I am having a hard time understanding some statistical issues in FSL...
I have analyzed my data at the third-level (across subjects) and know what
it shows me at the p < 0.05 level uncorrected. Am I correct in understanding
from this third-level analysis I can take the "resels" value that is in the
"smoothness" file, and divide this number into the number of voxels in my
mask? I know then I am to run the ptoz 0.05 -g <value from above> but what
exactly does this tell me? Is this more strict than the uncorrected stats?
I just need a little guidance as I am a bit confused. In addition, is there
a way to extract percent signal change from FSL data?
Thanks for any help!
|