Hi All,
The t-stat sub-brick for 3dANOVA3, type-4 bmean was
not correct. My fault.
This was due to part of the computation coming from
an uninitialized variable, which has been fixed, along
with the mis-labeling of the aBdiff sub-brick.
For a discussion, see :
[
afni.nimh.nih.gov]
Thanks to Debbie from U-Toronto for finding this problem.
The current source has been updated, along with some of
the binaries, as Bob has noted.
Sorry,
- rick