ANOVA VS GAMj

[kevinchan0903]

I am running a repeated measures ANOVA involving only 1 IV and 1 DV.

I have found that the F values and the df in Post-hoc test were different between the results from ANOVA and GAMj.

Indeed, the results from ANOVA is the same as SPSS output.

I am thinking the different results are due to the different estimators, am I correct?

[jonathon]

in GAMLj i assume you're using the mixed model? a mixed model isn't the same as an RM ANOVA i'm pretty sure. let me ask marcello.

jonathon

[mcfanda@gmail.com]

Yes, RM ANOVA and mixed give the same results if the repeated measures are almost perfectly spherical (sphericity assumption is met quite well). Otherwise, they are different