Current Research :
What can we learn from conspiracy narratives?
1.12.2025
Conspiracy theories have long attracted attention, but most research treats them as harmful or irrational beliefs, focusing on misinformation, distrust, or polarization. Less explored is what these narratives can tell us about societal anxieties, structural inequalities, and crises of trust. My work asks what we can learn about society by analyzing the narrative structures of conspiracy theories, rather than simply pathologizing them.
Using a large-scale dataset of German-language Telegram channels, I identify and track specific conspiracy theories as they emerge and spread. I then apply graph-based narrative analysis to uncover the structure of these narratives, examining how global movements like QAnon and localized ones like Reichsbürger differ in their storytelling and social appeal. This approach highlights how different social, political, and cultural contexts shape the stories people are drawn to, and what these narratives reveal about societal pressures, unmet needs, and collective anxieties.
ORION
01.09.2025
ORION – Orientation in Conspiration
Link to Research Project: https://idea-lab.uni-graz.at/de/forschungsprojekte/orientation-in-conspiration-orion/
While conspiracy theories were once considered mostly harmless curiosities, the COVID-19 pandemic highlighted the concrete social harm they can cause, particularly when amplified by online communities. In the ORION project, funded by the Austrian Science Fund (FWF) and netidee, we investigate the factors that contribute to the popularity and spread of conspiracy theories on social media.
Conspiracy theories are not inherently negative: they can act as social “watchdogs,” pushing elites to justify their actions. However, combined with the reach of social media, especially messaging platforms like Telegram, and coordinated efforts by external actors, their societal impact has shifted, often causing more harm than good. Rather than stigmatizing followers or relying on broad censorship, ORION seeks to understand the individual, social, and technological factors that drive the spread of these narratives, laying the groundwork for interventions at the level of platform design and algorithmic systems.
Our research focuses on German-language conspiracy theories on public Telegram channels, using the 2022 “Schwurbelarchiv,” which contains over 23 terabytes of mostly multimedia content from 2019–2022. A major part of the project is cleaning, structuring, and transcribing these data to enable automated text analysis and make the dataset usable for other researchers. By analyzing text, voice messages, and videos, we aim to uncover new insights into how digital platforms shape the formation and dissemination of conspiracy theories.
Publications in this Project:
- Mathias Angermaier, Elisabeth Höldrich, Jana Lasser, Joao Pinheiro Neto (2025)
The Schwurbelarchiv: a German Language Telegram dataset for the Study of Conspiracy Theories
ArXiv abs/2504.06318
- Elisabeth Höldrich, Mathias Angermaier, Jana Lasser, Joao Pinheiro-Neto (2025)
Characterizing the Dynamics of Conspiracy Related German Telegram Conversations during COVID-19
ArXiv 2507.13398
Mapping Credibility
07.12.2024
Mapping Trust: Credibility Through User Sharing Networks on Telegram
Detecting misinformation and identifying trustworthy sources online is crucial for public trust, democratic stability, and informed decision-making. Most existing credibility databases focus on high-traffic, English-language websites, leaving smaller, non-English sites underrepresented. To address this, we developed a user-based approach for German-language Telegram channels using the Schwurbelarchiv dataset of messages from conspiracy-related chats.
We identify “superspreaders”, users who share many links, and calculate credibility scores based on the reliability of the websites they share. This allows us to classify both users and previously unlisted websites, significantly expanding coverage of German-language sites.
The approach provides a scalable way to detect credible and non-credible sources beyond mainstream, English-language databases.
Speaking Truth or Belief: How Political Language Shapes Elections
24.11.2023
In recent years, politicians have shifted the way they communicate with the public, increasingly relying on social media to reach voters directly. This has changed the language of politics: rather than emphasizing factual accuracy, politicians often craft messages that convey their emotions and personal beliefs, presenting themselves as authentic and relatable. This shift blurs the line between truth-seeking speech and belief-driven communication, raising important questions about how language influences public perception and electoral outcomes. In my thesis, under Prof. Douglas Maraun and Prof. Jana Lasser, I analyzed how these communication styles correlate with electoral trends across House and Senate races and regions before the US midterm elections of 2022.