I would say yes, but I've seen plenty of pubs that do not correct. The bonferroni correction is fine. You could also try FDR (though on 4 points it might not make much sense).

On 01/10/2017 01:47 PM, neuroimage analyst wrote:

Hi,

I extracted thalamic, caudate, and putamen volume (left hemisphere), and cingulate thickness (right hemisphere) from freesurfer's parcellation. Then I compare these volumes and thickness across two groups using R and found significant differences in these measurements. Do I have to correct for multiple comparisons?

If yes, how do I do that? Does this make sense to divide the p-value/4 and if the p-value < p-value/4 then the test is Bonferroni corrected? If no, what do I say in my results that the test statistic is not inflated?

Thanks

Regards

--VM

---

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