That's great, thank you Steve for the clarification that it is grand mean scaled! Much appreciation.
Another question I have is from your reply:
"note that this is NOT the same as scaling each *timepoint* to a constant value."
- I was wondering, what is the difference between scaling each volume by the same amount, and scaling each timepoint? (How is it done generally statistically?) I know grand mean scaling should result in voxel means ~10,000, (as mean_func shows) but as I'm new to/still learning about FEAT, I don't understand how the mean can be changed by not scaling each timepoint constantly, and still preserve the timecourse changes from TR to TR.
Since we want to compare stimuli values across two runs, timecourse-wise, I just wanted to make sure we are using the correct file in filtered_func_data.nii.gz. We need blocks in run 2 to be on the same scale as blocks in run 1 for the stimulus, and have timecourses for each block be comparable.
Apologies if this is already explained somewhere in the boards/online, I have been searching but could not figure it out yet.
thanks!
Jake
|