Gender-Related Differences in Online Comment Sections: Findings From a Large-Scale Content Analysis of Commenting Behavior. (June 2023)

• Record Type: Journal Article
• Authors: Küchler, Constanze; Stoll, Anke; Ziegele, Marc; Naab, Teresa K.
• Is Part Of: Social science computer review. Volume 41:Number 3(2023)
• Page Start: 728
• Page End: 747
• Publication Date: 2023-06
• Subjects: online discussions -- user comments -- incivility -- gender -- discrimination -- machine learning -- word embeddings -- automated content analysis -- multilevel modeling; Social sciences -- Data processing -- Periodicals; Computers -- Social aspects -- Periodicals; Microcomputers -- Periodicals; Sciences sociales -- Informatique -- Périodiques; Micro-ordinateurs -- Périodiques; 300.285
• DOI: 10.1177/08944393211052042
• Languages: English
• ISSNs: 0894-4393

Abstract

Comment sections below news articles are public fora in which potentially everyone can engage in equal and fair discussions on political and social issues. Yet, empirical studies have reported that many comment sections are spaces of selective participation, discrimination, and verbal abuse. The current study complements these findings by analyzing gender-related differences in participation and incivility. It uses a sample of 303, 342 user comments from 14 German news media Facebook pages. We compare participation rates of female and male users as well as associations between the users' gender, the incivility of their comments, and the incivility of the adjacent replies. To determine the incivility of the comments, we developed a Supervised Machine Learning Model (classifier) using pre-trained word embeddings and word// frequency features. The findings show that, overall, women participate less than men. Comments written by female authors are more civil than comments written by male authors. Women's comments do not receive more uncivil replies than men's comments and women are not punished disproportionately for communicating uncivilly. These findings contribute to the discourse on gender-related differences in online comment sections and provide insights into the dynamics of online discussions.