Computes permutation test p-value

Description

Returns permutation test p-value; used in the construction of exact confidence intervals by the function exactCI

Usage

1
pval(eff, est, null, y0.a0, y1.a0, y0.a1, y1.a1, h, n, m.a0, m.a1, C2)

Arguments

eff

treatment effect of interest; either “DEa0”, “DEa1”, “IE”, “TE”, or “OE”

est

estimated treatment effect using estimators from Hudgens and Halloran (2008)

null

value of treatment effect of interest under the sharp null hypothesis

y0.a0

hypothesized vector \vec{y}(0;α_{0}) under the sharp null hypothesis

y1.a0

hypothesized vector \vec{y}(1;α_{0}) under the sharp null hypothesis

y0.a1

hypothesized vector \vec{y}(0;α_{1}) under the sharp null hypothesis

y1.a1

hypothesized vector \vec{y}(1;α_{1}) under the sharp null hypothesis

h

the number of groups out of k total to be randomized to strategy α_{1}

n

group size vector where element i=1,…,k is equal to the number of subjects in group i

m.a0

α_{0} randomization vector where element i=1,…,k is equal to the number of subjects in group i who would receive treatment if group i was randomized to strategy α_{0}

m.a1

α_{1} randomization vector where element i=1,…,k is equal to the number of subjects in group i who would receive treatment if group i was randomized to strategy α_{1}

C2

number of re-randomizations (experiments) to conduct in computing the null distribution of the estimator

Details

See equation (6) in Rigdon and Hudgens (2014)

Author(s)

Joseph Rigdon jrigdon@bios.unc.edu

References

Hudgens, M.G. and Halloran, M.E. “Toward causal inference with interference.” Journal of the American Statistical Association 2008 103:832-842.

Rigdon, J. and Hudgens, M.G. “Exact confidence intervals in the presence of interference.” Submitted to Statistics and Probability Letters 2014.

See Also

exactCI