Please join us for our upcoming Statistics Seminar on May 19th with Art B. Owen who is a Max H. Stein Professor of Statistics at Stanford University.
Title: Tie-Breaker Designs
Abstract: Companies may offer incentives to their best customers and philanthropists may offer scholarships to the strongest students. They can evaluate the impact of these treatments later using a regression discontinuity analysis. Unfortunately, regression discontinuity analyses have high variance. It is possible to get much more statistical efficiency using a tie-breaker design that works by triage: top subjects get the treatment, bottom subjects do not, and those in between have their treatment randomized. Statistical efficiency increases monotonically with the amount of randomization, causing an exploration versus exploitation tradeoff. This holds in a simple two-line regression model and also with nonparametric kernel regression-based methods. We have found D-optimal treatment probability functions for scalar and vector data. The conclusion is that when it is possible (and ethical) to randomize for a group of subjects, it is wise to do so. Based on joint work with Hal Varian, Google and Dan Kluger, Harrison Li, Tim Morrison, Stanford University. Some portions of this work were done as a paid consultant for Google and were not part of Art Owen's Stanford responsibilities. Subsequent work was done as a Stanford faculty member.