Dear SPM experts,
I´m reposting this with the hope that someone can help me to specify a flexible factorial design with a covariate of interest.
My dataset consists of T1 images at 3 time points and 2 groups of patients (AB, BA). I ran a longitudinal voxel-based morphometry analysis and a flexible factorial model to examine groupxtime interactions. In this analysis, I found some significant effects of groupxtime for grey and white matter anatomical regions that I was expecting. Here one example of a contrast vector (T contrast) for the interaction (AB>BA x TP2>TP1) is: [-1 1 0 1 -1 0].
Next I want to set up the same flexible factorial but including a covariate with the questionnaire's scores (FAB) at the 3 time points (centering). To do so, I specified a total of 3 covariates of interest, each one including the FAB score at a timepoint (and padded with 0´s for the other timepoints). Also, I specified the ICV as a nuisance covariate. All this added 4 extra regressors to the end of the design matrix. Then if, for example, I have to specify the same interaction including the FAB as covariate of interest (AB>BA x TP2>TP1 & FAB_T1 FAB_T2), the contrast vector (T contrast) would be: [-1 1 0 1 -1 0 1 1 0 0], where the last 4 columns represent the covariates included in the design.
I´d appreciate if someone could check the code I wrote for this analysis and give me some feedback. The lines for the contrast in the second example are L89-91: https://github.com/nomamoli/VBM_multisubject_multisessions/blob/master/L2_Flex_TP1_TP2_N26_TP123_WM_FAB_ICV.m
Thank you in advance for your time!
Best,
Noelia
|