samplesize.vanElteren: Compute required overall sample size for van Elteren test

Description Usage Arguments Details Value Examples

View source: R/samplesize.R

Description

samplesize.vanElteren uses the procedure suggested by Zhao et al. [1] to compute the required overall sample size for a van Elteren test comparing two groups A and B within strata [2].

Usage

1
2
samplesize.vanElteren(strata_fractions, treatment_fractions,
  strata_fractions_alternative, alpha, beta)

Arguments

strata_fractions

numeric vector of length n_strata with fraction of subjects enrolled to the respective strata

treatment_fractions

2-by-n_strata matrix where the first row corresponds to group A and the second to group B. Each column must sum to 1 and the k-th column holds the treatment fractions for the k-th stratum.

strata_fractions_alternative

2-by-n_categories-by-n_strata array. I.e. strata_fractions_alternative[g, , k] holds the expected category fractions in group g (= 1 or 2) for stratum k (= 1 ... n_strata); strata_fractions_alternative[g, , k] must sum to 1.

alpha

maximal type one error rate

beta

maximal type two error rate on specified point alternative

Details

[1] Zhao et al., (2008), "Sample Size Calculation for the van Elteren Test Adjusting for Ties", Journal of Biopharmaceutical Statistics, 18:1112<e2><80><93>1119.

[2] van Elteren, (1960). "On the combination of independent two sample tests of Wilcoxon", Bulletin of the Institute of International Statistics, 37:351<e2><80><93>361.

Value

N, single number giving the required overall sample size for the van Elteren test.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
# First example from [1], Table 3.
alpha       <- 0.05
beta        <- 0.2
strata_frac <- c(.78, .22)
n_strata    <- length(strata_frac)
treatment_frac   <- matrix(.5, nrow = 2, ncol = n_strata)
n_cat            <- 3
strata_frac_alt_BIFl28 <- matrix(c(
   0.62, 0.31, 0.07, # Duloxetine
   0.51, 0.39, 0.1   # Placebo
), nrow = 2, ncol = n_cat, byrow = TRUE)
strata_frac_alt_BIFgeq28 <- matrix(c(
   0.67, 0.33, 0.00, # Duloxetine
   0.27, 0.55, 0.18  # Placebo
), nrow = 2, ncol = n_cat, byrow = TRUE)
strata_frac_alt <- array(c(
    strata_frac_alt_BIFl28, strata_frac_alt_BIFgeq28
), dim = c(2, n_cat, n_strata))
samplesize.vanElteren(strata_frac, treatment_frac, strata_frac_alt, alpha, beta) # should be 228

kkmann/vanElteren documentation built on May 25, 2019, 8:29 p.m.