Subject
The effect of feelings generated by an ad on various measures of advertising effectiveness.

Type of Report
Report of empirical findings from the “Feelings Mechanism” project.

Objective
To examine the role of feelings generated by an advertisement in explaining consumers’ responses to the ad. The research contrasts the effect of consumers’ affective reactions to ads with the effect of their cognitive reactions and demonstrates that when assessing advertising effectiveness, it is important to measure both types of reactions.

Method
Reviews recent literature on cognitive and emotional responses to ads; demonstrates experimentally that consumers’ feelings about an ad are distinct from their semantic judgments of the ad’s characteristics; develops a model of both the direct and indirect effects of feelings on attitude toward the brand; empirically tests the model in an experiment that varies the number of ad exposures and the length of delay between final exposure and response measurement.

Audience
Advertising researchers in both industry and academia, as well as marketing managers.

Main Points
Historically, advertising-effectiveness research concentrated on the evaluative component of attitude. Subjects usually judged ads semantically by stating how well descriptive adjectives characterize an ad, provided written or verbal reactions to an ad in the form of cognitive responses, or both. While these procedures result in useful information about what subjects learn from ads, there was a concern that such “cognitive” procedures may miss entirely how the ad makes the subject feel. In his 1985 address to the Association for Consumer Research, Peter Wright framed the issue succinctly. He said that when asked to answer questions such as “Was that ad convincing?” respondents may be “responding as critics, not as audience members” (emphasis added). This project contrasts consumers “ad critic” responses to ads with the feelings that they personally experience in response to an ad.

The Power of Feelings
The Feelings Mechanism project set out to determine whether feelings generated by an ad were distinct from thoughts about the ad and whether such feelings represented a qualitatively different dimension of attitude. An important finding is that feelings contribute uniquely to attitude toward the ad, beliefs about the brand’s attributes, and attitude toward the brand. Further, the relative importance of feelings and judgments of the ad’s characteristics varies, based on the extent to which the ad is transformational and informational. In addition, scales representing three dimensions of feelings: Upbeat Feelings, Warm Feelings, and Negative Feelings; and three dimensions of judgments of the ad’s characteristics: Evaluation Judgments, Activity Judgments, and Gentleness Judgments–were developed using factor analysis. (Details of the scale development are included as an appendix with this report). An important finding is that positive and negative feelings are independent, i.e., they form separate factors. Thus, positive and negative feelings can co-occur. Positive and negative evaluations of the ad’s characteristics, however, are not independent. While an ad can make a viewer feel both happy and sad, it will not be judged as both good and bad.

This first part of the Feelings Mechanism project demonstrated the power of feelings in understanding advertising effects and resulted in a conceptual model of the role of feelings that includes effects on both affective and cognitive measures of advertising effectiveness.

The Impact of Feelings
The second part of the Feelings Mechanism project investigates the relationships among feelings experienced in response to an ad, judgments of the ad’s characteristics, brand attribute evaluations, attitude toward the ad, and attitude toward the brand. The present report focuses particularly on these relationships.

A model was developed that specified feelings would influence each of the other constructs, either directly or indirectly. The model was tested using data gathered in an experiment that varied how many times subjects saw new ads and how much time elapsed between exposure to the ads, which were embedded in programming, and measurement of the subjects’ reactions to the ads. Factor analysis of the subjects’ reactions produced in the same feelings scales found earlier Upbeat Feelings, Warm Feelings, and Negative Feelings; and the same judgment scales: Evaluation, Activity, and Gentleness.

Estimation of the model shows that Upbeat, Warm, and Negative feelings each have effects throughout the system. The particular manner in which each feeling influences attitude toward the brand varies. The most important finding is that both Upbeat Feelings and Negative Feelings have a direct effect on attitude toward the brand, demonstrating that the effects of feelings are not completely mediated by their effect on attitude toward the ad or brand attribute evaluations. These findings are robust with respect to multiple viewings of the ad, different measurement delay periods, and the particular ad being evaluated.

Implications
The feelings generated by an ad are important advertising responses that should be included in any battery of measures of advertising effects. Feelings are not only linked to the consumer’s attitude toward the ad; they are intricately linked to the consumer’s judgments of the ad’s characteristics, brand attribute evaluations, and attitude toward the brand as well.

This project adds to our understanding of how feelings influence certain advertising effectiveness measures. Upbeat, Warm, and Negative feelings have significant effects and influence both cognitive and the affective reactions, sometimes directly and sometimes indirectly. Thus, it is not enough merely to measure how an ad makes a subject feel in general. It is important to measure each dimension of feelings because the different feelings have countervailing influences that would be missed if a summary measure of feelings was used. Adding the three feelings scales to copy tests and other ad effectiveness research will add significantly to understanding how an ad works.

Researchers’ Comments
The researchers reflect:

“Our findings indicate that the effects are robust with respect to time. We are not sure, however, that the 10-exposure level used is sufficiently high to invoke repetition related changes in feelings. After many more exposures either in terms of total exposures or exposures per day or per hour–feelings may not have an effect, or their effect may be different than the ones identified here.

“An important aspect of the real world that must be considered in future research is that potential consumers of new products and services will eventually have more sources of information about a brand than just one particular ad. For instance, they may see different advertising executions, they may personally experience the product or service, or they may hear or read about the experience of others. Whether feelings evoked by an ad will exert the same influence as they did in our experiments in settings where other information is available remains to be established.”

About the Authors
Julie A. Edell and Marian Chapman Burke are Associate Professors at the Fuqua School of Business at Duke University.