Permutation tests for experimental data

What is it about?

We provide an intuitive introduction to permutation testing for data collected in controlled experiments. This methodology offers an attractive alternative to normal-theory and asymptotic statistical testing when working with small samples. Familiar rank-based tests like the Wilcoxon signed-rank test and Mann–Whitney U test are special cases of the general permutation framework that we think experimenters should adopt.

Featured Image

Photo by Naser Tamimi on Unsplash

Photo by Naser Tamimi on Unsplash

Why is it important?

Modern computing power makes continued reliance on older rank-based tests unwise in many applications. We demonstrate how researchers can construct tests that make more natural and more complete use of collected sample data.