How Emotional Media Reports Influence Attitude Formation and Change: The Interplay of Attitude Base, Attitude Certainty, and Persuasion

Abstract

This study examined the influence of an emotionally arousing writing style on attitude formation and change. It has been proposed that different writing styles induce attitudes based on either affect or cognition and with either high or low certainty. Previous work indicates that the interplay of these attitude characteristics determines the persuasiveness of emotional and rational media appeals. To test the hypotheses, participants in an experimental study read articles from a magazine about a fictitious attitude object. In the first step, 4 different types of attitudes varying in base and level of certainty were induced through a respectively manipulated article. In the second step, these attitudes were challenged by an additional article, which presented either an emotional or rational persuasive appeal. The results supported hypotheses on attitude induction through media stimuli and 3 of 4 hypotheses regarding the persuasiveness of emotionally and rationally written articles.

NOTES

Notes

1. The term “argument” carries a rational connotation and is therefore commonly used to describe rational information. In this article, we use the term to describe evidence, which can be presented using either rational or emotional language, speaking for or against an attitude object.

2. These participants indicated their suspiciousness about the cover story during debriefing. They did not, however, rate the articles to be of low external validity.

3. In a pretest (N = 11) participants rated all arguments used on three scales ranging from 1 to 5: argument valence (the argument is against Pure to the argument is in favor of Pure), argument tone (the argument is rational to the argument is emotional), and argument strength (the argument is not convincing to the argument is convincing). For the induction texts, we used only arguments rated in favor of Pure (M > 2.5). For the persuasive appeals, we used arguments that were rated as speaking against the product (M < 2.5). For emotional stimuli, we used arguments with an emotional orientation (M = 3.16), and for rational stimuli, we used arguments with rational orientation (M = 2.51). For the low certainty condition stimuli we used only arguments that were little or medium convincing (M < 3.7), whereas for the high certainty condition stimuli also highly convincing arguments were chosen.

Following the input of a reviewer, we conducted an additional test of argument strength (N = 42). In this test, participants rated all arguments on four items from the perceived argument strength scale by Zhao, Strasser, Capella, Lerman, and Fishbein (2011) (e.g., the statement regarding Pure is convincing, α = .92). In this test, the findings of our initial pretest were supported; that is, the arguments we used for the low certainty condition stimuli were rated as little or medium convincing (M < 2.5), whereas the arguments we included in the high certainty condition stimuli were rated as highly convincing (M > 2.5).

4. Alternatively, the hypothesis can be tested using the Time 2 attitude as the dependent variable and the Time 1 attitude as a covariate instead of using the attitude change index as the dependent variable. Likewise, this analysis revealed a significant three-way interaction between attitude base, attitude certainty, and writing style of the persuasive appeal (F(1, 156) = 8.56, p < .001, η_p² = .05).

5. The hypotheses were also tested controlling for the external validity ratings. Including the external validity ratings did not change the results in any way. Specifically, we still found a significant three-way interaction (F(1,155) = 12.89, p < .001, η_p² = .08) and support for Hypotheses 3a, 3b, and 3d.