Perm.CI: Permutation test confidence interval for a treatment effect on a
binary outcome
Description
Computes permutation-based confidence intervals for the
average treatment effect on a binary outcome in an experiment where
\(m\) of \(n\) individuals are randomized to treatment by design.
Usage
Perm.CI(data, level, nperm)
Value
tau.hat
estimated average treatment effect
lower
lower bound of confidence interval
upper
upper bound of confidence interval
Arguments
data
observed 2 by 2 table in matrix form where row 1 is
the treatment assignment Z=1 and column 1 is the binary outcome Y=1
level
significance level of hypothesis tests, i.e., method yields a 100(1-level)% confidence interval
nperm
number of randomizations to perform for each
hypothesis test
The permutation confidence interval results from inverting
\(O(n^{4})\) hypothesis tests where \(n\) is the total number of
observations in the observed 2 by 2 table. For each hypothesis test,
if \(n \choose m\) is less than or equal to nperm, \(n \choose m\)
randomizations are performed, but if \(n
\choose m\) is greater than nperm, a random sample with replacement of nperm randomizations
are performed.
References
Rigdon, J.R. and Hudgens, M.G. (2015).
Randomization inference for treatment effects on a binary outcome.
Statistics in Medicine, 34(6), 924-935.