You need much more than 5 iterations! If you are doing whole hemisphere analysis, then you can use the precomputed tables.
doug
On 09/24/2013 01:46 PM, ye tian wrote:
Dear Freesurfers, Sorry for having posted an incorrect question. I basically ran *mri_glmfit-sim *on both hemispheres and tried both *--sim-sign neg *and *--sim-sign pos*. Although I only had 5 iterations, I found no significant clusters on any case. Does that mean that I accept the null hypothesis? PS: I found in http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg23324.html that 10,000 iterations are commonly used. Thank you very much!
On Mon, Sep 23, 2013 at 9:47 PM, ye tian <tianye730@gmail.com mailto:tianye730@gmail.com> wrote:
Dear Freesurfers, I used the following command line to find significant clusters after glm analysis (before MC clusters attached) mri_glmfit-sim \ --glmdir lh.meditation_age.glmdir \ --sim mc-z 5 4 mc-z.negative \ --sim-sign neg --cwpvalthresh 0.05\ --overwrite I didn't find any, i.e., there are no clusters listed in file mc-z.negative.sig.cluster.summary. Does this mean that I should accept the null hypothesis for the whole cortex? Also, how many iterations of mc do people typically perform? I guess it doesn't make a difference if I do 1000 iterations instead of 5 anyways. Thank you very much! Sincerely, Ye
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer