Within-Subject Comparison (Paired T-Test)

Submitted by skuehn on Mon, 11/01/2010 - 01:53

Hi,

I am planning a study in which I want to measure two periods of resting state for each participants and want to compare them by means of paired-t-tests.
There are two ways of doing this: (1) stopping the scanner after the first resting period and restarting it for the second or (2) not stopping and restarting the scanner.
In task related fMRI one would not want to compare conditions that come from different scanning runs. Does the same apply for resting state? If so I assume that option (2) is better.
What would be your suggestion?

Best wishes,
Simone

Submitted by YAN Chao-Gan on Mon, 11/01/2010 - 19:20

For the option (1), how to define stopping the scanner? Turn off the entire machine?
In my opinion, you can have two resting-state fMRI scanning sessions as you usually doing in task fMRI studies. Turn off the machine is not necessary, but you need to stop the first scanning session and turn to the next scanning session.
Best wishes!

Submitted by ZangYF on Sat, 11/06/2010 - 17:37

I think you can also get one long session (say, 20 minutes). Then divide it into two periods and compare between them. It depends on what specific question you want to answer.

Submitted by skuehn on Sat, 11/06/2010 - 21:29

Thanks for your reply. My aim is to compare the two resting state sessions with one another (within-subject).
Say e.g. in the first part participants are instructed to think about a sad event (for 10 min) in the second part they are instructed to think about a happy event (for 10 min). My question is what is technically (concerning resting state analysis) the best solution. Use two fMRI runs each 10 min, or 1 run of 20 min with a short cue that informs the subject to switch.

Best
Simone

Submitted by ZangYF on Sun, 11/07/2010 - 08:37

In this particular case, I would suggest two separate and counterbalanced runs (sessions). You may also use one long session with the order of affective conditions counterbalanced. But the transitional time points (say, 20 s) should be discarded becaused this period of time is a mixed effect of the two affective conditions.