Independence Tests for Count Data — independence

Perform an independence test between two categorical variables, using either the chi-squared method or Fisher's exact method.

Usage

# S3 method for class 'formula'
independence_test(
  formula,
  data,
  method = c("default", "chisq", "exact"),
  correct = FALSE,
  ...
)

# Default S3 method
independence_test(
  x,
  method = c("default", "chisq", "exact"),
  correct = FALSE,
  ...
)

independence_test(x, ...)

Arguments

formula: a formula of the form y ~ x, where y and x are both factor variables. If not factors, they will be automatically converted. To perform test within subgroups, use y ~ x | g.
data: a data frame containing the values in the formula.
method: character string specifying which method to use. One of "default", "chisq", or "exact".
correct: a logical indicating whether Yates' continuity correction should be applied; used for chi-squared test only.
...: additional parameters, currently unused.
x: instead of a formula, provide the matrix of counts directly.

Details

By default, the chi-squared method with no continuity correction is used unless the minimum expected count under the null is less than 5, in which case Fisher's exact test is used.