Gendered Tweets: Computational Text Analysis of Gender Differences in Political Discussion on Twitter

Lingshu Hu,Michael Wayne Kearney
DOI: https://doi.org/10.1177/0261927X20969752
IF: 2.098
2020-11-06
Journal of Language and Social Psychology
Abstract:Politics is an area that is traditionally believed to be gender divided. According to intergroup communication theory, this gender-salient context might cause differences in political communications between genders. Moreover, the internet and social media, which creates a computer-mediated interactive context, might also impact the traditional gender discrepancies in political discourse. This study used Twitter trace-data and computational text analysis to examine such suppositions. By analyzing over one million tweets, we found that compared to men, women generally had a stronger sense of group awareness and cohesion and showed a desire to promote their tweets while avoiding addressing other users in political discussions. Women also focused on family- and home-related issues more than men did. These findings suggest that Twitter is not an ideal public sphere where differences and inequalities are eliminated, but it might be a counter-public sphere that promotes the voices and increases the publicity of marginalized groups.
communication,linguistics,psychology, social
What problem does this paper attempt to address?