Range statistics and equivalence tests

Abstract

This paper investigates range statistics and its application on the hypothesis testing of equivalence of mean of multiple systems. We develop an equivalence-testing procedure that allows unequal variances among systems and discuss how to estimate the required critical constant, that is, quantiles of a range. A table of critical constants is provided for implementing the procedure. An experimental performance evaluation demonstrates the validity and efficiency of the procedure.

Keywords:

• range statistics
• equivalence testing
• sample size allocation
• quantile estimation