Deceived, Disgusted, and Defensive: Motivated Processing of Anti-Tobacco Advertisements

ABSTRACT

A 2 × 2 experiment was conducted, where participants watched anti-tobacco messages that varied in deception (content portraying tobacco companies as dishonest) and disgust (negative graphic images) content. Psychophysiological measures, self-report, and a recognition test were used to test hypotheses generated from the motivated cognition framework. The results of this study indicate that messages containing both deception and disgust push viewers into a cascade of defensive responses reflected by increased self-reported unpleasantness, reduced resources allocated to encoding, worsened recognition memory, and dampened emotional responses compared to messages depicting one attribute or neither. Findings from this study demonstrate the value of applying a motivated cognition theoretical framework in research on responses to emotional content in health messages and support previous research on defensive processing and message design of anti-tobacco messages.

Notes

¹ A description of the stimuli is available upon request from the first author.

² The message order variable listed in the design showed no interactions with the independent variables on the dependent variables, and therefore was eliminated from the reported analyses.

³ Accuracy does not take into account the foils in its computation because it is the hit rate (% correct targets). Sensitivity is a ratio of hits to false alarms, and it takes into account the ability of the viewer to discriminate targets from foils. Criterion bias indicates that confidence in the decision a participant makes of whether or not an item was previously seen. A conservative (higher value) criterion means that there will be fewer false alarms (i.e., saying “yes” to a foil), but also there will be more misses (i.e., saying “no” to a target). Conversely, if a person adopts a liberal (lower) criterion, there will be more hits (i.e., saying “yes” to targets), but there will also be more false alarms. Nonparametric measures for sensitivity (A′) and criterion bias (B″) were used.