permTest: Permutation test

Description

Permutation test to test a hypothesis involving two samples.

Usage

permTest(x, ...)
# S3 method for default
permTest(
  x,
  group,
  statistic = mean,
  B = 9999,
  alternative = "two.sided",
  plot.hist = TRUE,
  plot.qq = FALSE,
  xlab = NULL,
  ylab = NULL,
  title = NULL,
  seed = NULL,
  ...
)
# S3 method for formula
permTest(formula, data = parent.frame(), subset, ...)

Value

Returns invisibly a vector of the replicates of the test statistic.

Arguments

x: a numeric vector. If the function is the mean (fun = mean) and x is a binary numeric vector of 0's and 1's, then the test is between proportions.
...: further arguments to be passed to or from methods.
group: a factor variable with two levels. If group is a binary numeric vector, it will be coerced into a factor variable.
statistic: the statistic of interest.
B: the number of resamples (positive integer greater than 2).
alternative: the alternative hypothesis. Options are "two.sided", "less" or "greater".
plot.hist: a logical value. If TRUE, the permutation distribution of the statistic is plotted.
plot.qq: a logical value. If TRUE, then a normal quantile-quantile plot of the resampled test statistic is created.
xlab: an optional character string for the x-axis label
ylab: an optional character string for the y-axis label
title: an optional character string giving the plot title
seed: optional argument to set.seed
formula: a formula of the form y ~ group where y is numeric and group is a factor variable.
data: a data frame with the variables in the formula.
subset: an optional expression specifying which observations to keep.

Methods (by class)

permTest(default): Permutation test
permTest(formula): Permutation test

Author

Laura Chihara

Details

Permutation test to see if a population parameter is the same for two populations. For instance, test \(latex\) where \(latex\) denotes the population mean. The values of the numeric variable are randomly assigned to the two groups and the difference of the statistic for each group is calculated. The command will print the mean and standard error of the distribution of the test statistic as well as a P-value.

Observations with missing values are removed.

References

Tim Hesteberg's website: https://www.timhesterberg.net/bootstrap-and-resampling

Examples

Run this code


permTest(states03$ViolentCrime, states03$DeathPenalty)

#using formula syntax
permTest(ViolentCrime ~ DeathPenalty, data = states03, alt = "less")

Run the code above in your browser using DataLab