Rank Procedures for Testing Linear Hypotheses in Repeated Measures Design

• Activity Number: 174
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Section on Nonparametric Statistics
• Author(s): Danush Wijekularathna*+ and Hossein Mansouri
• Companies: Texas Tech University and Texas Tech University
• Keywords: Dependent observations ; Rank tests ; Sub-hypothesis ; Estimating equation ; Wilcoxon rank statistic ; Linear regression

Abstract

Repeated measure data models arise when for each subject a vector of observations is taken. A unique feature of the repeated measures data is the correlation structure between observations. Often it is of interest to test hypotheses concerning the parameters of a linear model for such data. The parametric models for repeated measure data have been studied extensively in the literature. Several tests based on ranks are also available. In this presentation, we will formulate a class of rank tests for testing linear hypotheses for parameters of a linear model for repeated measures data and derive its asymptotic distribution. We will also present some results on small sample properties of the rank tests and compare them with their parametric competitors.