QUANTITATIVE RESEARCH METHODS WORKSHOP
Abstract: Modern empirical work often involves estimating effects for many individuals or groups. To increase accuracy of the estimates, it is common to employ shrinkage or regularization, such as shrinking noisy preliminary estimates toward baseline values using empirical Bayes methods, or using machine learning techniques. The gains in accuracy come at the expense of introducing bias into the estimates, which makes inference challenging. We develop a general method for constructing intervals in this setting: to automatically reflect gains from data-driven regularization, they are based on regularized or shrinkage estimators, but use a novel critical value to take into account the potential bias of the estimators. These intervals have an average coverage guarantee, covering a prespecified fraction (95%, say) of the true effects on average. We illustrate our methods with an application to the effects of U.S. neighborhoods on intergenerational mobility.
Michal Kolesár is a Professor of Economics at Princeton University, specializing in econometrics. He received his Ph.D. in economics from Harvard in 2013. His research focuses on developing methods for causal inference, such as developing robust methods for inference in regression discontinuity designs, and instrumental variables models.
This virtual workshop is open to the Yale community. To receive Zoom information, you must subscribe to the Quantitative Research Methods Workshop at this link: https://csap.yale.edu/quantitative-research-methods-workshop.
The series is sponsored by the ISPS Center for the Study of American Politics and The Whitney and Betty MacMillan Center for International and Area Studies at Yale with support from the Edward J. and Dorothy Clarke Kempf Fund.