Dear Developers/Users!
I think I need some expertise in applying comparison of groups of models (families) in the new Parametric Empirical Bayes framework.
I have 11 subjects with 2 sessions each (22 task fMRI in total) and I constructed 10 models. The models are designed to represent 2 groups: in 5 models the experimental variable "stimulates" region A; and region B in the other 5 models.
I thought I can create regressors for grouping the models into family A and B, but I think I can only make a design matrix for group effect of subjects. (It is still useful as I can have the inference of session differences.)
Not sure if my approach makes any sense, I may have some misconceptions. Does anyone have experience with grouping models in PEB?
Thank you for your time,
Csaba Aranyi
|