Gang Wrote:
-------------------------------------------------------
> Michael,
>
> In addition to the information of three
> measurements per subjects, could you provide more
> information? How many tasks or conditions? How
> many groups? More importantly, what tests or
> hypotheses are you trying to run?

Gang, thanks for your response.

For a given subject=

Group = 1 group, 15 subjects, 1 scan for each of the 3 visits

Scan, Visit1 = Block design, two stimulus types. Stimulus A = presented 15 times, Stimulus = B presented 15 times.

Scan, Visit2 = Block design, two stimulus types. Stimulus A = presented 15 times, Stimulus = B presented 15 times.

Scan, Visit3 = Block design, two stimulus types. Stimulus A = presented 15 times, Stimulus = B presented 15 times.

We want to test for reliability of spatial extent and the BOLD response across all 3 visits, for stimulus A and B, respectively. However, we are concerned with how the varied time between visits per subject may interfere with the reliability result. Would it inflate the intersession reliability? Is the reliability result independent of or dependent on the duration between visits? Could subject level variables be accounted for as covariates in this analysis? If it is dependent, how do we account for it? If it is independent of duration between visit, we should see the same activation maps with and without accounting for variability of time per subject?



Michael