Hi Livia,

I am not very familiar with the paired analysis (never used it and did not develop it). From a glance (others may know better) your files look OK, the first creates the paired differences, the second is for the group comparison to test if the group mean of the paired differences differs across groups. So this would be fine, and since you have

- exactly 2 time points per subject

- equally spaced in time

you should be fine with this analysis. You can of course try LME but I would not expect too much from it. Should be nearly identical, still it is a sanity check.

Best, Martin

On 05/21/2017 05:09 PM, Livia Liu wrote:

Dear Martin， Thank you for your reply. I used a paired analysis to do "two groups two times data of the longitudinal analysis".

But unfortunately, I did not make the results, and I use “mri_glmfit-sim” to find 0 vertices in mask.

My files are as follows.

Fsgd 1: (pairs.fsgd)

GroupDescriptorFile 1

Class Sub

Class Con

Input sub01_MR1 Sub

Input sub01_MR2 Sub

Input con01_MR1 Con

Input con01_MR2 Con

...

Fsgd 2: (paired-diff.fsgd)

GroupDescriptorFile 1

Class Sub

Class Con

Input sub01pair Sub

Input con01pair Con

...

mean.mtx: 1 -1

Are they correct in my Paired analysis? And I am trying LME now.

However, I see other people in the mailing list of the questions and replys, I realize that I may be wrong.

My two groups of two time are the experimental group (bs) and the placebo group (placebo).

I want to do “the crossover analysis” Just like:(postbs-prebs) - (postplacebo-preplacebo)

Should not I use longitudinal analysis for these two groups? What kind of analysis do I need to do? Please tell me where I should look in the wiki. Looking forward to reply, and thanks very much. Kind regards, Livia

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer