1
Preface
2
Introduction
2.1
A lab experiment
2.2
Challenges
2.3
DAG (Directed Acyclic Graph)
2.4
A simulated DGP
2.5
Discussion
3
Why Regression?
3.1
The best-fit line
3.2
Linear Regression Specification
3.3
Law of iterated expectation
3.4
Error term
3.5
Decomposition
3.6
Estimation
3.7
Running a regression
3.8
Standard Errors
3.9
An exercise
4
Regression and Gradient Descent
5
Causal Inference
5.1
Potential Outcome Framework: Neyman-Rubin Causal Model
5.2
Average treatment effect (ATE)
5.3
RCT
5.4
Average treatment effect on the treated (ATT)
5.5
An estimation example
5.6
Unconfoundedness assumption
5.7
Discussion
5.8
Reference
6
IPW and AIPW
6.1
A simple example
6.2
Aggregated Estimator
6.3
Propensity score
6.4
Estimation of propensity score
6.5
Using cross-fitting to predict propensity score
6.6
Propensity score stratification
6.7
Inverse Probability Weighting (IPW)
6.8
Comparing IPW with Aggregated Estimate
6.9
AIPW and Estimation
6.10
Assessing Balance
6.11
Cross-fitting
7
Difference in Differences
7.1
A Quick Introduction
7.2
Set up
7.3
An example: Evaluating the impact of Medicaid expansion on uninsured rate
7.4
Naive estimator
7.5
Canonical Difference in Differences Framework
7.6
DiD in multi-period set up
7.7
Conditional Parallel Trend Assumption
7.8
Some concerns with controls
7.9
The
\(2 \times 2\)
Difference-in-Differences Estimate
7.10
Event study model
7.11
Two way fixed effect (TWFE) Revisited
7.12
Various ways of estimation
7.13
Multi Period, Multi Group and Variation in Treatment Timing
7.14
Problem with TWFE in Multiple Group with Treatment Timing Variation
7.15
What is TWFE Estimating when there is Treatment Timing Variation?
7.16
Assumptions governing TWFEDD estimate
7.17
How Does Treatment Effect Heterogeneity in Time Affect TWFE?
8
Causal Forest
8.1
Introduction
8.2
Summary of GRF
8.3
Motivation for Causal Forests
8.4
Causal Forest
8.5
An example of causal forest
9
Heterogeneous Treatment Effects
9.1
Some ways to estimate CATE
9.2
Estimation
9.3
Some Remarks and Questions
Causal Inference
8
Causal Forest