Hello all,
A few questions regarding the function "contrast of connections" for BPA (fMRI). I am modeling the effects of three event types (so 3 DCM.B matrices).
1. To be sure I am using it correctly: if I want to test the contrast of two different effects on a single connection of interest (against a null of no difference in effect), in the GUI I use "Threshold: 0 Hz", "contrast for B", and then enter a "1" for the connection of interest in the first DCM.B matrix and a "-1" for the other matrix, leaving everything else as "0": is this correct?
2. After the above steps ^, in the GUI I get a nice plot of the posterior density of the contrast. However this does not appear to generate a file. Is there a way I can save and record the exact result (rather than just eyeballing the density plot), or otherwise find those numbers elsewhere?
3. In general, how does a contrast for differences in BPA parameters differ from doing a classical t-test (or ANOVA) on each subject's parameters for the model of interest (i.e., why specifically would results differ between the two?)
Thank you in advance for the help,
Jake Bowling
University of North Carolina at Chapel Hill
|