Dear Dr. Zeidman,
In our full DCM model, two regions receive the input (matrix C); the intrinsic connections are fully connected (matrix A); and the modulation of two (out of three) conditions (matrix B) test on all intrinsic connections. Within this model, we got a poor explained variance (between 0% to 7% for more than half of the cases). However, when we did not include matrix C, the explained variance has improved significantly (between 19% to 44% for all cases). Why does the exclusion of matrix C have significantly improved the explained variance? What could be the solution for this case?
Thank you,
Nofar
|