I don't often venture into the statistics part of SPM, but I'd expect it to be possible to set up such an analysis via many different routes. I guess you could use ANOVA, and specify that there are 4 groups, and then enter processing speed as a covariate. I don't know if you you would also want to model interactions between processing speed and group membership though. You should probably also model out other possible confounds (age, sex, TIV etc) too. Best regards, -John On 15 July 2014 17:36, Audreyana Jagger <[log in to unmask]> wrote: > Dr. Ashburner and colleagues, > > Sorry for the blank email everyone! > > I am trying to set up a statistical design for a VBM analysis within SPM. > I have 4 clinical groups, global gray matter values and I would like to use > processing speed (continuous variable) as covariate/regressor. I would like > to see how the groups' gray matter relates to processing speed. How would > you suggest setting up the analysis? Should I use a multiple regression or > an ANOVA? > Thank you! > > > Audreyana Jagger, B.S. > Graduate Student > Brain and Cognitive Sciences Program > Psychology Department > Southern Illinois University-Carbondale >