Skip to contents

Mirrors the procedure described in section 5.1 of the paper.

Usage

kruskal_then_permute(
  value,
  group,
  B_perm = 10000,
  alpha = 0.05,
  verbose = FALSE
)

Arguments

value

numeric vector of observations.

group

factor or character vector of group labels.

B_perm

number of permutations per pairwise test.

alpha

global significance level.

verbose

logical; emit a cli trace of the decisions.

Value

list(omnibus, pairwise, letters, decision); `pairwise` is a tibble and `letters` is a compact-letter display compatible with the paper's tables.