Research

My research agenda examines the intersection of digital technology and democratic discourse. Using computational methods and experimental designs, I investigate how algorithms and platform governance shape what citizens see, believe, and tolerate online.

I. Algorithmic Gatekeeping & Inequality

Do search and recommendation algorithms amplify structural discrimination?

  • Influence of Hate Speech About Refugees in Search Algorithms on Political Attitudes (2024) New Media & Society | Journal | PDF > This article provides causal evidence that exposure to hate speech in search engines increases hostility toward refugees—not among moderates, but specifically among individuals with an extreme right-wing ideology. Crucially, those with right-wing ideologies tend to trust hate speech about refugees more than those with a left-wing ideology. According to the study, these individuals are also more likely to self-select into such hateful content, which may lead to a reinforcement of hostile attitudes in the long run.

  • Biased Representation of Politicians in Google and Wikipedia Search (2021) Political Communication | Journal > Using a novel computational measure for detecting gender bias, this study demonstrates that search algorithms do not treat all candidates equally—revealing significant gender biases particularly among conservative politicians.

II. Toxicity & Content Moderation

Do citizens actually want a “sanitized” internet?

  • Toxic Speech and Limited Demand for Content Moderation on Social Media (2024) American Political Science Review | Journal | PDF > The study challenges the assumption that users universally desire stricter content moderation. Results from five experiments show that citizens largely tolerate toxic speech and only demand moderation for severe forms of toxicity (e.g., violent threats). However, these demands are context-dependent: users significantly differentiate their tolerance levels based on the specific target that is attacked. Furthermore, the study conceptualizes “uncivil,” “intolerant,” and “threatening” language, demonstrating that while users perceive these as distinct types of toxicity, they react similarly to uncivil and intolerant content—reserving demands for strict removal only for physical threats.

III. Disinformation & Conspiracy Beliefs

What drives susceptibility to conspiracy narratives?

  • Justifying an Invasion: When Is Disinformation Successful? (2024) Political Communication | Journal | PDF > Analyzing data from 19 countries regarding the Ukraine invasion, this study finds that a user’s pre-existing “conspiracy mindset” is a stronger predictor of believing hostile disinformation than their actual media diet.

  • Beliefs in Conspiracy Theories and Online News Consumption (2024) Journal of Quantitative Description: Digital Media | Journal | PDF > A comparative mapping of conspiracy beliefs across 17 democracies. Results show that national information environments play a crucial moderating role in the relationship between social media use and conspiracy beliefs.

Grants & Awards

My research has been supported by competitive funding and recognized by international awards.

Secured Funding

  • TUM Think Tank / HfP Seed Funding (2026) Principal Investigator | “AI-driven Empowerment in Political Discussions” (Pilot for ERC Grant application).
  • Re-Boot Social Media Lab (2022-2023) Co-Principal Investigator | “Platforms for the People” (€35k).

Research Honors

  • Finalist: WWTF Young Research Group (2024) Selected for the final round of the Vienna Science and Technology Fund call (approx. €1.6M volume), confirming the high quality of the research proposal.
  • Best Paper Award (2019) ACM Web Science Conference