I've set up a multisubject, fixed-effect analysis and have performed a few
contrast conjunctions among subjects to visualize regions shared by all
subjects for the contrasts in question. So far, so good. However, I have a
question about FWE/FDR correction and the expression of T-values for the
conjunctions. I have Z scores in the mid-3 to high-4 range, but these
translate to T-value ranges of 0.47 to 0.85. Are T-values in the case of a
multisubject, fixed-effect conjunction an expression of effect size?
Also, when given the option in the SPM2 GUI, I selected an FDR of .05, but
was not given an opportunity for a k threshold (which typically means no
suprathreshold voxels). However, I have a fairly significant blobbed glass
brain for the conjunction.
Regards to all,
Jeff
-------------------------------
Jeff Browndyke, Ph.D.
Duke University Medical Center
|