wilcoxon.mann.whitney {ANSM5}

Perform Wilcoxon-Mann-Whitney test

Description

wilcoxon.mann.whitney() performs the Wilcoxon-Mann-Whitney test and is used in chapters 6, 8, 9 and 12 of "Applied Nonparametric Statistical Methods" (5th edition)

Usage

wilcoxon.mann.whitney(
  x,
  y,
  H0 = NULL,
  alternative = c("two.sided", "less", "greater"),
  cont.corr = TRUE,
  CI.width = 0.95,
  max.exact.cases = 1000,
  nsims.mc = 1e+05,
  seed = NULL,
  do.asymp = FALSE,
  do.exact = TRUE,
  do.mc = FALSE,
  do.CI = TRUE
)

Arguments

x

Numeric vector, or factor with same levels as y

y

Numeric vector, or factor with same levels as x

H0

Null hypothesis value (defaults to NULL)

alternative

Type of alternative hypothesis (defaults to two.sided)

cont.corr

Boolean indicating whether or not to use continuity correction (defaults to TRUE)

CI.width

Confidence interval width (defaults to 0.95)

max.exact.cases

Maximum number of cases allowed for exact calculations (defaults to 1000)

nsims.mc

Number of Monte Carlo simulations to be performed (defaults to 100000)

seed

Random number seed to be used for Monte Carlo simulations (defaults to NULL)

do.asymp

Boolean indicating whether or not to perform asymptotic calculations (defaults to FALSE)

do.exact

Boolean indicating whether or not to perform exact calculations (defaults to TRUE)

do.mc

Boolean indicating whether or not to perform Monte Carlo calculations (defaults to FALSE)

do.CI

Boolean indicating whether or not to perform confidence interval calculations (defaults to TRUE)

Value

An ANSMtest object with the results from applying the function

Examples

# Examples 6.1 and 6.2 from "Applied Nonparametric Statistical Methods" (5th edition)
wilcoxon.mann.whitney(ch6$groupA, ch6$groupB)

# Exercise 12.4 from "Applied Nonparametric Statistical Methods" (5th edition)
wilcoxon.mann.whitney(ch12$feedback.satisfaction[ch12$PPI.person.2 == "Representative"],
  ch12$feedback.satisfaction[ch12$PPI.person.2 == "Researcher"],
  do.exact = FALSE, do.asymp = TRUE)