Moral Values Underpinning COVID-19 Online Communication Patterns

TL;DR

This study investigates how moral foundations shape COVID-19 discussions on Twitter over nearly two years by combining Moral Foundation Theory with large-scale network analysis. It detects 10 moral signals per tweet, derives four distinct user groups via Social-LLM embeddings, and assigns political leaning to explore morality-politics dynamics. The findings reveal strong moral homophily for care, fairness, and authority, with one group (Group IV) forming a potential moral echo chamber, while messages high in moral diversity or pluralism can more effectively bridge divides across groups. The work advances understanding of how moral foundations influence online discourse and network structure, offering actionable insights for health communication and design of diverse, dialog-friendly online ecosystems.

Abstract

The COVID-19 pandemic has triggered profound societal changes, extending beyond its health impacts to the moralization of behaviors. Leveraging insights from moral psychology, this study delves into the moral fabric shaping online discussions surrounding COVID-19 over a span of nearly two years. Our investigation identifies four distinct user groups characterized by differences in morality, political ideology, and communication styles. We underscore the intricate relationship between moral differences and political ideologies, revealing a nuanced picture where moral orientations do not rigidly separate users politically. Furthermore, we uncover patterns of moral homophily within the social network, highlighting the existence of one potential moral echo chamber. Analyzing the moral themes embedded in messages, we observe that messages featuring moral foundations not typically favored by their authors, as well as those incorporating multiple moral foundations, resonate more effectively with out-group members. This research contributes valuable insights into the complex interplay between moral foundations, communication dynamics, and network structures on Twitter.

Figures (8)

• Figure 1: Distribution of raw user moral scores.
• Figure 2: The average moral z-scores of each foundation for the four user groups.
• Figure 3: Partisanship breakdown of the four user groups. Blue bars represent left-leaning users and red bars represent right-leaning users
• Figure 4: Distribution of the user metadata features for the four user groups.
• Figure 5: TSNE visualization of 100,000 sampled user embeddings of the four user groups.