Event Date
Speaker: Dominik Rothenhäusler (Assistant Professor of Statistics, Stanford University)
Title: "Quantifying distributional uncertainty"
Abstract: How can we draw trustworthy scientific conclusions? It has been argued that trustworthy scientific conclusions require disparate sources of evidence. For example, in causal inference from observational data, it is common to compute regression-adjusted estimators for different choices of adjustment sets. Then stability investigations can be done by studying the estimator-to-estimator variability between sensible choices of adjustment sets. However, different methods might have shared biases, making it difficult to judge the theoretical guarantees of this practice. We formalize this issue by introducing a ``distributional uncertainty model", which captures biases in the data collection process. Distributional uncertainty is related to other concepts in statistics, ranging from correlated data to selection bias and confounding. We show that a stability analysis on a single data set allows to construct confidence intervals that account for both sampling uncertainty and distributional uncertainty.
Faculty webpage (links to Stanford University): https://statistics.stanford.edu/people/dominik-rothenhausler
Seminar Date/Time: Thursday April 13, 2023, at 4:10pm
Location: MSB 1147 (Colloquium Room)
Refreshments: 3:30pm (MSB 1147 courtyard)