Joy’s mediating effect on digital consumer engagement with advertising

From a computational perspective, advertisers stand the chance to benefit from a better understanding of a message’s framing effects on behavioral responses to an advertisement. Sentiment analysis, which relies on a text analytic process to determine the opinions, beliefs, and feelings of a speaker directed toward someone or something, has been shown to be useful in determining associations between words and discrete emotions. Surprisingly little research to date has examined the ad framing effects of emotion in an advertisement’s copy or the role that the emotion plays in relation to behavioral outcomes of engagement. Current research exploring discrete emotions tends to live in the news media, science communication, or health communication streams of literature with less attention given to how discrete emotions affect behavioral outcomes in advertising. Even fewer studies have considered the role of emotion in text, aside from examining how terms may evoke an emotion using survey-based methods. To address this gap, I introduce a text analytic process using a real-world advertising sample of Facebook advertisers. Using the NRC lexicon, I use natural language processing to measure the sentiment and emotional frames of a corporate brand advertising sample. Each advertisement was coded for a mean sentiment measure before being measured for discrete word associations of joy. Given the broader theoretical aim of this dissertation and the overabundance of literature focused on the negative effects of media, I focus on the positive emotion of joy and examine its mediating effects on digital consumer engagement with advertising. This dissertation contributes to advertising effectiveness literature, first by conceptualizing engagement behaviors alongside outcomes of advertising processing, and then by arguing for the importance of engagement as a measure of effectiveness in relation to emotions as a frame. My results extend previous emotion advertising literature by showing that joy-frames can mediate outcomes of advertising processing. Specifically, I show that words associated with joy in text mediate outcomes of clicks. Moreover, I show how word associations with positive, discrete emotions can mediate behavioral outcomes of digital consumer engagement with advertising and suggest how these findings can be used in a theoretical and applied manner.