Re: Effect sizes for Friedman's ANOVA


Ollie wrote:
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.

This equation can be used to calculate the phi-coefficient from the
chi-square value based on the corresponding 2x2 table. I do not see how
this relates at all to your original question about calculating an
effect size in the context of Friedman's test.

m00es

.