This item is only available for download by members of the University of Illinois community. Students, faculty, and staff at the U of I may log in with your NetID and password to view the item. If you are trying to access an Illinois-restricted dissertation or thesis, you can request a copy through your library's Inter-Library Loan office or purchase a copy directly from ProQuest.
Permalink
https://hdl.handle.net/2142/125707
Description
Title
Selected topics on design-based causal inference
Author(s)
Chen, Zhe
Issue Date
2024-07-11
Director of Research (if dissertation) or Advisor (if thesis)
Li, Xinran
Zhu, Ruoqing
Doctoral Committee Chair(s)
Li, Xinran
Committee Member(s)
Yu, Ruoqi
Simpson, Douglas G
Department of Study
Statistics
Discipline
Statistics
Degree Granting Institution
University of Illinois at Urbana-Champaign
Degree Name
Ph.D.
Degree Level
Dissertation
Keyword(s)
Randomization inference
Quantile of individual treatment effect
Instrumental variable
Observational study
Abstract
Design-based causal inference has emerged as a robust approach for impact evaluation of interventions, programs, and policies. These methods draw causal conclusions by relying solely on the building blocks of experimental designs, thereby requiring minimal assumptions and exhibiting good statistical properties. Design-based methods are applicable to both randomized controlled trials and quasi-experimental designs.
This thesis comprises three papers discussing design-based methodologies for causal inference across different study designs. The first paper proposes two enhanced randomization-based, finite-sample valid methods for inferring distributions and quantiles of individual treatment effects. The two methods demonstrate substantial power gain compared to the existing approaches in both completely randomized and stratified randomized experiments, and they can be further extended to sampling-based experiments as well as quasiexperiments constructed from matching. The second paper systematically reviews the role of randomizationbased inference in unraveling individual treatment effects in early phase vaccine trials and extends some results from the first paper to a special scenario where na¨ıve participants are not expected to exhibit responses to highly specific endpoints. We apply these methods to analyzing the immunogenicity data derived from HIV Vaccine Trials Network Study 086 and explored how different methods may facilitate decision-making and improve the evaluation of vaccine regimens. The third paper studies how to manipulate a continuous instrumental variable (IV) to facilitate causal inference in a design-based framework. We develop a non-bipartite, template matching algorithm that embeds observational data into a target, pair-randomized encouragement trial which maintains fidelity to the original study cohort while strengthening the IV. We then derive both randomization-based and biased-randomization-based inference of partial identification bounds for the sample average treatment effect in an IV-based matched pair design.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.
