Dear spm list,
we have acquired a dataset with fMRI and would now be interested in doing
a statistical analysis on a multiple conditions basis.
Briefly, we have one group (15 subjects) who performed each six different
tasks:
Baseline a, Imagery 1, Imagery 2, Baseline b, Perception 1 and Perception 2
We first performed a 1st-level single subject analysis with six conditions x
subject.
In a 2nd level analysis, we would like to test if the task component 1
(tested for by the contrast Imagery 1 versus Imagery 2 and Perception 1
versus perception 2) produces common activations.
We saw on the list that multiple regression analysis was proposed to
perform a 2 nd level conjunction. We first implemented two groups of
con???.img corresponding to each of our contrasts in providing two
covariates. However it was not possible to define contrasts whose sum
differs from zero [1 0] or [0 1].
Please could someone indicate us a way of performing a 2nd level
conjunction!
Angélique Mazard
UMR 6095 CNRS-CEA-Universités de Caen et Paris V
Cyceron, Bd Becquerel BP 5229
14074 Caen Cedex
tel: +33 (0)2 31 47 02 67
fax: +33 (0)2 31 47 02 22
|