Panos Toulis is an Associate Professor of Econometrics and Statistics and a John E. Jeuck Faculty Fellow at the Booth School of Business. His research interests include causal inference, experimental design, and statistical machine learning. His applied work primarily focuses on social networks, including school absenteeism, peer effects, crime spillovers, and taxation policy. He has received a faculty award from Adobe, an Economic Graph Challenge award from LinkedIn, the Tom Ten Have Award in causal inference, the Arthur P. Dempster Award from Harvard University, and the Google U.S./Canada PhD Fellowship in Statistics. His research has also been supported by NSF research grants.
This talk is sponsored by the School of Operations Research and Information Engineering, the department of Statistics and Data Science, and the Center for Data Science for Enterprise and Society.
Talk: Randomization Tests for Robust Causal Inference in Network Experiments
Abstract: Network experiments pose unique challenges for causal inference due to interference, where cause-effect relationships are confounded by network interactions among experimental units. This paper focuses on group formation experiments, where individuals are randomly assigned to groups and their responses are observed—for example, do first-year students achieve better grades when randomly paired with academically stronger roommates? We extend classical Fisher Randomization Tests (FRTs) to this setting, resulting in tests that are exact in finite samples and justified solely by the randomization itself. We also establish sufficient theoretical conditions under which general FRTs for network peer effects reduce to computationally efficient permutation tests. Our analysis identifies equivariance as a key algebraic property ensuring the validity of permutation tests under network interference.