Multiple comparison correction for 3dTcorr1D Spearman correlation

AFNI Message Board
1

Hi AFNI experts,
I used 3dTcorr1D to compute the whole-brain Spearman correlation between a neural index and a behavioral index. Here is my code:

3dTcorr1D -spearman \
                -prefix SubjExpmapnomap_BhvAcc_Spearman_${conditions[i]}_TR${TR_avg[$j]}_${Condtrts[$p]} \
                -mask /data/fmri/freesurfer/subjects/MNI152/SUMA/ribbon_mask.nii \
                SubjExpmapnomap_AIstimv17_stdGrayMatter_${conditions[i]}_TR${TR_avg[$j]}_${Condtrts[$p]}+tlrc \
                /data/fmri/locobjv2/results/avg_results/TwoExp_bhv_wholeandhalf/Expmapnomap_BhvAcc.txt

I obtained a map of correlation coefficients.

Could anyone tell me how to perform multiple comparison correction on these results?

Thanks!

2

An ad hoc solution is to convert the correlation value r to t-statistic using the formula t^2 = DF*r^2 / (1 - r^2) (where DF is the number of degrees of freedom), which lends to the following command line:

3dcalc -a r.value -expr '(ispositive(a)-isnegative(a))*sqrt(DF*a*a/(1-a*a))' -prefix t.value

Gang Chen

1 Like
3

Thanks!