Thanks for the quick reply Anderson. However, I prefer to use the factor effects model, since I have used it in all previous models of my study. Moreover, if I understand the GLM help (http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/GLM) correctly, it is recommended to use the factor model since "the advantage of the factor effects approach is that the grand mean is not modelled, and thus there is no problem with rank-deficient design matrices".
So, further help is highly appreciated. Thanks in advance!
Henk
|