Skip to content Skip to navigation

Statistics Colloquium

Event Type: 
Dr. Richard Hahn, Booth School of Business, University of Chica
Event Date: 
Thursday, December 15, 2016 -
3:30pm to 5:00pm
SMLC 356
General Public
Statistics group

Event Description: 

Title: Bayesian Causal Forests


Abstract: In this talk I will describe a semi-parametric Bayesian regression model for estimating heterogeneous treatment effects from observational data. Standard nonlinear regression models, which may work quite well for prediction, can yield badly biased estimates of treatment effects when fitted to data with strong confounding. The new Bayesian causal forest model is able to eliminate this adverse bias by jointly modeling the treatment and the response conditional on control variables. Two empirical illustrations are given, analyzing the impact of smoking on medical expenditures and the impact of abortion laws on future crime rates.

Bio:  P. Richard Hahn is Associate Professor of Econometrics and Statistics.  His research develops computational methods for modeling complex real-world data.  Currently he is developing statistical tools for analyzing data from personal health technology to inform training and recovery programs for professional athletes.  His research has appeared in the Journal of theAmerican Statistical Association, the Annals of Applied Statistics, the Journal of Business and Economic Statistics, and the Journal of the Royal Statistical Society.  Outside of academia, Hahn has statistical consulting experience in diverse areas, including politics, management, marketing and biotech.  Hahn earned his PhD in statistical science from Duke University.

Event Contact

Contact Name: Gabriel Huerta

Contact Email: