Message-Elicited Brain Response Moderates the Relationship Between Opportunities for Exposure to Anti-Smoking Messages and Message Recall

Abstract

Campaign success is contingent on adequate exposure; however, exposure opportunities (e.g., ad reach/frequency) are imperfect predictors of message recall. We hypothesized that the exposure-recall relationship would be contingent on message processing. We tested moderation hypotheses using 3 data sets pertinent to "The Real Cost" anti-smoking campaign: past 30-day ad recall from a rolling national survey of adolescents aged 13-17 (n = 5,110); ad-specific target rating points (TRPs), measuring ad reach and frequency; and ad-elicited response in brain regions implicated in social processing and memory encoding, from a separate adolescent sample aged 14-17 (n = 40). Average ad-level brain activation in these regions moderates the relationship between national TRPs and large-scale recall (p < .001), such that the positive exposure-recall relationship is more strongly observed for ads that elicit high levels of social processing and memory encoding in the brain. Findings advance communication theory by demonstrating conditional exposure effects, contingent on social and memory processes in the brain.

Keywords: Health Campaigns; Mentalizing; Message Effects; Neuroimaging; Youth, Memory.

Figure 1

Overview of fMRI study task and brain regions of interest. (Left panel) Participants viewed a 4-second preparation countdown and were then instructed to view one of the 30-second “The Real Cost” ads. Participants then rated their intention to share the ad on social media and were asked to close their eyes and reimagine the ad in their mind’s eye. Each participant completed these tasks in the same order for all 12 ads; however, the order in which the ads were presented was randomized. The current study focused on neural response during the ad exposure period (outlined in red); sharing ratings and neural responses during the sharing and reimagining portions of this task were not assessed in this study. (Right panel) Neural response was measured in (a) social-processing regions and (b) memory-encoding regions. For additional details on brain regions of interest, please see Supplemental Figure S1. FMRI = functional magnetic resonance imaging.

Figure 2

Association between past 8-week TRPs and predicted cued ad recall, at higher and lower levels of ad-elicited neural response in social-processing and memory-encoding regions, controlling for whole-brain neural response and potential covariates. In each figure, the blue line illustrates the relationship between past 8-week TRPs and cued recall, as predicted by (a) the social-processing moderation model and (b) the memory-encoding model (see Table 1), for ads that elicited higher levels (+0.5 SD) of neural response in the corresponding brain regions. The orange line illustrates the relationship between past 8-week TRPs and predicted recall for ads that elicited lower levels (-0.5 SD) of neural response in corresponding brain regions. Results reflect predicted relationships between TRPs, cued recall, and neural response in these regions, holding whole-brain neural response and potential covariates constant. TRP = target rating points. SD = standard deviation.