The Student-Newman-Kuels procedure is a well-known step-down multiple comparisons procedure with critical values based on the Studentized range distribution. The False Discovery Rate is a Type I error rate for multiple comparisons, intermediate in stringency to the weak familywise error rate (experimentwise error rate) and the strong familywise error rate. We show that SNK controls FDR.
Copyright 2020 Elsevier B.V., All rights reserved.
- Multiple comparisons
- Simultaneous inference