ABSTRACT
So far we have studied general linear models (GLM) that are constructed at each voxel. In this chapter, we explore the multiple comparisons issue that is necessary to properly threshold statistical maps for the whole image. The multiple comparisons are crucial in determining overall statistical significance in correlated test statistics over the whole brain. In practice, t-or F -statistics in adjacent voxels are correlated. So there is the problem of multiple comparisons, which we have simply neglected up to now. For multiple comparisons that account for spatially correlated test statistics, various methods are proposed: Bonferroni correction, random field theory [388, 394], false discovery rates [35, 36, 134] and permutation tests [261].
