Friedman test with between-subject

[Piotr Lewczuk]

Good afternoon,
Is there something like a non-parametric equivalence for repeated measurements ANOVA considering also between-subject effects (or split-plot ANOVA)? Something like Friedman's test but including between-subjects? A typical example would be: 20 participants divided into two groups according to their sex (between-subjects); from each person three measurements are performed (within-subjects).
Sorry if this already has been discussed, and thank you in advance for your kind support.
Piotr Lewczuk

[Joseph Coveney]

You would use Friedman's test for the within-subjects effects and van Elteren's test for the between-subjects effects. You can do both with the user-written command emh, which is available from SSC.

[Piotr Lewczuk]

Thank you!
If I have to use another test for between-subject effects anyway, what is the advantage of van Eltern's test over Kruskal-Wallis (or Mann-Whitney for two groups, like two sexes)?
Best greetings,
Piotr

[Piotr Lewczuk]

By the way: are you sure you mean "emh"? It says it calculates extended Mantel-Haenszel statistics, and I'm not sure if this is what I mean.
Greetings,
Piotr