Dear SPM ers:
I am trying to find correlations between changes in binding of a PET
ligand [(+ or -) baseline + (- or +) Activation] and the plasma level
of a drug that induced the "activation" , it is understood that
baseline have zero plasma drug level.
(there are 10 subjects (A to J) with one PET baseline and one PET activation).
In SPM96 I was using: ANCOVA (should I have used instead PROPORTIONAL
SCALING ? ), multisubject:conditions and covariates, mean centering
(=m) the drug (=d) level (and entering
1) A to J data, baseline followed by Activation
2) covariate [ 0 (md)A 0 (md)B ..............0 (md)J ]
3) contrasts [0 0 1] and [0 0 -1] for negative and positive correlations.
In SPM99, I tried to replicate the same design using multisubject:
conditions and covariates, however I had 3 options for global norm:
ANCOVA BY SUBJECT, PROPORTIONAL SCALING and NO NORMALISATION.
ANCOVA BY SUBJECT will give zero residual degrees of freedom. So the
only options remaining are iPROPORTIONAL SCALING and NO NORMALISATION
Is there a way to get around that in SPM99 to get an ANCOVA for this
10 paired studies or is ANCOVA inappropriate and why.
Thanks for any help
Didine
--
Badreddine Bencherif, MD
Department of Radiology
Division of Nuclear Medicine
Johns Hopkins University School of Medicine
601 N. Caroline St. / JHOC 4230
Baltimore, MD 21287-0855
Phone : (410) 614-2787
Pager : (410) 283-2050
Fax : (410) 614-1977
email : [log in to unmask]
|