STAT 41530 Topics in Causal Inference
Description
In this course, we will have a brief introduction of both the potential outcome framework and the causal directed acyclic graph (DAG) for causal inference. We will discuss topics including confounding, instrumental variables (IV) and mediation analysis, with the applications of causal inference in genetics and epidemiological research.
We will follow the following book for the first 5 weeks:
- Causal Inference, What If by Hernán MA, Robins JM
Course Materials
| Week | Date | Topic | Slides | Extra reading materials |
|---|---|---|---|---|
| 1 | 2022-01-11 | Potential outcome framework: definition and randomized experiments | Lecture 1 | – |
| 1 | 2022-01-13 | Potential outcome framework: observational studies | Lecture 2 | – |
| 2 | 2022-01-18 | DAG: Markov assumption, d-seperation and connection with potential outcome framework | Lecture 3 | Pearl’s slides |
| 2 | 2022-01-20 | DAG: do-operator, backdoor and frontdoor criteria | Lecture 4 | Pearl’s slides |
| 3 | 2022-01-25 | DAG: selection bias. Review | Lecture 5 | – |
| 3 | 2022-01-27 | Estimation and statistical inference | Lecture 6 | Dr. Fan Li’s slides |
| 4 | 2022-02-01 | Instrumental Variables | Lecture 7 | – |
| 4 | 2022-02-03 | Causal mediation analysis | Lecture 8 | Tyler’s causal mediation lecture |
| 4 | 2022-02-08 | Lord paradox | Lecture 9 | Cox and McCullagh 1982, Holland and Rubin 1983, Pearl 2016 |
| 4 | 2022-02-10 | Comparisons between potential outcome and DAG | Lecture 10 | Imbens' paper, Pearl’s feedback on the paper |