sampleSizeMcNemar: Sample size for McNemar test

In felix-hof/biostatUZH: Misc Tools of the Department of Biostatistics, EBPI, University of Zurich

Sample size for McNemar test

Description

Compute the sample size to test the null hypothesis of the McNemar test.

Usage

sampleSizeMcNemar(p1, p2, sig.level = 0.05, power = 0.9)

Arguments

p1	Assumed marginal probability of the first row.
p2	Assumed marginal probability of the first column. Has to be larger than p1 and 0.5.
sig.level	Significance level.
power	Power, i.e., 1 - probability of type II error.

Details

Given two samples of paired binary observations, e.g., results of some positive/negative test on the same experimental units, we want to assess the null hypothesis whether the probability of positives by the first method is equal that of positives by the second method.

Value

Vector of min, max, and mid sample size as explained in Lachenbruch (1982).

Author(s)

Kaspar Rufibach
[e-mail]

References

Lachenbruch (1992). Sample Size for Studies based on McNemar's test Statistics in Medicine, 11, 1521–1525.

Examples

# example from Lachenbruch (1982), Table II, first row
sampleSizeMcNemar(p1 = 0.8, p2 = 0.9, sig.level = 0.05, power = 0.9)

felix-hof/biostatUZH documentation built on Sept. 27, 2024, 1:48 p.m.