Gaussian prepivoting for finite population causal inference

• Published in: Journal of the Royal Statistical Society. Series B, Statistical methodology Vol. 84; no. 2; pp. 295 - 320
• Main Authors: Cohen, Peter L., Fogarty, Colin B.
• Format: Journal Article
• Language: English
• Published: Oxford Oxford University Press 01.04.2022
• Subjects: Averages; covariance; cumulative distribution; Distribution functions; Hypotheses; Inference; Null hypothesis; pivotal quantity; Randomization; randomization tests; Regression analysis; rerandomization; sampling; sharp null; Statistical analysis; Statistical methods; Statistical tests; Statistics; stochastic dominance; testing; weak null

In finite population causal inference exact randomization tests can be constructed for sharp null hypotheses, hypotheses which impute the missing potential outcomes. Oftentimes inference is instead desired for the weak null that the sample average of the treatment effects takes on a particular value...