Dear SPM community,
I have a statistical question about the comparability between fixed effect designs (1st level) at the subject level.
Is it necessary that the number of estimated column in the design matrix (1st level) to be the same between subjects to make a group analysis?
The degree of freedom of 1st level contrasts sent to random effect (group analysis) are not the same between subject 1 (12 estimated column) and subject 2 (10 estimated column). Is it a problem of comparability for these subjects? Should I use always exactly the same design matrix between participants to make a group analysis?
Michaƫl
|