Hi,
I would like to model all medication effects as a covariate on second level.
The patient group has three main groups of medication (A, B, C). I was
thinking of adding A, B and C as three EV's (orthogonalised wrt to patient
group EV) and do an F-test on all three. Would that be the proper way to do
it? Thanks.
Regards,
Benny Liberg
|