When you say that I've already defined my groups in my contrast are you referring to the 92 rows and 3 columns relating to group assignments (design matrix)? Does that mean I do not need to demean my data because I'm modeling it in the design matrix?
In reference to the 7th contrast Con > Rel, Scz. We've had issues with this interpretation. Previously I've run cluster thresholding and then used the mask from that contrast to pull the % signal change at the subject mean level. Additionally I've tried running my task contrast of interest (let's say A-B) as a group mean and then pulled the % signal change and the subject level.
Do you have any suggestions about how to compare 3 groups?
Also once I have my randomise input I was told I need to re-threshold it in FSL view and set the intervals for activation shown from .95-1. Is that correct?
-Edward H. Patzelt
University of Minnesota
|