Hi,
I am using SPM2 to do some VBM. Recently, I started playing around with the
design definition structures. As I was going through the design definition
structures for the basic statistics designs (contained in spm_spm_ui.m) I
noticed that all of the basic designs, with the exception of the t-tests,
contain a replication factor. Indeed, when I inspected the SPM structure
after running spm_spm_ui (selecting, say, multiple regression with constant)
I found that the SPM.xX.sF and SPM.xX.I fields displayed the replication
factor, and that the replication factor levels always just corresponded to
the number of the scan. However, the HForm and BForm specifications in the
design definition structures always leave out the replication factor, so the
replication factor never has an effect on the design matrix.
Now, please excuse me if this is an ignorant question (which it might be
because I haven't used SPM to analyze fMRI data yet), but why is the
replication factor included at all if it is going to be ignored in the design
matrix? Also, I would think that a replication factor would not make any
sense for a VBM analysis, right? So, should I be making my own design
definition structures to do VBM, or does it not matter because the
replication factor is ignored? Some clarification would be greatly
appreciated.
Cheers,
Blake Richards
|