LLMs for Behavioral Science

Symposium

The one-day symposium will take place Tuesday, February 11th, 2025 at the Faculty of Psychology, University of Basel (Missionsstrasse 64a, 4055 Basel, Seminarraum 00.010), and will be dedicated to exploring the potential of large language models (LLMs) in behavioral and social sciences. This event is designed for those who are beginning to work with LLMs or are already using them and want to deepen their understanding, gain new insights, and discuss innovative applications. Through a series of talks and discussions, participants will engage with cutting-edge research, learn about practical implementations, and explore how LLMs can address key challenges in fields such as psychology, economics, and political science. The symposium aims to foster a collaborative environment where researchers can exchange ideas about what LLMs can help achieve in scientific inquiry.

Schedule

Time Session Details
10:50 Welcome & Introduction
11:00 Language models accurately predict risk perception
MSc Zakir Hussain, Cognitive and Decision Sciences, University of Basel
11:30 Turning a radio show into research data: A field report on challenges in speech-to-text automation
Dr. Max Theisen, Cognitive and Decision Sciences, University of Basel
12:00 LLM-rated deservingness predicts crowdfunding success
Prof. Dr. Dominik Meier, Global Philanthropy, University of Basel
12:30 Lunch Break
14:00 Simulating social media using Large Language Models
Prof. Dr. Petter Törnberg, University of Amsterdam
14:30 Examining participant diversity in psychological research using LLMs
Dr. Anna Thoma, Center for Adaptive Rationality, Max Planck Institute for Human Development
15:00 Shaping parliamentary debate: Measuring the centrality of party leaders in populist discourse with LLMs
MSc Clint Claessen, Political Science, University of Basel
15:30 Break
16:00 Are minorities punished more harshly for underperformance? Evidence from Premier League soccer (via Zoom)
Prof. Dr. Alexandra Siegel, Political Science, University of Colorado Boulder
16:45 Computational analysis of 100K choice dilemmas” (via Zoom)
Prof. Dr. Sudeep Bhatia, Computational Behavioral Science, University of Pennsylvania
17:30 Apéro