Mood and Reactions to Political Advertising: A Test and Extension of the Hedonic Contingency Hypothesis

Abstract

Research in a variety of contexts has shown that the mood or emotional state of receivers may affect responses to communication messages. This study tested the interaction between mood and message on persuasiveness in the context of political television advertising. Using a 2 (happy or sad mood) by 2 (hedonic consequences: negative or positive) experimental design, a two-way interaction between hedonic consequence and mood was predicted, such that happy people would be more persuaded by positive ads than negative ads. In addition, sad people, it was expected, would be equally persuaded by either ad. Cognitive processing measures, attitudes, and candidate evaluations were used for measuring effects. Data indicated that happy people were more persuaded by positive political advertisements than negative political advertisements. Sad people, however, were equally persuaded by both negative and positive appeals.

Notes

Note. Means with same subscripts indicate statistically significant differences at the p < .05 level. These differences are based on 95% confidence intervals.

We agree whole heartedly with an anonymous reviewer who pointed out that we should have reported Scott's Pi (or Cohen's Kappa) for the assessment of intercoder reliability for both the recall and the cognition data. Unfortunately, the third author trained the coders and held the coding data for this study. She passed away in the midst of our study write-up and we do not have access to the raw data. Our collaborator provided us with these statistics prior to her untimely death.

We want to thank an anonymous reviewer for suggesting we control for party preference of the message recipient. We ran each of the above dependent variables through an ANCOVA with mood and hedonic consequences as the independent variables and self-reported political party (Democrat, Republican, or Independent) as the covariate. In all cases, the covariate was not statistically significant, revealing that these data are independent of party preference.

Additional information

Notes on contributors

Monique Mitchell Turner

Monique Mitchell Turner, Department of Prevention and Community Health, The George Washington University.

Jill Cornelius Underhill

Jill Cornelius Underhill, Department of Communication Studies, Marshall University.

Lynda Lee Kaid

Lynda Lee Kaid, College of Journalism and Communications, University of Florida.