Hi all,
I've been running into this problem quite frequently with results assessment
in SPM5 and wondered whether anybody else had the experience or a solution.
I've updated my SPM5 as of about 2 months ago....
On about 20% of the times I hit the results button and chose an SPM.mat from
some new folder (so far only for second level models and never yet for first
level models) I get the specified contrasts for a completely different, but
recently interrogated second level model (eg looking to assess A-B in
controls, I get the contrasts specified for the A-B model for patients).
Once the wrong contrasts come up, they remain there when you re-attempt to
load them. If you click through, thinking that it's just the label that's
different, you get a resulting map which looks somewhat like the original
map you were going for, except that all of the significant voxels are red
with no color-coding of significance. The only way I've gotten around this
(and this certainly is a good way to duplicate effort and waste time) is to
re-run my batch specification for that model and re-set up the contrasts of
interest.
any ideas????
many thanks,
Amit
|