Clickbait for climate change: comparing emotions in headlines and full-texts and their engagement

Abstract

ABSTRACT Anthropogenic climate change remains a polarizing topic. As most social media users share articles solely relying on the headline, this raises the question of how emerging digital media reporting – especially in the headlines – shapes the perception of climate change issues and engages audiences. Guided by the dual-systems emotion model and discrete-emotions model, this study compared emotion words used in headlines versus full text among climate change articles – and their social media engagement, using computational methods. Findings suggested that climate change support headlines were more likely to use fear words while denial headlines were significantly more likely to contain emotion words, negatively-valenced words, as well as words for anger, anticipation, disgust, sadness, and surprise. Regarding the full text, denial articles were more likely to contain emotion words, negatively-valenced words, and many discrete emotions related words than support articles. A denial article’s engagement was predicted by the total number of emotion words contained in its headline, whereas a support article’s engagement was predicted by negatively-valenced words and words for fear used in its headline. Emotions contained in the full text did not predict support and -denial articles’ engagement. Findings provide practical guidance on how to increase the engagement level of climate change articles.