Dear SPM experts,
I have a couple of general questions on spectral DCM implemented on resting state data, after implementation of a single fully connected model on two groups :
- I'm looking at Ep.A for estimates of effective connectivity between two groups. Anything else I should look for to compare two groups?
- I optimized each model for each subject before doing t statistics on effective connectivity parameters between the two groups. Is that OK or should I use the non-optimized model? I read that the model needs to be exactly the same between the two groups so that the analysis is valid, and I don't know if optimization on the subject level may change the structure of the model.
Thanks in advance.
George
|