Dear SPM experts,
I tried for the first time the new mixed-effect module implemented in SPM12b.
My first level analysis contained four sparse-sampling sessions per subject with one t-contrast of interest per session respectively. I have two groups I'd like to compare using these 4 contrasts to perform a group x session analysis. I successfully used the fullfactorial module before but understand that the mixed-effects module would also take the within-subject error into account (similar to GLMflex I guess) and would thus yield more accurate results.
FFX specification and estimation worked well. Then I specified and ran the MMX model, which terminated with the message that the SPM.mat file was succesfully safed. Now this may be a naive question but the manuals says "SPM will create a subdirectory DIR/mfx and place an SPM.mat file there. This is the MFX model". However, no such subdirectory was created but instead, the FFX SPM.mat file seemed updated after the MFX specification job terminated sucessfully.
Before I was wondering how I would define my groups and sessions as I did in the fullfactorial model but now I would need to get past the MFX estimation step first.
Any suggestions?
Many thanks in advance,
Boris
|