StatMethRank (version 1.3)

independence.test: Nonparametric Rank Tests for Independence

Description

This function performs a nonparametric test of ranking data based on the correlation. This function can be applied to the ranking data with missing ranks and tie ranks.

Usage

independence.test(X1, X2, method = c("spearman", "kendall"))

Arguments

X1

a vector, using NA to stand for the missing ranks

X2

the same as X1

method

whether the test is based on Spearman correlation or Kendall correlation

Value

a list of the test statistics

References

Rank Correlation Methods for Missing Data, Mayer Alvo and Paul Cablio Nonparametric Rank Tests for Independence in Opinion Surveys, Philip L.H. Yu, K.F. Lam, and Mayer Alvo

Examples

Arith = c(14, 18, 23, 26, 27, 30, 40, NA, NA)
Lang = c(28, 14, 46, NA, 53, NA, 54, 50, NA)
independence.test(Arith, Lang, method = "spearman")
independence.test(Arith, Lang, method = "kendall")