History of AFNI updates  

|
September 29, 2014 10:20AM
> I am not sure which p-value or any threshold value is appropriate, because the correlation
> map changes with different threshold value (qutorange/manual, p-value, q-value, etc.)
> To show the results of correlation map, I wonder which threshold should be used.

The p-value you see underneath the slider bar is the voxel-wise probability. You can choose any threshold or p-value you feel comfortable with for general visualization. If you want to consider to control for multiple testing errors, use either FDR q-value or FWE corrected value for thrsholding. To obtain the minimum cluster size, run 3dClustSim with a mask and the FWHM values.

> If I apply FDR curve menu, is correlation map affected? When I choose FDR curve, q-value
> appears, but there is no change in the correlation map.

When you threshold the activation patterns with a q-value of, for example, 0.05, there is no change in the correlation map? No change compared to what?

Gang
Subject Author Posted

P-value in AFNI

Seul September 25, 2014 10:56PM

Re: P-value in AFNI

gang September 26, 2014 03:19PM

Re: P-value in AFNI

Seul September 26, 2014 04:46PM

Re: P-value in AFNI

gang September 26, 2014 05:10PM

Re: P-value in AFNI

Seul September 28, 2014 03:40PM

Re: P-value in AFNI

gang September 29, 2014 10:20AM

Re: P-value in AFNI

Seul September 29, 2014 03:50PM

Re: P-value in AFNI

gang September 30, 2014 10:28AM