Ninad,
There are, no doubt, more experienced minds around the SPM table than I,
but I'll give this one a shot.
If I'm correct, then you're going to have to run two design matrices:
one for those participants who do have a 'miss', that includes misses as
an additional regressor in the model, and one for participants that
don't have a 'miss', that doesn't include this regressor. SPM won't
accept '0' as an onset vector - because '0' means that there is nothing
at all to model (and so including '0' as a vector is really the same
thing as not including that regressor in the model at all).
When you run your contrasts, you'll be able to use all participants
regardless of which model was used. Just be careful to ensure that you
do your contrasts properly: the columns in your design matrix may differ
across the two models.
Hope that helps (and hope it's right!)
Matt Shane
Olin Neuropsychiatry Research Center
Institute of Living
Hartford, CT, USA
>>> Ninad Gujar <[log in to unmask]> 12/08/06 4:07 PM >>>
Hi SPMers,
Question: Can we enter a regressor value for a onset zero or null ? And
how
does the F-contrasts work in such a scenario ? Does it ?
Is there any other way to work around this ?
I am running a batch script for an e-r experiment. I am modeling subject
misses, but subjects do not always always make misses in each session.
Now,
I want to use the contrast vectors across subjects, it would be
convenient
to include the regressor for misses in the model for each session and
for
each participant. I've entered '0' for the onset vectors in these
cases,
and 0 for th regressor but SPM does not seem willing to accept it,
Also, it does not seem to create an F-contrast for the regressor applied
to
an array with one elemt in it. One of the session events has just one
regressor value for the corresponding onset time.
Could someone please advise the way to get around this.
Guesses, examples, and partial answers are very welcome.
Thanks in advance.
- Ninad
Confidentiality Notice
This e-mail message, including any attachments, is for the sole use of
the intended recipient(s) and may contain confidential or proprietary
information which is legally privileged. Any unauthorized review, use,
disclosure, or distribution is prohibited. If you are not the intended
recipient, please promptly contact the sender by reply e-mail and
destroy all copies of the original message.
|