How Message Fatigue toward Health Messages Leads to Ineffective Persuasive Outcomes: Examining the Mediating Roles of Reactance and Inattention

Abstract

Message fatigue refers to a state of being exhausted and tired of prolonged exposure to similarly-themed messages (e.g., anti-obesity messages; So, Kim, & Cohen, 2017). This study tests a mediational model that accounts for how one’s preexisting fatigue toward anti-obesity messages may contribute to two different types of resistance—reactance and disengagement—toward an incoming anti-obesity message, which, in turn, reduce intention to adopt weight-management behaviors advocated in the message. The proposed model was tested in an experimental study (N = 312) involving a sample of overweight or obese adults in the United States. In the meditational model, reactance significantly mediated the negative effects of message fatigue on intention to adopt only one of four weight-management behaviors promoted in the message. However, inattention, which was an operationalization of disengagement, significantly mediated the negative effects of message fatigue on behavioral intention to adopt all four weight-management behaviors. This study urges future research on message fatigue and resistance to persuasion to consider disengagement with a message as a significant barrier to effective health communication and to devise ways to increase engagement with messages communicating “overtaught” health issues.

Notes

¹ Threat to freedom was not manipulated in the stimuli because this study did not purport to examine the effects of reactance induced by message features (e.g., forceful language; Quick & Considine, 2008). Since the message was not manipulated to induce variations in perceived threat to freedom, induction check for perceived threat to freedom, as it is typically done in reactance studies, is not relevant or feasible. That said, we wanted to check our central assumption that fatigued individuals would consider an exposure to another message they are tired of hearing as threat to freedom. We modified the two-step process model of reactance (Quick & Considine, 2008) to reflect the context involving message fatigue as the antecedent to resistance in message processing. Since pre-existing anti-obesity message fatigue was postulated as an antecedent to perceived threat to freedom upon exposure to an anti-obesity message, we expected those who have higher levels of pre-existing anti-obesity message fatigue to be more likely to perceive exposure to another anti-obesity message as greater threat to freedom (i.e., step 1). The aroused perception of threat to freedom would, in turn, lead to reactance and inattention, depending on where one locates the blame for the threatened freedom (i.e., message source or message environment; step 2). The modified two-step model showed an acceptable fit to data: χ²(11) = 28.48, p < 0.001, RMSEA = 0.07, (0.04, 0.11), CFI = 0.92, TLI = 0.86, SRMR = 0.05. As predicted, message fatigue, was positively associated with perceived threat to freedom (β = 0.29, p < 0.001; step 1), which, in turn, was positively associated with reactance (β = 0.78, p < 0.001) and inattention (β = 0.22, p = 0.002; step 2). Thus, the assumption that greater anti-obesity message fatigue would lead individuals to perceive an exposure to an anti-obesity message as threat to freedom received empirical support.

² Since difference in experimental conditions is not the focus of the present study, a one-way MANOVA was conducted to determine if the experimental conditions had effects on inattention, counterargument, anger, and four behavioral intention variables. No significant difference was found between the experimental and control groups on the combined dependent variables, Wilks’s λ = 0.98, F(7, 300) = 1.03, p = 0.41. The results of a series of independent samples t-tests showed that participants in the experimental groups reported lower levels of intention to use small plates (M =5.11, SD = 1.91) than those in the control group (M = 5.56, SD = 1.43, t(118.20) = 2.07, p = 0.04). Thus, use of data from all five groups was deemed justifiable as long as the effects of the presence a frame message on intention to use small plates were controlled for in the analyses. Age and gender were also included as covariates (see Table 1 for the correlation matrix) in primary analyses.

Additional information

Funding

This work was supported by the University of Georgia Research Foundation [grant number 2964]