Estimation and testing methods for causal inference with interference
Abstract/Contents
- Abstract
- Causal inference is one of the core areas of research in modern data science that allows researchers to determine whether a specific intervention or treatment has an effect on an outcome. In its most basic form, causal inference is concerned with understanding the "cause and effect" relationship between variables. This requires going beyond correlation to understand whether changing one variable leads to a change in another. The gold standard for inferring causality is the randomized controlled trial, which randomly assigns subjects to a treatment or control group and compares outcomes. While randomized controlled trials provide us with data to do causal inference, the subsequent statistical analysis often relies on a key assumption known as the Stable Unit Treatment Value Assumption (SUTVA). This assumption states that the treatment of one unit (or individual) does not affect the outcome of another unit. However, in many real-world situations, this assumption does not hold, leading to what is called interference or a violation of SUTVA. Interference can occur in various contexts such as social networks, where the treatment of one person can influence the outcomes of others, or in marketplace, where treatment of one entity can impact other entities of same type. Understanding and handling interference is a critical and complex aspect of causal inference, and it necessitates more advanced methods to correctly estimate causal effects. This dissertation offers new methodologies and theoretical results to address key issues in causal inference with interference. Specifically, we develop inferential results for causal effect estimators in panel experiments under interference, introduce novel estimation methods for causal effects with network experiments and tackle the problem of detecting interference in online controlled experiments with increasing allocation.
Description
| Type of resource | text |
|---|---|
| Form | electronic resource; remote; computer; online resource |
| Extent | 1 online resource. |
| Place | California |
| Place | [Stanford, California] |
| Publisher | [Stanford University] |
| Copyright date | 2023; ©2023 |
| Publication date | 2023; 2023 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Author | Han, Wu |
|---|---|
| Degree supervisor | Imbens, Guido |
| Degree supervisor | Ugander, Johan |
| Thesis advisor | Imbens, Guido |
| Thesis advisor | Ugander, Johan |
| Thesis advisor | Owen, Art B |
| Degree committee member | Owen, Art B |
| Associated with | Stanford University, School of Humanities and Sciences |
| Associated with | Stanford University, Department of Statistics |
Subjects
| Genre | Theses |
|---|---|
| Genre | Text |
Bibliographic information
| Statement of responsibility | Kevin Han. |
|---|---|
| Note | Submitted to the Department of Statistics. |
| Thesis | Thesis Ph.D. Stanford University 2023. |
| Location | https://purl.stanford.edu/yh822by5730 |
Access conditions
- Copyright
- © 2023 by Wu Han
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...