Hi Bettina,
>HI, everyone!
>Is it possible to make a conjunction analysis if I have performed a group
>analysis with one sample t-tests?
>I have a 2x2 design and would like to see the activation that is present
>on both levels of one factor.
>Calculating the results in spm, I cannot click on more than one contrast
>as the other t-tests have been calculated in different folders (otherwise
>the spm.mat-files would have been overwritten.
>Or do I need to perform an ANOVA?
>Thanks for your help
>Bettina
>
>
From a 2x2 design (A1 A2 B1 B2) I see several possibilities, depends of
your expe, for ex:
- if you want to see what's happening in both A1 & A2, you can make the
conjunction at the first level and then use the contrast at the 2nd
level (one-sample t test)
- if you want to look at the conjonction of the relative contribution to
each level of A in comparison with B then you can do it through an ANOVA
at the 2nd level with 3 contrasts C1=A1-B1, C2=A2-B2 and C3=C1UC2
be also aware of the type of conjunction, null or global ..
Hope it helps,
Best
Cyril
|