rankb {IncomPair}

Rank Based Tests for Incomplete Paired Data

Description

The function performs testing the hypothesis of equality of means for the incomplete pairs setting data. The function uses a rank-based procedure for parameter estimation and hypothesis testing when the data are a mixture of paired observations and independent samples. The rank-based methods combine Wilcoxon signed-rank statistics and Wilcoxon-Mann-Whitney two-sample procedures. These methods were developed by Dubnicka, Blair and Hettmansperger (2002).

Usage

rankb(xp, yp, xu, yu, mu = NULL, method = "Ranku",
      alternative = "two.sided", verbose = TRUE)

Arguments

xp, yp

(non-empty) numeric vectors of data values of the the complete pairs

xu

a numeric vector of data on x only

yu

a numeric vector of data on y only

mu

a number indicating the true value of the mean (or difference in means if you are performing a two sample test)

method

a character string specifying the different type of methods, must be one of "Ranku" or "Rankw"

alternative

a character string specifying the alternative hypothesis, must be one of "two.sided" (default), "greater" or "less"

verbose

if TRUE, show the test used, null and alternative hypotheses in addition to the p-value

Value

A S4 object containing the following components:

Title

a character string describing the test used

Nhypothesis

a character string describing the null hypothesis

Ahypothesis

a character string describing the alternative hypothesis

Pval

the p-value for the test

References

Dubnicka, S. R., Blair, R. C., & Hettmansperger, T. P. (2002). Rank-based procedures for mixed paired and two-sample designs. Journal of Modern Applied Statistical Methods, 1(1), 6.

Examples

n=20
n1=15
n2=10
r=0.8
xp=rnorm(n)
yp=r*xp+(1-r)*(rnorm(n))
xu=rnorm(n1)
yu=rnorm(n2)
mu=0
rankb(xp,yp,xu,yu,mu,method="Rankw",alternative="two.sided")