0.09940 and 0.10338 seem pretty close to me:). The summary file has Hi and Low which is the 95% confidence interval assuming a binomial model at the mean prob
On 9/9/2020 3:24 PM, freesurferlist@gmx.com wrote:
External Email - Use CautionThanks, I was looking for the higher resolution as, from my understanding, the p-value of a cluster is usually reported as “p =“ rather than “p <“ (but this may be not worth the extra computational time for cluster 1).
Just as a follow up question, what about cluster 2 for cwp0.1 and cwp0.999. This has a CWP of 0.09940 and 0.10338 respectively. Should we be expecting convergence by 10,000 iterations for this cluster?
On 10/9/20 at 12:18 am, Douglas N. Greve wrote:
The issue is that the chance of this event happening by chance is very low, around 2-14 times out of 10,000. You are getting this variation because you are only using 10000 iterations. But why would you need higher resolution to your p-value? It will definitely be below 0.05 needed for publication.
On 9/8/2020 8:45 PM, Paul wrote:
External Email - Use Caution
Dear Douglas, Thanks for looking in this. Please see files attached. Also, I have used Freesurfer 5.1 for all the processing (i.e recon-all and editing etc) and have used Freesurfer 6.0 for the statistical analysis (i.e. mris_preproc, mri_surf2surf, mri_glmfit) and the patch version of mri_glmfit-sim (i.e ftp://surfer.nmr.mgh.harvard.edu/pub/dist/freesurfer/6.0.0-patch/mri_glmfit-sim). ftp://surfer.nmr.mgh.harvard.edu/pub/dist/freesurfer/6.0.0-patch/mri_glmfit-sim Kind regards, Paul *Sent:* Wednesday, September 09, 2020 at 1:01 AM *From:* "Douglas N. Greve" dgreve@mgh.harvard.edu *To:* freesurfer@nmr.mgh.harvard.edu *Subject:* Re: [Freesurfer] mri_glmfit-sim using --cwp 0.05, --cwp 0.1, --cwp 0.999 On 9/7/2020 10:52 PM, Paul wrote:
External Email - Use Caution Dear Freesurfer Experts, I have few questions regarding CWP values from mri_glmfit-sim using --cwp 0.05, --cwp 0.1, --cwp 0.999. Using the 3 three commands below (A, B and C): A. mri_glmfit-sim --glmdir lh_status.glmdir --perm 10000 1.3 abs --cwp 0.05 --2spaces --bg 1 B. mri_glmfit-sim --glmdir lh_status.glmdir --perm 10000 1.3 abs --cwp 0.1 --2spaces --bg 1 C. mri_glmfit-sim --glmdir lh_status.glmdir --perm 10000 1.3 abs --cwp 0.999 --2spaces --bg 1 I end up with the same significant ClusterNo 1. (in terms of Max, VtxMax, Size(mm^2), MNIX, MNIY, MNIZ, NVtxs, WghtVtx, Annot) However, the CWP is 0.00040, 0.00140, and 0.00020 for A, B and C respetively. I was expected after 10,000 iterations, the CWP would be more closely matched or even convergent. Could you please let me know if this result is typical after 10,000 iterations? Also, should I be using more iterations (say 20k or more 100k?) to reach convergence across A, B, and C or is it usual to just report A, B or C?I don't understand. All you are changing is the CWP threshold? The final CWP value for a cluster should not be changing at all. If this is the case, can you send me the summary file for each?
Kind regards, Paul _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Mass General Brigham Compliance HelpLine at http://www.massgeneralbrigham.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
freesurfer@nmr.mgh.harvard.edu
-
Douglas N. Greve