power.w.test: Power and sample size of the Wilcoxon Mann Whitney U test

In floatofmath/adaperm: Permutation tests for adaptive designs

Description

This function computes the power and sample size according to the asymptotic power formula given in [(2.23) p.71 of Lehmann (2006)]. If is.null(n) the sample size required to achieve the target power is computed using bisection search.

Usage

power.w.test(n = NULL, p1 = NULL, p2 = NULL, p3 = NULL, propn = 1/2,
  sig.level = 0.025, power = 0.9, silent = F)

Arguments

n	Sample size in the control group
p1	probability that X < Y
p2	probability that X < Y and X < Y'
p3	probability that X < Y and X' < Y
propn	proportion of the total number of observations in the control group
sig.level	significance level of the test
power	desired target power
silent	should hints be suppressed

Value

Either the required number of control group observations to achieve the targe power, or the power of the WMW test under the specified alternative and given sample size.

Author(s)

float

floatofmath/adaperm documentation built on May 16, 2019, 1:18 p.m.