We are also wondering what the current advice is on this matter.
We have 3 sessions per subject which all contain the same experimental manipulations (i.e. the conditions in each session are the same). The subject remained in the scanner and the break between sessions was brief (less than a minute).
Is it preferable to:
A) Run separate DCMs for each session, then a FFX DCM for each subject, then RFX at the group level.
B) Concatenate the 3 sessions prior to pre-processing, and include additional session regressors in the GLM (modelling both session effects and the transitions between sessions)
Also, an explanation of the pros and cons associated with each option would be informative. E.g. Do the benefits of concatenating outweigh the potential issues with high pass filtering?
James.
|