matching.test: Matching test

In stat.extend: Highest Density Regions and Other Functions of Distributions

Description

matching.test performs the matching test.

Usage

matching.test(
  x,
  size,
  null.prob = 0,
  alternative = "greater",
  approx = (length(x) > 100)
)

Arguments

x	Sample vector containing values from the generalised matching distribution
size	The size parameter (number of objects to match)
null.prob	The null value of the probability parameter
alternative	The alternative hypothesis
approx	A logical value specifying whether to use the normal approximation to the distribution

Details

The matching test considers a situation where a person attempts to match a set of objects to a corresponding set of positions. The null hypothesis is that the matching is performed at random and the alternative hypothesis is that the matching occurs with some ability on the part of the player.

For data vectors x that are not too large we perform an exact test by computing the exact distribution of the mean number of matches. If the number of data points is too large then we perform an approximate test using the normal approximation to the distribution of the mean number of matches. The parameter max.m sets the maximum number of data points where we perform an exact test.

Value

An htest object giving the output of the matching test

stat.extend documentation built on Nov. 23, 2021, 5:06 p.m.