11-15-2012 08:11 PM
I am trying to obtain partial eta squared values, as a journal is asking for this. I used SAS PROC MIXED and have not been able to find anything about this. Does anyone know of a way to do this, or a round about way?
11-16-2012 08:00 AM
I wonder if this can be done, as partial eta squared values are usually calculated using sums of squares. One approach would be to calculate the noncentral F value, and then apply (partial eta squared) = (noncentral F value)/(noncentral F value + number of observations). I got this formula from the documentation for the GLM procedure. There are also formulas for a variety of estimators of the noncentral F.
Hope this helps.