Re: Effect sizes for Friedman's ANOVA

In the previous reply the symbols weren't printed so the expression is as follows:

rho=sqrt(Chi²(1)/N) and it seems to be larger than one for some of my variables.

Maybe I am mistaken in the way to compute N.

Rosenthal considers N to be the "size of the study" or 'other sampling units'. Does this means that in a repeated measures design

N = (k*n)

with k= number of repetitions and n the number of subjects????

or is it simply the number of subjects???
.