I think you should not have any problem unless something fails ;)
Could you send the error output from Matlab?
What do you mean by 'Matlab got no maps'? Have you checked the path in which
SPM.mat is stored? You could give a glance at:
http://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind04&L=spm&D=0&I=-1&P=145677
if your error is similar to the one described there, then you should change the
path where you store your 'con' files. Otherwise you should describe your error
a little more.
(you should use the same 'con' files as in a two-sample t-test, but sepparating
them correctly in two sepparated tests).
Cita de "[log in to unmask]" <[log in to unmask]>:
> Hi SPMers,
>
> I would like to look at differences between controls who never used
> drugs and those with a past history of drug dependence.
>
> I started the analysis with a two-sample t-test;
> Using a contrast of [1 -1] I got the areas where activity is greater
> in group 1(ex-users) compared to group 2(controls). There were
> significant areas particularly in the temporal lobe. The contrast
> of [-1 1] showed no statistically significant areas. Theoretically, I
> was expecting there to be areas of hypometabolism in group 1(ex-users)
> vs. group 2(controls).
>
> To further investigate the difference in activity, I wanted to look
> at the one-sample t-tests for each group. Either Matlab crashed during
> the process or got no maps. Is my analysis procedure wrong? What files
> should be used in one-sample t-tests?
>
> I have very small group, 7 patients in Group1, and 8 patients in Group2, all
> images (one scan/one condition) are normalized to our template, smoothed and
> globally normalized.
>
> Thank you,
>
> Jeremy
>
________________________________________________________
Juan José Lull Noguera - jualulno_at_upvnet.upv.es
[MI - Medical Imaging Area]
BET - Bioengineering, Electronics and Telemedicine Group
UPV - Politechnical University of Valencia - Spain
________________________________________________________
|