Abstract
This study examined threatening background visuals and voice-over differences in televised alcohol warnings. Participants ( N v = v 401) viewed four television adwarning pairs embedded in sports programming. Between-subjects conditions included a threatening visual behind the warning, a non-threatening visual, a plain background, and a no-warning control. Use of a male or female announcer, each using a relatively warm or imperative voice quality, were manipulated within-subject factors. Outcomes included knowledge, cognitive responses regarding warning content and presentation, and perceptions of the risk of the product (beer) shown in the ad. All warnings increased post-test knowledge of alcohol risks relative to control; the largest increase resulted when warnings were accompanied by the threatening visual. Threatening visuals also increased positive thought elaborations about the warning message content and presentation. However, the threatening visuals did not reliably impact perceived risk of beer use, suggesting that threatening visuals influenced responses and knowledge by increasing attention to the warning rather than by increasing perceived threat. These findings are consistent with the EPPM's proposed initial threat appraisal mechanism (Witte, 1992), and further suggest that threatening visuals serve as a heuristic cue motivating increased processing of message content (see Eagly & Chaiken, 1993), rather than an influence via a peripheral route as suggested by the ELM (Petty & Cacioppo, 1986). Interactions between drinking level and announcer gender and vocal quality variables were also found, but contributed only trivially to explained variance.