Effect of message approach and image size on pictorial health warning effectiveness on cigarette pack in Indonesia: a mixed factorial experiment

Abstract

Background: Cigarette consumption remains high and increasing in Indonesia. The government implemented a pictorial health warnings requirement of 40% cover of the pack (front and back) using fear appeal messages. Objective: Our study aims to assess the effectiveness of cigarette pictorial health warnings by message and size. Methods: We conducted a mixed factorial experiment online study using three messaging approaches (fear vs. guilt vs. financial loss) and two picture sizes (40% vs. 75%) among 209 smoking participants. Sociodemographic variables included gender, education, income, employment status, and marital status. Data analysis used a mixed model ANOVA to see the main effect and interaction effect on dependent variables. For subgroup analysis, we used t-test and one-way ANOVA. All analyzes were in SPSS 22. Results: We found significant differences in the three message types, in which fear and guilt have higher effectiveness than financial loss. By subgroup, the guilt message was more compelling among female smokers and married smokers. The financial loss message was effective among lower-income smokers. We found no difference in pictorial health warning effectiveness by image size, potentially because participants could zoom in/out the cigarette pack image on the screen. Conclusions: Our finding supports more diverse message types in pictorial health warnings in Indonesia and other countries.