hi all
I have 3 groups of subjects each making 3 fmri sessions with the same task. Each group execute a different task.
In order to average the 3 sessions (instead of merging them, as here suggested several times) i make a 2nd level fixed analysis.
In the FSL documentation is suggested to perform it at group level, in order to better estimate within sessions variance.
It is better to make one separate 2nd level analysis for each group (supposing different variance for each group)? or is it always better one single analysis with all the subjects ungrouped?
thanks in advance
cheers
Alberto
|