Dunn's test (post-hoc pairwise comparisons after Kruskal-Wallis)

Dunn's test: all pairwise rank-based comparisons following a Kruskal-Wallis test (kruskalTest), using the shared tie-corrected rank variance and a chosen multiple-comparison adjustment. It is the non-parametric counterpart of Tukey's HSD (tukeyHSD). The z-statistics and adjusted p-values reproduce base R's Dunn formula, the PMCMRplus package and scikit_posthocs.posthoc_dunn to machine precision.

Usage

dunnTest(x, ..., p_adjust = c("bonferroni", "none", "holm", "BH"))

Arguments

x

Either a named list of k >= 2 numeric group vectors, or the first group's numeric vector (with further groups passed through ...).

...

Additional numeric group vectors when x is a single vector.

p_adjust

Multiple-comparison adjustment: "bonferroni" (default), "none", "holm" or "BH".

Value

A data frame with one row per pair: comparison, z, p_value (two-sided, unadjusted) and p_adj, carrying k, n as attributes.

See also

kruskalTest, tukeyHSD

Examples

dunnTest(list(a = rnorm(20), b = rnorm(20, 1), c = rnorm(20, 2)))
#>    comparison        z      p_value        p_adj
#> b         b-a 3.032947 2.421777e-03 7.265332e-03
#> c         c-a 5.359716 8.335280e-08 2.500584e-07
#> c1        c-b 2.326769 1.997758e-02 5.993274e-02