It's hard to say what the right thing to do is. In principle, the larger
activation you are seeing in 4.5 could be false negatives, and you
probably don't want something that is going to include those. You can
get to something closer to the 4.5 whitening by adding the following to
your mkanalys
So we won't actually be doing group analysis on this except ROI analysis.
We are using this data to look at overlap in activations on individual
subject levels. Since the v5 whiten seems much more extreme than v4.5, I'm
not really sure how to interprete this in terms of what I need to do for
accur
selxavg3-sess checks to see whether preproc has been run and runs it if
it is needed, so all that stuff can go on in the background. It does
look like the differences you are seeing come from the whitening. This
attempts to account for temporal correlation in the noise. Whether you
need it or
Hmmm, not sure what's going on with that. Can you verify that the
registration is correct in the v5 results? The only other thing that I
can think of is that the whitening may be a bit different between 4.5
and 5.0. Try turning off the whitening in v5 (-no-whiten when running
mkanalysis-sess).
