Sentiment, Political Identity and Engagement Dynamics in 2024 U.S. Election Discourse on X

Abstract

Nigar Garajamirli

Political communication in election times is increasingly influenced by emotional displays, partisan affiliations, and platform-centric engagement dynamics. The 2024 U.S. presidential election fostered a highly divided online atmosphere where discussions about Donald Trump attracted significant public interest and engagement. This research examines the interplay of sentiment, political identity, thematic communication trends, and audience interaction in Trump-focused political discussions on social media. Employing a computational social science approach, the research examines a dataset comprising 18,172 political tweets linked to two political communities engaged in conversations regarding the 2024 U.S. presidential election. The analytical method integrates sentiment analysis, descriptive statistics, inferential testing, correlation analysis, regression modeling, and comparative TF-IDF text evaluation. The results indicate statistically significant variations in sentiment among political groups, with one group demonstrating a consistently more positive emotional expression in political discussions. Nonetheless, analysis of effect size shows that these differences are still fairly minor in practical terms. Sentiment showed restricted predictive capability for engagement outcomes, whereas political affiliation and message traits displayed somewhat stronger—yet still modest—connections with likes and retweet actions. Furthermore, thematic analysis revealed significant lexical variation among political communities, indicating the presence of politically divided communication contexts centered around unique actors, narratives, and ideological frameworks. In general, the results indicate that political identity has a greater influence on engagement patterns than just emotional tone. The findings emphasize the intricacies of political communication related to Trump- centric discussions and show that online participation during elections is influenced by various interrelated elements beyond mere sentiment.

PDF