I'd model each event with its duration. This should work for PPI as
well. Just make sure you have a sufficient number of trials in each
imaging run. Having only a few trials will cause problems with
modelling due to collinear PPI and Task regressors with limited
trials.
Best Regards, Donald McLaren
=================
D.G. McLaren, Ph.D.
Research Fellow, Department of Neurology, Massachusetts General Hospital and
Harvard Medical School
Postdoctoral Research Fellow, GRECC, Bedford VA
Website: http://www.martinos.org/~mclaren
Office: (773) 406-2464
=====================
This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED
HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is
intended only for the use of the individual or entity named above. If the
reader of the e-mail is not the intended recipient or the employee or agent
responsible for delivering it to the intended recipient, you are hereby
notified that you are in possession of confidential and privileged
information. Any unauthorized use, disclosure, copying or the taking of any
action in reliance on the contents of this information is strictly
prohibited and may be unlawful. If you have received this e-mail
unintentionally, please immediately notify the sender via telephone at (773)
406-2464 or email.
On Fri, Jul 27, 2012 at 5:17 PM, Bob Spunt <[log in to unmask]> wrote:
> Dear SPM experts,
>
> I am currently designing an fMRI task with which I plan to closely examine
> condition-dependent connectivity using PPI. My understanding is that
> blocking stimuli by condition is the most powerful approach. My concern,
> however, is that I expect different conditions to yield different RTs to the
> stimuli. Moreover, I am interested in group differences in
> condition-dependent connectivity and am unable to ensure that RTs are
> matched across groups. Would a solution to this issue be to design the task
> in a quasi-self-paced manner, such that the average ISI (within-block) would
> be identical both across conditions and across subjects?
>
> If this isn't clear, what I mean is this: Normally, one would specify a
> fixed stimulus duration, say 5 seconds, followed by a fixed ISI (say, 2
> seconds) (or in a mixed design, a variable ISI). Assume that in Condition A,
> the average RT is 2 seconds, and in Condition B, the average RT is 4
> seconds. In the former condition, the ISI would effectively be 5 seconds,
> while in the latter, the ISI would effectively be 3 seconds. The modeling
> issue is that the boxcar over the entire block of trials would be a better
> model for Condition A then it would be for Condition B. Hence, differences
> might arise simply because of model fit rather than because of actual
> differences. The simple solution seems to be to define the ISI with respect
> to the offset of the previous stimulus (in this case, when the subject
> responds). Obviously, this would yield different block lengths for different
> conditions and different subjects. But is this an issue? My understanding is
> that in event-related designs with variable duration stimuli (e.g., when the
> trial duration is defined by subject response), the variable epoch model is
> in fact the best model (cf. Grinband et al., 2008). Would the same
> conclusion apply to a block design?
>
> Any thoughts would be much appreciated, thanks!
>
> Cheers,
> Bob
>
>
>
> -----------------------------------------------------------
> Bob Spunt
> Postdoctoral Scholar
> Emotion and Social Cognition Laboratory
> California Institute of Technology
|