[1 1 0 0] will give you the average effect of your regressor across
both sessions, which I think is what you want.
[1 -1 0 0] will tell you where the effect of your regressor differs
between sessions (as a t test, this would tell you where the effect is
bigger in the first session than the second).
Hope this helps,
On Tue, Mar 10, 2009 at 10:13 AM, SUBSCRIBE SPM Anonymous
<[log in to unmask]> wrote:
> Dear all,
> I have a question about specifying the contrast in single-subject
> multisession data.
> Due to the scanner issue I'd to break a single run into two sessions. I'd
> read many older messages and found the suggestion is to treat them
> separately as two sessions, and not to pool them together as a single run. I
> tried with SPM5 by choosing:
> Specify 1st-level => Data&design => New Subject/Session (add twice) => ...
> and specify the scans
> When specifying the contrast, should one use [1 1 0 0 ] or [1 -1 0 0 ]?
> Or should I use "Specify 2nd-level " instead?
> p.s. The design matrices look like these (.jpg).
> Any suggestions are appreciated