Hypothesis Testing of Population Percentiles via the Wald Test with Bootstrap Variance Estimates
Year: 2016
Type: article
Abstract: Testing the equality of percentiles (quantiles) between populations is an effective method for robust, nonparametric comparison, especially when the distributions are asymmetric or irregularly shaped. Unlike global nonparametric tests for homogeneity such as the Kolmogorv-Smirnov test, testing the equality of a set of percentiles (i.e., a percentile profile) yields an estimate of the location and extent of the differences between the populations along the entire domain. The Wald test using bootstrap estimates of variance of the order statistics provides a unified method for hypothesis testing of functions of the population percentiles. Simulation studies are conducted to show performance of the method under various scenarios and to give suggestions on its use. Several examples are given to illustrate some useful applications to real data.
Source: Open Journal of Statistics
Authors: William D. Johnson, Jacob E. Romer
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W2267634951