Dear SPMers:
When I was doing DCM, in order to obtain the effective connectivities, we applied the random effects of Bayesian model averaging (BMA) to acquire subjects’ connectivity estimates across all models (42 models) based on the group winning family. Because I want to analysis the correlation between the connectivities and behaviors. However, for some subjects, the BMA estimates of connectivities are negative. What does the "negative connectivities" mean?
All the bests,
Daiming
|