distraction, received marginal support. Distraction was hypothesized to enhance persuasiveness, perceived realism, similarity, attractiveness, and identification, and reduce distrust and
counterarguing. Because distraction had no multivariate effect, its effect on the individual dependent variables was not examined. That said, in the stepdown analysis, controlling for its effects on the higher priority dependent variables (persuasiveness, distrust, identification,
attractiveness, realism, and similarity), distraction had a marginally significant (p < .01) effect on counterarguing, such that it was reduced in the distraction condition.
The effects of distraction can be complex. Research suggests that when a message is strong, distraction tends to interfere with persuasion (i.e., by reducing favorable thoughts), but when a message is weak, distraction can enhance persuasion (i.e., by reducing counterarguing;
e.g., Petty & Brock, 1981). Whereas research on distraction usually manipulates message strength, an attempt was made to control message strength in the present research by selecting ads with equivalent strength. However, the results suggest some ads were stronger and more persuasive than others. To examine the possibility that the effects of distraction differed depending on the perceived persuasiveness of the ad, a series of follow-up analyses was
conducted. The results of these analyses suggest distraction did not interact with persuasiveness in a linear fashion, but when the persuasiveness variable was divided into three groups of
roughly equal size, an interesting pattern emerged. Among the two-thirds of respondents who found the target ad moderately or highly persuasive (mean ratings of 4 or higher), distraction had no significant effect on the valence of thoughts. However, among those who found the ad
relatively unpersuasive (mean ratings less than 4), distraction significantly reduced counterarguing.
Research Question 1b
Research question 1b asked how, if at all, barriers to motivation and processing ability interact in their effects on the dependent variables.
Hypotheses 4 through 6: Interactions. Distraction was hypothesized to have a greater effect on relevant dependent variables for the informational ad. Additionally, the effect of the vulnerability manipulation was expected to be most pronounced when ability was not
compromised (informational ad, no distraction). None of the two-way interactions was significant in the multivariate analysis. As such, their effects on the individual dependent variables were not evaluated.!
Mediators and Moderators of Persuasiveness Research Question 2a
Research question 2a addressed potential mediators and moderators of the relationships between manipulated vulnerability, distrust, and persuasiveness.
Hypotheses 7 and 8: Mediation. Perceptions of persuasibility and third-person effects (general and ad-specific) were hypothesized to mediate the effect of the vulnerability
manipulation on persuasiveness. Although participants in the demonstrated vulnerability condition rated the ad more persuasive, this effect remained after accounting for the potential mediators. That is, the effect of the vulnerability manipulation on persuasiveness was not fully
mediated by perceptions of persuasibility, perceptions of personal vulnerability to media messages in general, or perceptions of vulnerability to the target ad specifically.
General third-person effects were not influenced by the vulnerability manipulation, but had a counterintuitive negative relationship with persuasiveness, indicating that greater relative perceived vulnerability to media messages was associated with greater persuasiveness. Ad-specific third-person effects were influenced by the vulnerability manipulation, but not in the expected direction. Individuals whose vulnerability was demonstrated rated themselves less affected by the target ad than those in the control condition. Like general third-person effects, ad-specific third-person effects had a negative relationship with persuasiveness, such that greater perceived vulnerability to the target ad was associated with greater persuasiveness. Because both indirect paths (paths a and b) were significant and the direct effect of the vulnerability
manipulation on persuasiveness (path c’) was closer to zero than the total effect (path c), ad-specific third-person effects partially mediated the effect of vulnerability on persuasiveness.
However, paths a and b were opposite to the directions hypothesized. These findings suggest that perceptions of vulnerability to the target ad worked against effect of the vulnerability
manipulation, but the effect nonetheless remained strongly negative. In other words, the vulnerability manipulation had the expected negative effect on persuasiveness in spite of its effect on perceived vulnerability, not because of it.
A follow-up analysis was conducted to explore why the vulnerability manipulation achieved the desired effect (reduced persuasiveness) if it did not work through the hypothesized mechanism. The results of this post hoc analysis suggest the valence of generated thoughts partially mediated the effect of the vulnerability manipulation on persuasiveness. That is, the
vulnerability condition elicited less favorable thoughts in response to the target ad, which was in turn associated with reduced persuasiveness.
Hypotheses 9 and 10: Moderation. Perceptions of persuasibility and third-person effects (general and ad-specific) were hypothesized to moderate the relationship between distrust and persuasiveness. These hypotheses were partially supported. As anticipated, distrust was negatively associated with persuasiveness. The strength of this relationship was not influenced by perceived persuasibility or general third-person effects, but ad-specific third-person effects did attenuate the relationship. As hypothesized, the negative relationship between distrust and persuasiveness was strongest when third-person effects were low or negative. In other words, distrust, or the perception of manipulative intent, was more strongly related to resistance among those who perceived themselves to be vulnerable to the target ad.
Hypothesis 11: Mediation. Thought valence was hypothesized to partially mediate the relationship between perceptions of manipulative intent (i.e., distrust) and persuasiveness. This hypothesis was supported. Distrust was associated with less favorable thoughts, which was in turn associated with reduced persuasiveness. All paths in the mediation model were statistically significant. The direct path from distrust to persuasiveness remained significant after accounting for the indirect path through thought valence, but was substantially reduced, indicating partial mediation.
Research Question 2b
Research question 2b addressed the extent to which the mediated relationship described under Hypothesis 11 is affected by variations in ability.
Hypothesis 12: Moderated mediation. Thought valence was hypothesized to mediate