Different Positive Feelings Leading to Different Ad Evaluations: The Case of Coziness, Excitement, and Romance

This study contributes to the debate about the valence-based versus the multidimensional view of feelings. The differential impact of three different contextand ad-induced feelings on ad effectiveness was compared. Support for the multidimensional view of feelings was found in the sense that adand context-evoked coziness, excitement, and romance had a different impact on ad evaluations. In addition, a significant interaction effect between adand context induced feelings indicated that ads that were exciting, romantic, and cozy scored best in a feeling-congruent context. Beginning in the 1980s, several researchers have demonstrated that, next to cognitions, feelings could also play an important role in the formation of judgments on advertising (Batra and Ray 1986; Burke and Edell 1989; Fabrigar and Petty 1999; Geuens and De Pelsmacker 1998; Holbrook and Batra 1987; Holbrook and O'Shaughnessy 1984; Stout and Leckenby 1986; etc.). Not only should the feelings induced by an ad itself be taken into account; the feelings people experience before watching ads are also important. The latter can be caused by many dif ferent things, such as the program, the article or ad pod in which the ad is embedded, or an event that took place before watching the ad. In the remainder of this paper, we refer to these emotions as Reviewing the literature, it seems that most researchers who have attempted to answer the question of how context-induced or ad-evoked feelings influence our thinking and judgment have focused on the comparison between the effect of versus feelings (e.g., Ayelsworth and MacKenzie 1998; Goldberg and Gorn 1987; Kamins, Marks, and Skin ner 1991). However, more and more researchers (e.g., Babin, Darden, and Babin 1998; Mitchell et al. 2001; Raghunathan and Pham 1999) are convinced that the valence dimension does not suffice and more dimensions need to be taken into account to understand the differential effects of different emotions of the same valence. We would like to contribute to the debate on the merits of the unidimensional (valence-based) versus multidimensional views of feelings by investigating whether three contextand ad-induced feelings of the same valence, namely, coziness, romance, and excitement, have a different impact on ad evaluations. A second objective is to study the interaction effects between context-induced and ad-evoked feelings. Impact of different feelings of the same valence Valence-Based Versus Multidimensional Views of Feelings The majority of the studies on the impact of and feelings on ad evaluation are based on the assumption that all feelings are bipolar and can be placed on a continuum going from positive at one end to negative at the other (e.g., Green, Goldman, and Salovey 1993; Russell and Carroll 1999). One important implication of this valence-based view of feelings is that different feelings of the same valence, such as sadness, anger, and fear, or coziness, romance, and excitement, would exert similar influences on judgment and choice. Recently, however, an increasing number of researchers (e.g., Babin, Darden, and Babin 1998; Mitchell et al. 2001; Raghunathan and Pham 1999) have questioned the assump tion that all and all feelings are equal and lead to similar effects on information processing and judgment formation. In their view, dimensions other than valence alone should be considered in accounting for the specificity of dif ferent emotions. With respect to context-induced feelings, for example, Mitchell et al. (2001) investigated the impact of Tine Faseur (master's in marketing analysis and planning, Ghent University, Belgium) is a Ph.D. student at Ghent University. Maggie Geuens (Ph.D., University of Antwerp, Belgium) is an assistant professor of marketing, Ghent University, Vlerick Leuven Gent Management School. The authors thank the Editor, Russell N. Laczniak, and three anonymous reviewers for their constructive remarks and suggestions on earlier versions of this manuscript. Journal of Advertising, vol. 35, no. 4 (Winter 2006), pp. 129-142. ? 2006 American Academy of Advertising. All rights reserved. ISSN 0091-3367 / 2006 $9.50 + 0.00. DOI 10.2753/JOA0091-3367350409 This content downloaded from 157.55.39.127 on Wed, 29 Jun 2016 04:47:22 UTC All use subject to http://about.jstor.org/terms 130 The Journal of Advertising context-induced happiness, sadness, and anger on the effective ness of a neutral ad. They concluded that the feelings of anger and sadness lead to different effects, although these effects are more similar to one another than they are to the feeling of happiness. Raghunathan and Pham (1999) investigated the impact of anxious, sad, and neutral feelings, induced by means of an empathy task, on decision making with respect to gambling and job selection. They found that affective states of the same valence have a different impact, in the sense that and sadness convey distinct types of information to the decision-maker and prime different goals. While anxiety primes an implicit goal of uncertainty reduction, sadness primes an implicit goal of reward replacement (Raghunathan and Pham 1999, p. 56). Anxious people were thus more inclined to choose for low-risk/low-reward options, whereas sad people preferred the high-risk/high-reward options. Also with respect to ad-evoked feelings, there are indica tions that not all feelings of the same valence lead to similar ef fects. According to Batra and Ray (1986), affective responses to advertisements can be divided into three different categories of feelings, namely, (1) surgency, elation, vigor/activation (SEVA), (2) deactivation, and (3) social affection. These cat egories do not appear to be equally important in the formation of attitudes and purchase intentions. Burke and Edell (1989) classified 54 ad-evoked feelings into three main categories, which they called upbeat, warm, and feelings. Each feeling category had a different influence on attitudes through its distinct relation with different cognitive judgments. Previous studies provide some initial evidence that differ ent feelings of the same valence can lead to different effects. This would imply that we cannot extrapolate the known effects of, for example, happiness and sadness to all and feelings, respectively. As Lerner and Keltner say: Valence-based approaches may sacrifice specificity in the ser vice of parsimony (2000, pp. 475-476). In other words, the valence-based view of emotions may oversimplify emotional experience and may conceal the distinctions among or emotions.