dear SPMers
I'm correlating brain metabolism (FDG-PET) with perfomance at a memory test in patients with dementia
in doing this I need to keep into consideration the influence of age, sex, MMSE score and a measure of cognitive reserve on one or both of the two variables of interest (brain metabolism and the memory score)
having so many nuisance covariates will probably reduce degrees of freedom in the SPM analysis?
I was therefore thinking to regress the memory score to age, MMSE and cognitive reserve scores, and then use the resulting standardized (and 'covariated') score for the correlation analysis in SPM including only sex as a covariate in SPM
do you think this would be correct...?
thank you
FR
|