Effects of Narratives and Affects in Digital Headlines on User Responses

The way in which a headline is packaged and presented affects audiences’ attentive, cognitive, and emotional responses, even when the substance of the headline remains the same. By data mining digital headlines and associated user response metrics from three major video-sharing websites in China as well as conducting subsequent content analysis, this study found that liking, sharing, and commenting were associated with headline viewing or attention. Moreover, both the main and interaction effects of narrative (vs. numerical) evidence and negative affect on viewing were significant. Negative affect moderated the mediating effects of viewing on the relationships between narrative (vs. numerical) evidence and liking, sharing, and commenting. In addition, all of these effects were significant, even after controlling for the effects of the word length of the headline, source influence, and differences between publication and scraping dates. Based on these findings, a new concept called the “Negative Narrative Effect” (NNE) is introduced. The implications of these findings for data journalism are also discussed.