Do textual styles in cigarette health warning labels matter?

OBJECTIVES: To determine whether different textual styles (testimonial vs. didactic) on cigarette health warning labels influence responses to the warning labels.
METHODS: A field experiment was conducted with adult smokers (n=584) and 15- to 18-year-old adolescents (n=593) in Indonesia using intercept recruitment strategies. Respondents were randomly assigned to one of two conditions: 1) warning labels with didactic text (i.e., smoking causes lung cancer); or 2) warning labels matched on all characteristics except that they contained testimonial text (i.e., “I am suffering from lung cancer because of smoking”). For each condition, respondents were randomly assigned to evaluate warnings for two out of eight possible health topics, with each topic containing four to six warnings that systematically varied key characteristics (i.e., presence vs. absence of picture; pictorial types: graphic organ damage, personal suffering, symbolic representation of risk). Respondents rated each warning for overall effectiveness, believability, relevance, and negative emotional arousal. Linear mixed effects models regressed ratings on dummy variables for textual conditions, for image types, and the interaction between them, both for the entire population and when estimating models separately for adults and adolescents.
RESULTS: In linear mixed effects models, effectiveness ratings were significantly higher for didactic than testimonial text only among adolescents. When estimating model for entire population, significant interactions were found between condition and age group, and between condition and image type. Similarly, believability ratings were also significantly higher for didactic than testimonial only among adolescents and when considering pictorial images. Negative emotional arousal and relevance ratings were not significantly affected by condition. However, when estimating relevance model for adolescents, significant interaction was found between condition and smoking status.
CONCLUSIONS: Didactic text style appears to influence youth responses to cigarette warning labels. Future research should determine its effects on cessation-related outcomes.