Regression discontinuity design — how does it work?
hardAnswer
- Treatment assigned by a threshold on a continuous running variable X (e.g. score ≥ 60 → scholarship).
- Compare Y just above vs just below cutoff — near cutoff, individuals are 'as-if random'.
- Sharp RDD: treatment deterministic at cutoff.
- Fuzzy RDD: probability jumps but not to 1 — combine with IV.
- Local linear regression around cutoff + optimal bandwidth (Imbens-Kalyanaraman).
- Foundational in policy evaluation, admission thresholds, credit-score cutoffs.
Check yourself — multiple choice
- Same as OLS
- Treatment assigned by threshold on continuous X; compare just above vs below cutoff (as-if random); sharp RDD or fuzzy RDD + IV; local linear regression with optimal bandwidth
- Random
- Not real
RDD: exploits threshold in running variable; local linear regression + optimal bandwidth.
#causal-inference#quasi-experimental
Practise Statistics Fundamentals
215 interview questions in this topic.