Hi,
I am trying to compile a first level analysis using data from three different studies, each study has a main contrast of "Social > NonSocial" and I want to combine the data from the three studies to investigate which areas of the brain show "Social > NonSocial" activation. Two of these studies were collected using one MRI acquisition protocol (dual echo fMRI) and one study was collected using a different MRI acquisition protocol (spin echo fMRI). The issue is that the spin echo acquisition protocol produces much smaller activations across the brain compared to the dual echo protocol. I want to be able to compare the data between these three studies but accounting for (or weighting out) the overall effect of acquisition.
I've been doing some research into this and believe that using a covariate of acquisition protocol would be the best way in which to do this. Would anyone be able to advise me on whether this would be the correct approach?
My plan is to run a second level analysis (one-sample t test) inputting the Social > NonSocial contrast from each of the three studies and then add Acquisition Protocol as a covariate but weight the spin echo scans higher than the dual echo scans (e.g., 1 0.5 0.5), so that the higher power of the dual echo scans doesn't exclude the data from the spin echo data. Is a valid approach?
Best wishes,
Grace
|