I am using a set of pilot fMRI task data to identify ROIs that have significant activation in response to one condition vs. another. I then plan to test hypotheses in another sample with these pre-defined ROIs.
For these pilot participants, most participants completed two scans about a month apart. Ideally, I’d like to include both scans for each participant instead of throwing out data, but am unsure about what the best approach to is including multiple sessions of the same participant. My current plan is to just run the level 1 models collapsed across both scans for each person, and include a regressor for session # in these models. Does anyone have any other recommendations or know of any papers that have dealt with or done something similar? Thanks!