The aim of this dissertation is essentially twofold: (1) to identify and correct some misunderstandings regarding Type I error control that are common in the field of psychology, and (2) to compare via simulation different multiple-testing procedures that can be used in a few specific experimental designs.