Two new classes of nonparametric tests for scale parameters

Abstract

Testing of homogeneity of populations is a very useful phenomenon in several fields. After knowing that the two populations have the same variability or not, public/private sector organizations can execute their plans accordingly. In this paper, two new classes of nonparametric tests are proposed to test whether the scale parameters of two populations are the same or not. For each class of tests, we determine which member of the class provides the maximum efficacy; and compare its performance with some of the existing tests. Monte Carlo simulation study is carried to see the performance of both tests in terms of power. An illustrative example is also provided to see the implementation of the proposed tests.

KEYWORDS:

• Nonparametric tests
• scale parameters
• simulation study
• illustrative example

Mathematics subject classifications:

• 62E20
• 62G10
• 62G20
• 62P10

Acknowledgments

The authors thank the anonymous referees for their valuable suggestions, which lead to an improvement of the paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).