Dear group:
In performing a cluster-level analysis, I have read that the minimum
appropriate z-score(uncorrected) is 2.00. With a z score of 2, what would
be the minimum potential spatial extent probability value that might still
keep the false positive rate relatively low? I thought I read in the spm
notes that the lowest spatial extent probability value that still gave a
reasonably small rate of false positives was 0.15. Is this accurate or too
lenient?
Thank you very much for any help.
Jeff Lorberbaum
Medical University of South Carolina
PS- If it helps, I am performing an fMRI experiment in which I have 40
active minus rest condition comparisons, voxel size of 4 x 4 x4 mm,and a
smoothing kernal of 8 x 8 x8.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|