CSS (Computational Social Science) is located at the intersection of social science and computer science. Social phenomena are the main focus of our research. They are, however, analyzed using a variety of data types, sometimes diverging from established data sources like panel surveys, sometimes complementing those. In particular, this data includes process-generated data, such as scientific texts, patents, minutes of plenary proceedings, or messages on social media channels. The variety of data affords a variety of methods as analytical stategies.
We apply text-as-data approaches (e.g., Erhard et al. 2025), statistical models of social networks (e.g., Windzio & Heiberger 2024), or agent-based modeling (e.g., Kaffai & Heiberger 2021) to explain social phenomena like populism, bullying or disease spread. A specific focus of our group's work lies on the combination of computer linguistic methods with survey data (e.g., Erhard et al. 2021 or Heiberger et al. 2021).
Recent advances of Large Language Models (LLM) provide yet another field of research for Sowi VII, as we explore the "social science fit" of LLMs, i.e, we are commited to ensuring the validity and reliability of data, and the theoretical applicability of the methods.
Sowi VII is responsible for teaching statistics at the Bachelor-level and computational methods at the Master-level. We aim to offer a modern approach to learning statistical modeling from basic understandings of data to first experiences of state-of-the-art methods. In all our classes, we emphasize practical, hands-on experiences from a social science perspective.
