Causal Inference II
Causal Graphs (DAGs) and Instrumental Variables Methods
By Li Ge in Causal Inference
July 10, 2020
Objectives
In the second part of this series, I will emphasize the graphical understanding of the causal models, and introduce the instrumental variables methods to estimate local (or complier) causal effects, which can be used to address unmeasured confounders.
Table of Contents
- Introduction to Causal Effects (Review)
- Confounding and Directed Acyclic Graphs (DAGs)
- Confounding
- The Basics of Graphical Models
- DAG examples
- Paths & Associations
- D-Separation
- Backdoor Path Criterion
- Disjunctive Cause Criterion
- Matching and Propensity Scores (Review)
- Inverse Probability of Treatment Weighting (Review)
- Instrumental Variables Methods
- Instrumental Variables
- Randomized Trials with Noncompliance
- Local Average Treatment Effect
- Observed Data
- Exclusion Restriction
- Monotonicity
- Complier Average Causal Effect (CACE)
- IVs in Observational Studies
- Two Stage Least Squares