Hi, all.
First and foremost, thank you Alfonso for the added CONN preprocessing and simple model methods text generator feature. I do think this will aid in reproduceability of any study using the CONN platform.
One of the questions I have, however, is related to how CONN
sets up mixed models in SPM. How should one describe how SPM
handles analysis of say a MVPA regression model w/ 4 principal
components and the following model:
B/W subjects (0 1 0) ---> 0 subjects; 1 duration variable; 0
denotes age covariate
W/N subjects (-1 1) ---> post-tx > pre-tx
four eigenmap PCI components (1 0 0 0; 0 1 0 0; etc.)
We're performing regression, but the model is structured like a
mixed-model MANCOVA, I believe.
If one were to run this analysis outside of CONN and only w/n SPM,
then how would you describe the statistical model and its
components in the methods?
Warm regards,
Jeff Browndyke