Dear SPM users,
My problem concerns an experiment with 2 populations, Family and
Controls,
where each subject of the group Family is paired with exactly one of
the
other group (15 couples).
The Family group is also divided in two subgroups: disease (5
subjects) and
not (10 subjects).
I would like to write in SPM5 a design matrix to study differences
in these
two subgroups (disease and not) but taking in account of the
differences
with the matched Controls.
It sounds like a group comparison via paired t-test by I'm not sure
how to
design it.
I wonder if it can be possible to write this experiment using a
single
design matrix.
I tried with the flexible factorial design, defining three factors:
Pairing
(with family subject and matched control), Family and Disease.
The I defined main effect of pairing, Family and Disease and
interaction
Family-Disease.
Resulting design matrix is attached.
Is this a right design? What kind of contrast can be defined?
Thanks in advance for any suggestion,
Marco
|