This study is designed to continue a nation-wide evaluation of early impacts of federally mandated alcohol warning labels, by adding longer-term data collection points to three existing survey series undertaken from 6 months before to 18 months after the policy intervention. The investigation focuses on beliefs, attitudes, intentions, and alcohol consumption and risk-related behaviors that the label health messages were intended to affect--such as drinking when driving or when pregnant. Between the 1989 baseline and 1991, three annual telephone surveys were conducted in the U.S. general population with surveys in Ontario, Canada, serving as a nonintervention control. A small quarterly state survey in Utah by collaborating investigators during the same period supplemented the national data by providing fine-grained results on the time-patterning of label awareness, recall and hazard perceptions. New data, collection will be accomplished in 1993 and 1994, with the third grant year for continued analysis of the comprehensive five year dataset. For the U.S. study, a split-panel design will include 1,000 U.S. cross-sectional cases each year (1993 and 1994) and an additional initial sample of 1,000 target individuals followed longitudinally for 1 year. Simultaneously, cross sectional surveys of 1,000 cases each will be done in Ontario. Six quarterly telephone surveys of 400 cases in Utah will also be done to extend the state series. Together, these surveys extend the evaluation period to 4.5 years post intervention, completing the policy assessment originally intended. Results from several sources indicate that penetration of the warnings has not yet plateaued; it is expected that in the continuation period several specified outcome measures will attain detectable levels. The first study aim is to provide definitive information on the efficacy of the warning label. Extended time and added measures will increase capacity to detect hypothesized aggregate impacts in beliefs, attitudes and behavior using a cross-sectional pre-test multiple post-test nonequivalent control design. Secondly, using a 1 year longitudinal component, warning label influences are modeled for two target groups at risk for drunk driving and drinking when pregnant: men and women 18-40. The study will advance our understanding of warning processes, by modeling how health messages like those on container labels, in stores, or. on radio and TV, have a preventive influence on target groups. The third major aim is to study in the long-term diffusion across population subgroups of awareness and recall of health messages on the labels and in advertising, and their possible associations with reductions in risk behaviors or adoption of protective behaviors. This information will advance theory which can be applied to tailoring, focusing, and improving warnings. Subsidiary aims are to continue monitoring public opinions and newspaper coverage relevant to labeling and other alcohol control policies.
Showing the most recent 10 out of 13 publications
