Gang Wrote:
-------------------------------------------------------
> It's still not clear to me whether you are
> considering Stimulus (with two levels, A and B) as
> an explanatory variable. In other words, are you
> interested in comparing the two stimuli, or do you
> just want to analyze each stimulus type
> separately?
>
Yes, we will want to analyze each stimulus type separately.
> Regarding modeling the variability of visit
> intervals, what kind of hypothesis are you
> envisioning? Suppose that, among the three
> consecutive visits for a subject, the time
> intervals (between visit1 and visit2, and between
> visit2 and visit3) are 4 and 7 days. First of
> all, what kind of impact of an interval on an
> upcoming visit? Linear? Secondly, how about
> visit1? Are you going to assign 0 days or
> something else (e.g., infinity) for it?
Well, we are hypothesizing that across the brain, there are roi's where you will see reliable activation across all 3 visits.
Initially, we were thinking this was a residual/within-subject variability issue.
But if we have duration between visits in days, couldn't we use duration as a covariate? The tricky part is coding the visit as you said. How would that work? Let's assume the impact of an interval is linear
Michael