>
>
> | In the framework of spm99, is there any sense in which movement between
> | sessions is more benign than movement within a session? (I assume a
> | session is a time-contiguous set of scans; thus, one could have many
> | "sessions" in one afternoon at the scanner.) For example, if I decide I
> | don't like movement more than (say) 1mm/1 degree within a session, should I
> | extend this judgement to between sessions?
>
> It all depends how correlated the movement is with the effects of interest
> modelled by the design matrix. If you are not looking for interractions
> between sessions, then movement between sessions should not be a problem.
> This is especially the case if the different sessions are modelled in the
> same way as different subjects would be.
>
> Best regards,
> -John
>
However, movement between sessions would effectively
spatially blur the true pattern of activation (when assessed across
sessions). This would be so even though, as John points out, within-voxel
error variance would not necessarily be affected.
Sincerely,
Eric
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|