| Literature DB >> 31099904 |
Abstract
In multiple testing scenarios, one of the more popular definitions of the error rate is the familywise error rate (FWER). The per-family error rate (PFER) is often not considered. The PFER deserves more attention than it currently receives. Both of these concepts were formulated by Tukey. We highlight some of the good and bad qualities of the FWER and the PFER. We find the FWER and PFER for four commonly used multiple comparison procedures.Keywords: Simes test; closed testing; linear programming; partitioning principle; simultaneous inference
Year: 2019 PMID: 31099904 DOI: 10.1002/sim.8190
Source DB: PubMed Journal: Stat Med ISSN: 0277-6715 Impact factor: 2.373