Hi Jack,
At second level I usually test for the significance of the temporal
derivatives of each lower-level EV. Very often I see significant effects of
the temporal derivatives at second level.
You can do it as follows:
First make sure that temporal derivatives are included in your first level
model (this is the default). When entering the first level contrasts, use
'Real EVs' instead of 'original EVs' . This allows you to calculate the
temporal derivative contrast. For example: 0 1 gives you the temporal
derivative contrast if you have only 1 EV and its temporal derivative in
your first level model. Then you perform a second level analysis on these
copes.
In you do a one sample t-test on these copes you can calculate two contrasts
(+1 and -1), giving you the average positive ('fast response') and negative
('slow response') effect of the temporal derivative.
Hope this helps,
Serge.
-----Original Message-----
From: Jack Grinband [mailto:[log in to unmask]]
Sent: Thursday, May 27, 2004 7:31 AM
To: [log in to unmask]
Subject: [FSL] PE for temporal derivative
Hi All,
I just wanted to confirm that FEAT only includes the non-derivative portion
of the model in its
hypothesis testing and that the parameter estimate associated with temporal
derivative is not
passed up to the group level. Is this correct?
If the derivative portion explains a significant portion of the variance,
you can get an artifactual
decrease in the non-derivative parameter estimate? Is this right?
I recently read a paper (Calhoun et al, 2004) which suggested using a
hypothesis test that includes
the parameter estimate for the derivative term. Does anyone have any
opinion on this?
Finally, is the PE for the derivative saved during the FEAT analysis? Where
could I find it?
thanks,
jack
|