Impact of discrete emotions on audience engagement with climatechange videos on Chinese TikTok (Douyin)

Abstract

In this study we explored the effect of the discrete emotions of short-video presenters on audience engagement with climate-change videos on Chinese TikTok (Douyin). We performed an automated content analysis of 510 science communication videos, with a focus on the emotional expression of the presenter and interactions of emotions. Our findings revealed that the negative emotion of anger dominated climate-change videos and significantly influenced audience engagement. To our surprise, the presence of the emotion of happiness in the audio led to decreased engagement. We also uncovered the critical role of complex emotional interactions in affecting audience engagement, with combinations of positive and negative emotions being particularly influential. Furthermore, we employed facial emotion recognition and the SpeechBrain machine-learning technique for emotion detection in the videos, a methodological approach that is innovative and ensures objective and accurate analysis, as well as offering new possibilities for research.