Hi Jon ,

I think you should use a Hotelling ´s T2 statistic at each parcellation unit to account for the two variables you are using. Then you need to correct for the # of parcels. Alternatively, you may use False discovery rate correction. It is less stringent than Bonferroni correction.

Cheers Jorge Luis

--- El vie, 20/11/09, Jonathan DuBois jonathan.m.dubois@gmail.com escribió:

De: Jonathan DuBois jonathan.m.dubois@gmail.com Asunto: [Freesurfer] Bonferroni multiple comparison correction Para: freesurfer@nmr.mgh.harvard.edu Fecha: viernes, 20 de noviembre, 2009 18:37 When correcting for multiple comparisons with Bonferroni for FS parcilation statistics, is it necessary to adjust for all parcellations (35) or just the variables (in our case mean thickness and surface area)? Assuming that no prior hypothesis is used to constrain the number of parcellations.

Thanks for you input!

Jon

-----Adjunto en línea a continuación-----

---

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