This is a bug in QDEC. I'm surprised it has not surfaced before. It is
supposed to be looking in "fwhm09" not "fwhm9". One work-around is to
create a symlink, ie,
cd /usr/local/freesurfer/average/mult-comp-cor/fsaverage/rh/cortex/
ln -s fwhm09 fwhm9
otherwise, you will have to run mri_glmfit-s
Thanks Doug. After using stage#1 of the longitudinal 2-stage model to
create cross-sectional measuers of e.g. thickness change from tp1 to tp2
(e.g. thickness-pc1), I then did analyses in QDEC on these measures to see
where in the brain such changes are
- different from zero (within each group)
It only needs to be accounted for when you compare across all 3 groups,
in which case you'd have to use mri_glmfit
doug
On 6/17/14 12:07 AM, Tudor Popescu wrote:
Hi Doug,
Thanks... Presumably I'd just leave each possible pair of groups at a
time in the qdec table, and delete rows correspon
Hi Doug,
Thanks... Presumably I'd just leave each possible pair of groups at a time
in the qdec table, and delete rows corresponding to subjects of the
remaining group? And then do stats on, e.g. "long.thickness-rate" to
compare rate of change between the current pair of groups?
Doesn't variabilit
Yes, you could do each separately.
doug
On 6/16/14 8:10 PM, Tudor Popescu wrote:
Dear FS list,
I have a data set with 3 groups (2 treatments, 1 control), each with
equally-spaced time-points (pre and post structural scan). I've done
the 3 longitudinal pre-processing steps, and stage#1 of th
Dear FS list,
I have a data set with 3 groups (2 treatments, 1 control), each with
equally-spaced time-points (pre and post structural scan). I've done the 3
longitudinal pre-processing steps, and stage#1 of the two-stage model, and
I would prefer to run stage#2 (cross-sectional analysis of the di
