores. A multiple discriminant analysis produced two significant functions: political attitudes representing 92 percent of the variance among groups, and freedom of expression, representing 6 percent. The NONE group expressed conservative political attitudes and the FNOW group, liberal political attitudes as predicted. Attitudes expressed by members of FQUIT were in the direction expected but not to the degree predicted. The attitudes of NOOP and LQUIT groups were not explained by impression-management theory. Alternative theories, experimenter effects, and variables not considered in the present sample were discussed. Recommendations for areas requiring attention in future research included: 1) differential social norms; 2) values weighting subjective norms; 3) behavioral intent; 4) elapsed time from behaviors to attitudes expressed, and 5) the level of specificity among attitudes and behaviors.