Dear Klaas and DCM experts
We have a question about how use of BMA when our goal is to contrast 2 specific connections:
In your paper (Penny et al. Plos Computaional Biology, 2010, p.7) it says:
"Model averaging can also be restricted to be within subject
(using equations 30/31 and 32 only). Summary statistics
from the resulting within-subject densities can then be entered into
standard random effects inference (eg using t-tests) [19]".
Does that mean: use the FFX BMA within each subject, and then use the resulting parameters in a standard t-test?
Is there a way to use RFx BMA to compare between connections, or the above is the only way?
Thanks a lot
Tali Bitan
University of Haifa
|