Does self-efficacy information improve the effects of pictorial fear labels on cigarette packages: real-life experiment

Abstract

Background. The effectiveness of scary health warning labels on cigarette packages is heavily debated. It is posited that fear arousing warnings will be effective only when self-efficacy is high. Shortly before pictorial labels were introduced in the Netherlands, we tested whether pictorial labels that combined negative health consequences with self-efficacy information were more effective. Methods: Dutch smokers (18-82 years) were randomly assigned to one of four warning label conditions: pictorial negative consequences (FEAR), pictorial self-efficacy information (SE), pictorial negative consequences and self-efficacy combined (COMBI), and textual negative consequences (TEXT). Each condition contained eight different labels. Participants were asked to put self-adhesive stickers on their own cigarette packages during four weeks. To check, they were asked six times to report the code on the stickers (368 participants passed this test). Main dependent variables were quit intention, quit attempt and cigarettes per day. ANCOVA was conducted (covariates were education, addiction, quit attempt in last year, first wave quit intention). Findings: At the last wave, results showed a main effect of condition on quit intention (p=.039) and quit attempt (p=.031), but not on number of cigarettes smoked (p=.551). Post-hoc tests showed that pictorial fear labels were more effective than textual fear labels. The SE and COMBI labels were not significantly more effective than pictorial and textual fear labels. Self-efficacy at first wave did not interact with condition effects. Discussion: Findings support the use of pictorial warning labels that concentrate on negative health consequences. Self-efficacy information did not improve the effect of warning labels.