Dear FS list (dear Martin),
I have a 3 groups x 2 timepoints data set, and I did
analyses in QDEC comparing my groups pairwise (2 at a time)
in terms of the pre-to-post percentage change (pc1) in
various surface-based metrics like thickness and area.
Since a group contrast is a bit difficult for readers to
interpret (it would read "where is the percentage change from
pre to post greater for group 1 than for group 2?"), I am
wondering what is, in fact, best to report, aside from showing
images of the significant clusters. What commands would I have
to use, exactly, to for instnace, compute the group means for
pre and for post values for e.g. thickness, i.e. absolute
values rather than differences? I think I only have the pc1's
calculated, since if I select thickness (rather than
thickness-pc1) in QDEC, it tells me it cannot find the
smoothed files.