Colloquium Series: Dominik Rothenhaeusler

Date and Time

February 9, 2026
12:00PM - 01:00PM EST

Location

Maxwell-Dworkin 134A/B

Our upcoming event for the Statistics Colloquium Series is scheduled for Monday, February 9 from 12:00 – 1:00pm (ET) and will be an in-person presentation at Maxwell-Dworkin 134A/B. Lunch will be provided to guests following the talk. This week's speaker will be Dominik Rothenhaeusler of Stanford's Department of Statistics.

Random Distribution Shift in Effect Generalization: Implications for Reweighting, Tuning, and Data Collection

Generalizing scientific findings across populations is central to science and policy. A common approach is to reweight observed covariates, implicitly assuming that only the covariate distribution changes across settings. Analyzing 680 studies across 65 sites, we show that shifts in the outcome conditional on covariates (Y | X) are common, but their magnitude can be predicted from changes in observed covariates. These patterns motivate a random distribution shift framework for effect generalization across studies and sites.

Building on this model, I’ll discuss implications for three problems:

(1) reweighting, where unstable weights can amplify noise and hybrid weighting/pooling can be preferable; (2) tuning nonparametric regression, where distributional uncertainty reduces effective sample size and alters optimal smoothing; and (3) prioritizing data collection, where covariate-only summaries can help identify which new sources will most improve predictive performance.

This is joint work with Ying Jin, Naoki Egami, Yujin Jeong, Anna Lyubarskaja, and Ivy Zhang.

Dominik Rothenhaeusler is an Assistant Professor of Statistics at Stanford University. His research focuses on distribution shift and causal inference, with broader interests in heterogeneous data, high-dimensional statistics, and graphical models. He earned his Ph.D. from ETH Zürich in 2018 under the supervision of Nicolai Meinshausen and Peter Bühlmann and completed a postdoctoral fellowship with Bin Yu at UC Berkeley. He is a recipient of the Royal Statistical Society’s David Cox Research Prize, the Dieter Schwarz Early-Career Faculty Award, and honors including Chamber Fellowship and the David Huntington Dean’s Scholar distinction.