Dear DCM experts
Recently we got the results of model comparisons in two different data sets
In both data sets we get the same pattern: an extreme preference for the
The weird thing is that it is unanimous across all 40+ subjects, with no
This is true even when we go to ridiculously simple models.
When looking at the 'relative cost' per region (in the model comparison output)
there is a unanimous preference for the most complex model for all regions, in
It seems like the model preferences here are determined only by apriori factors
while the specific data in each VOI plays very little role.
Does this pattern of results give any hint for where the problem could be?
I'd appreciate any help