tukey's test
Given a bunch of samples, what are the significances of the differences for all pairwise means?
That is: the inputs are an array of \(N\) samples (each sample is also an array). The output is an \(N \times N\) matrix, where each \((i,j)\) entry is the p-value for the difference between the i-th and j-th sample.
1. comparison with dunnett's
- dunnett's test makes a smaller claim: "all these samples are different than this one control."
- or, in another light: "all these samples are worse than this one control", which you might care about if you want to show that your control beats everything else.
- if this is what you are looking for, dunnett's is the more powerful test