主题：Inference on Maximal Heterogeneous Treatment Effect in Observational Studies
主讲人：郭心舟 香港科技大学
主持人：王国长 暨南大学
时间：2023年4月25日（周二）10：00-11：30
地点：暨南大学石牌校区经济学院大楼（中惠楼）102室
摘要
The broader availability of high-dimensional observational data provides opportunities for scientist to better understand treatment effect heterogeneity; in particular, maximal heterogeneous treatment effect, when randomized clinical trials are unavailable due to cost and ethical constraints. In modern drug development, statistical inference on maximal heterogeneous treatment effect plays an essential role in identifying and confirming the most promising subgroup and/or treatment strategy. Even if the heterogeneous treatment effects can be estimated accurately in observational studies, researchers might iteratively search for the maximal heterogeneous treatment effect and then only report the results with the most positive effects (without proper adjustments). Such practices naturally induce the bias and pave the way for spurious conclusions in subgroup analysis and multiple treatment comparisons. To prevent false scientific discoveries, we propose a debiased inference framework to evaluate the maximal heterogeneous treatment effect after a data-dependent search. Our proposed inference procedure is easy to compute, asymptotically sharp and generalizable to a variety of complex scenarios such as the case when the candidate subgroups are overlapped. We demonstrate the merit of the proposed method through extensive simulation studies and by analyzing the UK Biobank data.
主讲人简介
Xinzhou Guo is an Assistant Professor in the Department of Mathematics at the Hong Kong University of Science and Technology. He received his B.S. in Applied Mathematics from Peking University and Ph.D. in Statistics from the University of Michigan. Prior to joining HKUST in 2021, he did a postdoc at Harvard University. His main research interests are subgroup analysis, resampling methods, precision medicine and regulatory decision-making.
编辑|麦嘉杰
初审|王国长
审核|郑贤
（来源：暨南大学经济学院微信公众号）
