Can it be appropriate to apply an AlphaSim-estimated cluster threshold to whole brain voxelwise group analyses (an F test in a 3-level ANOVA specifically) that were calculated in programs more devoted to random field theory approach to thresholding? My thought was that if the Gaussian filter one tells AlphaSim to apply to its simulations along x,y, and z dimensions with the -fwhmx etc options match the estimated smoothness of the data in the group analysis, this might be OK.