Causal Inference I
Matching and Inverse Probability of Treatment Weighting
By Li Ge in Causal Inference
July 3, 2020
Objectives
In the first part of the series, I will cover the basics of causal inference by defining causal estimands using potential outcomes framework, and introduce the methods of matching and inverse probability of treatment weighting to address measured confounders.
Table of Contents
- Introduction to Causal Effects
- What makes a relationship causal?
- Potential Outcomes and Counterfactuals
- Causal Effects
- Caveats of Causal Effects
- Causal Assumptions
- SUTVA
- Consistency
- Ignorability
- Positivity
- Causal Estimands
- Confounding and Directed Acyclic Graphs (DAGs)
- Confounding
- DAG
- Matching and Propensity Scores
- Observational Studies
- Matching
- Matching Procedures
- Propensity Score
- Balancing Score
- Estimated Propensity Score
- Propensity Score Matching
- Inverse Probability of Treatment Weighting (IPTW)
- Intuition for IPTW
- IPTW Estimator
- Marginal Structural Models
- IPTW Estimation
- Regression-Based Estimation
- Doubly Robust Estimators
- IPTW in Practice