The Mediating Role of Message Engagement in the Extended Parallel Process Model

Theoretical Background and research questions/hypothesis:  Although fifty years of research has yielded mixed results, fear appeals remain one of the most common strategies to modify behavior in public health campaigns. While some studies substantiate the effectiveness of fear appeals (e.g., Beck, 1984, Insko, Arkoff, & Insko, 1965; Stainback & Rogers, 1983), others demonstrate their ineffectiveness (e.g., Janis & Feshbach, 1953; Kohn et al., 1982; Krisher, Darley, & Darley, 1973) and still others document mixed results (e.g., Hill & Gardner, 1980, Rogers & Mewborn, 1976, Witte, 1992). The Extended Parallel Process Model, the most recent model to describe how, why, and under what circumstances fear appeals work, posits that when individuals see a fear appeal, they first evaluate the severity of the threat and their susceptibility to it. If both are considered moderate to high, they then judge the efficacy of the recommended response by asking, first, how well it works (response efficacy) and second, if they can perform it easily (self-efficacy). If both threat and efficacy are rated highly, the appeal is considered effective and behavior change is likely. But, if the threat is too severe, individuals will not move on to a judgment of the recommended remedy and the message will be ignored or defensively avoided.

Methods:  Capitalizing on the H1N1 flu pandemic, this study examined the role of message engagement in the EPPM model by exploring the relationships among perceived threat, perceived efficacy, severity, susceptibility, self-efficacy, response efficacy,  message engagement and behavioral outcomes. After exposing 180 participants to an H1N1 fear appeal, evaluations of the threat, efficacy and engagement of the message, as well as intentions of making behavioral change (e.g. likelihood of getting vaccinated, protective behaviors, talking to others) were self-reported. Engagement was measured with fifteen questions about how much the respondent concentrated on and exerted mental effort while exposed to the message, as well as the evaluation of the credibility, effectiveness, persuasiveness, and intrinsic interestingness of the public service announcement.

Results: The results show that message engagement mediates the relationship between severity, self-protective behaviors, taking precautions and subsequent behavioral outcomes. Also, severity and response efficacy were related to likelihood of vaccination and engagement, and severity and response efficacy were found to be co-predictors of vaccination likelihood.

Conclusions:  Some studies have shown a positive linear relationship between fear and attitude, behavioral intention and behavioral change (Nabi, Roskos-Ewoldsen, & Dillman Carpenter, 2008; Sutton, 1982) while others have shown a curvilinear relationship where, if the level of fear is raised too high, the message will be rejected (Witte, 1994; Witte & Allen, 2000).The results of this study suggest that if a message is emotionally engaging, it can be extremely fearful and still be effective in eliciting behavioral intentions, potentially stretching the critical point at which a message is considered effective.

Implications for research and/or practice: Rather than attempting to pinpoint and balance unpredictable levels of fear and efficacy, creators of fear appeals should first consider the engagement of the message itself. Practical and theoretical implications are discussed.