TL;DR
This study investigates how critical online communities in Germany, particularly on Telegram, engaged with mainstream information during the COVID-19 vaccination discourse. By comparing Telegram, X (formerly Twitter), and news data, the researchers found that while mainstream sources were present in these critical communities, their role and prevalence differed significantly from X, suggesting "information voids" that complicate infodemic management.
What Problem Does This Paper Address?
The COVID-19 pandemic highlighted the critical role of information in public health crises, leading to an "infodemic" – an overwhelming supply of information of varying quality. This can lead individuals to struggle with making informed health decisions and can undermine public health measures like vaccination. While the spread of misinformation on platforms like Telegram has been studied, the role of mainstream information (from government and established media) within these critical communities remained unclear. This paper aims to bridge this gap by understanding how mainstream information is used, shared, and perceived in these often-opposing online spaces, providing context for managing future infodemics.
Methods
The researchers employed a mixed-methods approach, combining quantitative and qualitative analyses across three platforms: Telegram, X, and online news.
- Text Clustering (BERTopic): Used to identify popular topics within the vaccination discourse on Telegram and X.
- Keyword-based Classification: A two-stage filtering process to specifically identify posts related to vaccination and then to categorize them into five key subtopics (death, long COVID, mandatory vaccination, measures in schools, virus variants).
- Link Sharing Analysis: Quantified the prevalence of mainstream information by measuring the frequency of shared links from government and major media outlets.
- Correlation-based Time Series Analysis: Examined the similarity in discourse dynamics between Telegram, X, and mainstream news over time to understand how synchronized their discussions were.
- Thematic Analysis: Qualitatively analyzed a sample of posts to understand the underlying reasons and motivations for sharing mainstream information within these communities.
Datasets
The study utilized the following datasets:
- Telegram: Public posts from channels linked to the "Querdenken" protests and the far-right, collected via the Telegram API.
- X (formerly Twitter): A 1% sample of all public posts in the German language, streamed directly from the X API.
- News Headlines: Data from the Global Embedded Metadata Graph (GDELT project), covering news articles from various global sources.
The time period for data collection was April 2019 to February 2023, chosen to align with the availability of the X data.
Key Insights and Ideas
- Platform Differences Matter: The study strongly suggests that Telegram and X, despite both being social media platforms, host distinct information ecosystems with different engagement patterns regarding mainstream sources.
- "Information Voids" as a Driver: The researchers propose that differences in how mainstream information is used across platforms might indicate "information voids" – situations where reliable information is insufficient, creating opportunities for alternative or misleading content.
- Nuanced Engagement with Mainstream Sources: Critical communities don't always outright reject mainstream information. They may use it for various purposes, including criticism, but also for understanding developments.
- Context is Crucial for Risk Communication: Effective communication during health crises needs to go beyond simply disseminating information; it must understand how different communities interpret and utilize that information.
Key Results
- Prevalence of Mainstream Links: On average, 58% of Telegram posts and 21% of X posts contained an external link. Of these, 11%-35% of Telegram posts and 44%-60% of X posts linked to mainstream sources.
- Discourse Dynamics: Correlation analysis showed varying degrees of synchronization between Telegram, X, and mainstream news. Mandatory vaccination and virus variants showed strong positive correlations, indicating synchronized discourse, while long COVID showed weak associations.
- Thematic Analysis of Sharing: Five themes emerged for why mainstream information was shared:
- Observing News (Telegram: 29%, X: 21%): Sharing links with minimal commentary.
- News Comments (Telegram: 35%, X: 35%): Commenting on shared content without questioning its credibility.
- Directed Accusations (Telegram: 29%, X: 12%): Criticizing media or authorities.
- Participation (Telegram: 7%, X: 4%): Inviting engagement in events or petitions.
- Reference in Discussion (X: 24%): Using sources to support claims within discussions (exclusive to X).
Limitations and Future Work
- Approximation of Mainstream Sources: The study approximated mainstream sources by their popularity, acknowledging that a more nuanced examination of political positions within the media landscape could strengthen findings.
- Single Network Assumption for Telegram: The analysis treated Telegram communities as a single network, and future work could explore sub-communities for a more detailed breakdown.
- Limited Scope of Topics: While focusing on popular vaccination-related topics, the discourse was more complex and included other themes.
- Future Directions: The authors suggest extending the framework to other social media platforms for a comprehensive view, investigating information pathways within social media communities, and exploring how political positions within the mainstream information ecosystem differ. They also highlight the need for tailored strategies to address diverse information needs and build trust in risk communication.
Summary updated