SPM Community-
I would appreciate some advice on constructing a proper repeated-measures anova. I have 19 subjects who were scanned:
1) pre placebo
2) post placebo
3) pre drug
4) post drug
The pre-post sessions are 1/2 hour apart, with the placebo/drug visits a week apart. I generated relevant contrast images for each of the 4 above.
With the flexible factorial design, I can construct a 2x2 within-subject model that seems correct, and that I believe is consistent with examples from earlier list postings and the document by Jan Glascher and Darren Gitelman. However, the given degrees of freedom is 54 for the interaction, which seems high. This also gives a slightly more liberal result than obtaining the interaction by 2 Imcalc subtractions and a t-test.
Could I have set up the model incorrectly or am I not getting the correct degrees of freedom? Or is it ok?
Thanks in advance for any advice!
Jason Tregellas
Assistant Professor
University of Colorado Denver
Denver VA Medical Center
|