Lars Tjelta Westlye wrote:

Hi,

I have a couple of questions regarding mri_glmfit-sim (mc-z).

1. Is it correct that --sim-sign abs builds up a null distribution of max

cluster sizes regardless of the direction of the effects in the simulated data while --sim-sign pos/neg only counts the largest positive/negative cluster for every iteration? Or is the sign only used in the subsequent correction procedure by mri_surfcluster and not affecting the null distribution?

It does affect the null.

2. Under which circumstances would it be advisable to choose --sim-sign

abs instead of pos/neg and vice versa?

When you want to do an unsigned test. It is a little odd because it can pool voxels that are both pos and neg into one cluster, but this is the same as any unsigned test (eg, F).

doug

Thanks a lot, Lars

---

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