 |
 |
Dear Cole Please see https://en.wikibooks.org/wiki/SPM/Parametric_Empirical_Bayes_(PEB)#Example_design_matrices Let us know if that doesn't clear things up. Best P From: SPM (Statistical Parametric Mapping) <[log in to unmask]> On Behalf Of COLE JOHN COOK Sent: 18 June 2019 19:51 To: [log in to unmask] Subject: [SPM] Contrasts for PEB Model Dear SPM Users, I was wondering the best way to construct a contrast for a DCM PEB model after BMA. My model currently has three groups. The model matrix looks something like this 1 0 0 1 1 0 1 0 1 ... So the second column is one if the subject is in group 2 and zero if the subject is in either group 1 or 3 and the third column is one if the subject is in group 3 and zero if the subject is in either group 1 or 2. From this, I can easily see the differences between groups 1 and 2 from the coefficient for the second column and between groups 1 and 3 from the coefficient for the third column. However, I would like to see what the differences are between groups 2 and 3. Is this something that can be easily implemented? Would you recommend doing something akin to a contrast for GLM? Any thoughts would be greatly appreciated! Please let me know if you have any questions. Thank you very much. Have a wonderful day! Cole Cook