Post-hoc of Friedman test in multcompare function

Hello, when you perform a multcompare post-hoc analysis after a friedman test:

[p,tbl,stats]=friedman(X,1) %where X is a matrix NXM with M>2

multcompare(stats)

which kind of test is implemented in the multcompare function?

Answer 1

[the cyclist]

As explained in the documentation for multcompare, this is determined by the 'CType' (critical value type) input. The default is Tukey's honest significant difference criterion.