On Small Sample Properties of Simultaneous Inference Based on Rank Statistics

• Activity Number: 90
• Type: Contributed
• Date/Time: Sunday, August 4, 2013 : 4:00 PM to 5:50 PM
• Sponsor: Section on Nonparametric Statistics
• Author(s): Hossein Mansouri*+
• Companies: Texas Tech University
• Keywords: Multiple Comparisons ; Rank Estimation ; Ranking After Alignment ; Familywise Error Rate

Abstract

An extensive amount of literature exists on pairwise comparisons based on rank statistics for the one-way layouts. Extensions are also made to factorial designs. These Techniques are developed to provide robust alternatives to the classical methods. However, simultaneous inference based on ranks for linear functions of the parameters of a linear model has not received much attention in the literature. In this presentation, we discuss simultaneous inference for the parameters of a linear model based on rank statistics. We establish the asymptotic properties of the simultaneous procedures based on ranks. Using Monte Carlo simulation, we study the small sample validity and robustness of the rank-based simultaneous inference.