Why is a doubly-robust estimator useful?
hardAnswer
- Combines outcome model μ̂(x) and propensity model ê(x).
- Formula: ATE = E[μ̂(1, X) - μ̂(0, X) + T(Y - μ̂(1, X))/ê(x) - (1-T)(Y - μ̂(0, X))/(1-ê(x))].
- Consistent if EITHER μ̂ OR ê(x) is correctly specified — hence 'doubly robust'.
- AIPW, TMLE are standard implementations.
- Modern default in observational causal inference; also basis of DR-learners / DML with ML nuisance models.
Check yourself — multiple choice
- Random
- Combines outcome + propensity models; consistent if EITHER is correctly specified; foundation of AIPW / TMLE / DML with ML nuisance functions
- Same as IPW
- Not real
Doubly robust (AIPW / TMLE): consistent if outcome OR propensity model right.
#causal-inference
Practise Statistics Fundamentals
215 interview questions in this topic.