Potential outcomes framework — Rubin causal model.
hardAnswer
- For each unit i and treatment T, define (outcome if untreated) and (outcome if treated).
- Individual causal effect: — fundamentally unobservable (fundamental problem of causal inference).
- Estimands: ATE = E[Y(1) - Y(0)], .
- Identification requires: (1) SUTVA (no interference / one version of treatment), (2) unconfoundedness , (3) overlap.
- Foundation of modern causal inference.
Check yourself — multiple choice
- Random
- , potential outcomes; ATE = E[Y(1) - Y(0)]; identification needs SUTVA + unconfoundedness + overlap
- Same as OLS
- Not real
Rubin: , potential outcomes; ATE / ATT under SUTVA + ignorability + overlap.
#causal-inference#theory
Practise Statistics Fundamentals
215 interview questions in this topic.