Affective valence in posts on X (formerly Twitter) and evaluation of a politician’s image

Abstract

The aim of this study was to examine whether the use of affective language by politicians impacts the social evaluation of their image, as measured by the semantic differential method developed by Cwalina et al. In the study, Polish participants (N = 958) evaluated the profiles of three well-known Polish politicians from different parties: Bosak (right-wing), Trzaskowski (center), and Biedroń (left-wing). Participants made evaluations before and after reading hateful, kind or neutral tweets about refugees from Ukraine. We controlled for participants’ political views and demographic variables. The results confirmed changes in evaluations of the image of politicians depending on the language they used. The ‘positivity bias’ identified in the study provides encouragement to use kindness speech in public discourse, showing that this type of communication can improve a politician's ratings.