Overall, the results of the study were in line with the notion that smoking cues weaken the association between argument strength and effectiveness of the PSAs. Specifically, stronger antismoking arguments should yield PSAs perceived to be more effective and they do in this study, but presence of smoking cues weakened this association.
Smoking-related visual scenes in antismoking PSAs are often deployed to improve message relevance to smokers because messages highly relevant to target audiences draw more attention and are more persuasive (Petty & Cacioppo, 1979
; Roser, 1990
). As well, due to automatic processing nature of visuals, smoking cues become important message features for such PSAs to efficiently deliver the main arguments. Also, the smoking cues themselves are sometimes a part of the core argument being made in the PSAs and not merely “window dressing.” In such cases, such cues might function as an important complement to the overall argument. We are testing these ideas in ongoing studies. In the meantime, for persuasive messages to be effective, arguments employed in PSAs need to produce favorable and minimize unfavorable thoughts about cessation. The present investigation reveals that smoking cues can generate—at least in some circumstances—unfavorable effects on antismoking PSAs by reducing the impact of strong arguments.
Our analyses show that PSAs with stronger antismoking arguments are seen as more effective irrespective of the presence of smoking cues. However, the associations between argument strength and outcome variables—valenced thought, persuasiveness, and transportation—were mitigated by smoking cues. Interestingly smoking cues alter smoker participants’ unfavorable thoughts—wanting to continue smoking. When there is no cue in the PSAs, a negative association appears between argument strength and unfavorable thoughts. The stronger the argument becomes, the less likely the viewers are to say they have thoughts about wanting to continue smoking. However, when cues are present in the PSAs, the association between argument strength and unfavorable thoughts is flat to slightly positive, undermining the general effect of argument strength against smoking.
Our study did not have a standard measure of smoking urge, only a measure of unfavorable thoughts about cessation. As well, smoking cues were shown to alter the pattern rather than create significant increase in unfavorable thoughts. Nonetheless, our finding is unexpected and alarming because smoking cues activate unfavorable thoughts, functioning as potential distracters and inviting smokers to think about their smoking desires, although they are presented in the context of antismoking content. Using a standard measure of smoking urge, Kang et al. (2009)
also showed that smoking urge increased from baseline when smoking cues appeared in such PSAs with weak arguments.
The goal of antismoking PSAs is to reduce cigarette smoking, and smoking cues are commonly used to help PSAs to be more engaging and persuasive to the target audience. From a practical perspective, our data suggest that the addition of smoking cues may not always be effective in advancing the PSA's intent. These results imply that inclusion of smoking cues should be carefully considered in designing antismoking PSAs in concert with PSAs’ level of argument strength.
Despite the consistent set of findings observed over a large set of antismoking PSAs, the current study has some limitations. First, our measures of effectiveness are not measures of behavior, intention, or belief change (but see Dillard et al., 2007
). Nevertheless, messages that are emotional, transporting, and create thoughts of quitting have been shown to be linked to increased smoking cessation (Durkin et al., 2009
). Second, our study is observational, allowing the factors in PSAs to vary freely, although with random assignment to persons. To minimize potential confounds, we control a wide array of individual and message characteristics and employ a large number of messages avoiding the problem of a single (or small set) of PSAs that could be unique.
Evidence about the role of smoking cues in antismoking PSAs is growing. The present study indicates that the effectiveness of PSAs is mitigated by the presence of smoking cues across a large number of PSAs and for a large number of person and message characteristic controls. Future research needs to test these effects in an experimental context and to test behavioral outcomes beyond self-reported urges and perceived effectiveness.