Hi - you didn't explicitly say whether you expect (and are interested in)
a change between the two time scans for a given subject. Anyway, yes I
think you're on the right track; you want a design similar to the two
final examples at
and you want to either have the temporal factor as one of the ANOVA
factors, or take out the subject means as with the paired t-test example
higher up on that page.
Hope that makes sense! Cheers, Steve.
On Fri, 19 Nov 2004, Alison Burggren wrote:
> I have a fairly complicated design and have some questions about setting it up in
> FSL. I have 8 subjects, scanned twice, some of whom are on medication or not (fixed
> variable A) and some of whom carry a genetic risk factor (random variable B). Would
> it be correct to run a two sample paired t-test with this population but if so how do I
> model the other two variables, medication and genetic factor. Or would it be more
> correct to model this as a mixed effects anova in FSL? I thought that if I treat time as
> a third variable then I'm losing the power of modeling the contrast as a paired t-test
> (since the same subjects are scanned twice). I'm looking for theoretical input, not
> necessarily the design written out in full. Thanks for your time!
> Alison Clements Burggren, Ph.D.
> UCLA Brain Mapping Center
Stephen M. Smith DPhil
Associate Director, FMRIB and Analysis Research Coordinator
Oxford University Centre for Functional MRI of the Brain
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve