Developed by Judea Pearl (Causality, 2000) from the path analysis of Sewall Wright (Correlation and Causation, 1921).
Maps assumed causes as a directed graph of variables and arrows, showing which factors to control for, which to leave alone, and whether an effect can be estimated at all.
When to reach for it
When a causal claim has to rest on observational data with no experiment possible, and its hidden assumptions about confounders need to be laid bare.
How it runs
- List the treatment, the outcome, and candidate covariates.
- Draw an arrow wherever one variable directly causes another.
- Confirm no variable causes itself through a chain.
- Trace the back-door paths from treatment to outcome.
- Choose controls that block those paths, avoiding common effects.
- Add likely unmeasured confounders and check again.
Where it sits in SOLVE eX
Go to the source
- Causality: Models, Reasoning, and Inference, Cambridge University Press; second edition 2009, 2000
- Correlation and Causation, Journal of Agricultural Research 20, 1921