Targeting the smoker in an anti-smoking campaign.

Between May and September 1983, 1,661 smokers from a random sample of the populations of Sydney and Melbourne were interviewed in their homes. Of the first group, 219 were followed up 12 months later, representing a 75% response rate. Intention, measured by perceived likelihood to quit, was validated as a predictor of a later attempt to quit smoking in the cohort study. Males reported likelihood to quit more often than females. Perceived importance of smoking as a community health problem was also important in predicting attempts to change smoking status, indicating the possible importance of an agenda-setting role for the mass media in promoting change. The key finding was the interaction between health beliefs and social influence in predicting level of intention. On their own, health beliefs showed no relationship to perceived likelihood to quit, and social influence could be counterproductive. However, taken together, these two variables were strongly predictive of change. This suggests that a combination of these two messages should be used in anti-smoking campaigns.