Title: Letting go of multiplicity control: don't worry, you never liked it anyway

Abstract: Textbooks, editorial policies, and hundreds of articles have warned us of the dire consequences of not controlling for Type I errors when conducting multiple tests of statistical significance (dubbed multiplicity control). Researchers have proposed dozens of intricate methods for controlling a variety of different error rates (e.g., familywise error rate, false discovery rate), each purporting to control the risk of a false positive at an acceptable level. However, the practice of multiplicity control has weak theoretical support, a point that has been ignored for more than 25 years. In this talk I summarize the case against multiplicity control, including how the replication crisis might have put the final nail in the multiplicity control coffin. But sleep easy ... who ever enjoyed carving up their α anyway?