I was trying to play with ETAC today. I ran a two group 3dTtest++ with -ETAC on seed-based resting state data. I had done the preprocessing with a 6mm blur so I didn’t use the -ETAC_blur option.
For the results output, I am unclear if I should set the threshold bar to 0.05 or if I should set the threshold bar to a lower value, like 0.002.
For example, how should I interpret a “significant” cluster at per-voxel p= 0.05 with a voxel size of 10, with 3.5 mm isotropic voxels?