Lee Friedman wrote:
> Thanks Gang.
>
> I do mixed models all the time with SAS PROC MIXED and there
> is always a residual term that SAS estimates that is not
> individual for each contrast but is general to the entire
> model.
> Perhaps this residual term would be the proper term to measure
> the smoothness of.
>
> I can send some SAS output if you like.
>
> Lee
A general residual term seems ideal for this situation. How do you calculate it? What is SAS PROC MIXED?
