Dear Ana,
If you can't find an answer, you could use fMRI-CPCA
https://www.nitrc.org/projects/fmricpca
http://www.cnoslab.com/cpca.html
it uses a FIR model, so that a network-level HDR shape is extracted
for every condition and subject separately. All the condition
comparisons can be carried out in a standard statistics package like
SPSS on the HDR shapes. The 9 level factor would just be one repeated
measure with 9 levels.
Best,
Todd
--
Todd S. Woodward, Ph.D.
Cognitive Neuroscience of Schizophrenia Lab (CNoS)
http://www.cfri.ca/our-research/researchers/results/Details/todd-woodward
http://www.cnoslab.com/donate.html
604-875-2000 x4724
*****
Associate Professor
Department of Psychiatry
University of British Columbia
BC Mental Health and Addiction Research Institute
Vancouver, Canada
On Tue, Dec 1, 2015 at 1:55 AM, Ana Paqui Palenciano
<[log in to unmask]> wrote:
> Good morning everyone!
>
> We have a 2x2 factorial design within-subjects, modelling the events with FIR. We want to run a repeated-measures ANOVA at the second level using the flexible factorial model of the SPM. So, in that model, we included: (1) Subject Factor, (2) Factor A, with 2 levels; (3) Factor B, with 2 levels; and (4) a "Time Bin" factor, with as many levels as time bins we defined in the FIR model (9 levels), as it was recommended on the SPM12 manual.
>
> In a previous discussion (https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1509&L=SPM&P=R70297&I=-3&d=No+Match%3BMatch%3BMatches), we read that the correct way to assess the effect of our manipulation, when the FIR model was being used, was attending to the interaction of that factor with the time bin factor. However, we don't know exactly what vector/matrix of contrasts we should introduce. All the examples we found in papers (Gläscher & Gitelman, 2008) only include factors with 2 or 3 levels. So, we don't know how to generalize that to our case, with a 9 level factor. Could anyone help us?
>
> Thank you!!
>
> Ana Paqui
|