Compute sample size for two sample Wilcoxon (Mann-Whitney) test

Share:

Description

Compute sample size to test the hypothesis that two samples come from the same population against that Y's tend to be larger than X's.

Usage

1
2
sampleSizeWilcoxTwoSample(a = 0.05, b = 0.2, c = 0.5, 
    pxy = 0.75, two.sided = TRUE)

Arguments

a

Significance level of test.

b

Desired power of test.

c

Proportion of observations in group 1: c = m / (m + n) (c = 0.5 means equally sized groups).

pxy

A value for the probability P(Y > X).

two.sided

If TRUE a two-sided test is assumed, otherwise one-sided.

Details

Given two independent samples X_1, ..., X_m and Y_1, ..., Y_n, we want to test the hypothesis that the two samples come from the same population against that Y's tend to be larger than X's.

Value

m

Sample size of the first group.

n

Sample size of the second group.

Author(s)

Kaspar Rufibach
kaspar.rufibach@gmail.com

References

Nother, G.E. (1987). Sample Size Determination for Some Common Nonparametric Tests JASA, 82, 644–647.

Examples

1
2
3
4
5
6
7
8
# compute sample size for some pxy's
pxys <- c(0.65, 0.7, 0.75)
dat1 <- matrix(ncol = 3, nrow = length(pxys))
colnames(dat1) <- c("P(Y > X)", "m", "n")
for (j in 1:length(pxys)){dat1[j, ] <- c(pxys[j], 
    sampleSizeWilcoxTwoSample(a = 0.05, b = 0.1, c = 0.5, 
    pxy = pxys[j], two.sided = TRUE))}
dat1

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.