Data_Sheet_1.XLSX

Abstract

Owing to the proliferation of rumors on social media, it is necessary to disseminate debunking information to minimize the harm caused by rumors. Using content analysis, sentiment analysis, and regression analysis, this study examined the mediating role of follower count in the relationship between the debunker’s identity and sharing behavior and explored the relationship between the text sentiment of debunking information and sharing behavior based on data on the spread of three rumors that circulated extensively on social media. Using an ordinary account as the reference, we found that the mediating or suppression effect (i.e., when direct and indirect effects are significant and opposite) of follower count in the relationship between debunker’s identity (celebrity, media, or government) and sharing behavior was significant. The three test identities (celebrity, media, and government) had more followers than the ordinary account, which resulted in a significant positive effect on the number of reposts. Debunker’s identity did not have a positive effect on the sharing of debunking information when controlling for mediating variables. Debunking information with emotional overtones (positive or negative) was shared more widely compared with information with neutral emotions, and the dominant emotional polarity was different in the three different rumors. These findings contribute to the generation of debunking information content, which can aid in the development of effective communication strategies and improvement in the efficiency of crisis management.