Dear all,
I have the same questions. Any hint on this is greatly appreciated.
Best regards,
Feng Zhou

On Wed, May 17, 2017 at 3:35 AM, Helka Ozelo <[log in to unmask]> wrote:
Dear SPM experts,

In a DCM - RS analysis to a single subject I've specified and estimated a full model with 4 nodes (PCC, mPFC, LIPC and RIPC), but it´s not clear to me:

1. How could I get the 64 possible models (2^n(n-1)/2) from the full model? Should I specify them one by one?

2. How to reduce the number of models (Bayesian model reduction) at SPM in order to compare them?

I really appreciate any help you can provide.

Helka