Non-parametric Kruskal Wallis data example

Artificial dataset to show that the p-value obtained for the Kruskal Wallis is only valid _after_ the distributional form has been checked to be the same for all groups.

Format

An artificial data frame with 18 observations in each of three groups.

x

measurements for group 1

y

measurements for group 2

z

measurements for group 3

Source

Data example found on the internet

Examples

data(kwdata)
newdata <- stack(kwdata)
kruskal.test(values ~ ind, newdata)
#> 
#> 	Kruskal-Wallis rank sum test
#> 
#> data:  values by ind
#> Kruskal-Wallis chi-squared = 7.3553, df = 2, p-value = 0.02528
#>