Causal Inference in R: Introduction
> who_are_we(c("lucy", "malcolm", "travis"))
The three practices of analysis
- Describe
- Predict
- Explain
Normal regression estimates associations. But we want counterfactual, causal estimates:
What would happen if everyone in the study were exposed to x vs if no one was exposed.
For causal inference, we need to make sometimes unverifiable assumptions.
Today, we’ll focus on the assumption of no confounding.
Tools for causal inference
- Causal diagrams
- Propensity score weighting
- Propensity score matching
- G-methods & friends
Other tools for causal inference
- Randomized trials
- Instrumental variables & friends
- TMLE and other ML for causal inference
Resources
Causal Inference in R: Our book! Free online.
Causal Inference: Comprehensive text on causal inference. Free online.
The Book of Why: Detailed, friendly intro to DAGs and causal inference.
Mastering ’Metrics: Friendly introduction to IV-based methods
Causal Inference in R: Introduction