Dear Cindy
> We have a data set consisting of 6 depressed subjects and 6 control
> subjects with 2 baseline scans and 2 scans following a pharmacological
> challenge. We have Hamilton Rating Scale for Depression scores (HRSD)
> for each subject. Is it worthwhile to do a covariate analysis with the
> HRSD score for the two groups given the small sample size of each
> group?
Yes indeed. If you have a score for each subject (as opposed to each
scan) than you can look only for a score x pharmacological challenge
interaction. Doing this for both groups may be redundant in the sense
that depressed subjects will have higher scores and this simple reduces
to the group x challenge interaction. However there may be extra
variance accounted for by within group score x challenge effects that
you could look for.
I hope this helps - Karl
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|