Permutation-Based Tests of Perfect Ranking

TL;DR

This paper enhances three perfect ranking tests in ranked set sampling using permutation methods, increasing their power and generalizing existing tests across different cycles.

Contribution

It introduces permutation-based extensions to existing perfect ranking tests, improving their effectiveness and unifying different approaches.

Findings

Permutation approach increases test power.

Two tests are equivalent to existing literature tests.

Generalizes tests across different cycles.

Abstract

We improve three tests of perfect ranking in ranked set sampling proposed by Li and Balakrishnan (2008) using a permutation approach. This simple way of extending all three concepts to comparisons across different cycles increases the power. Two of the proposed tests are equivalent to tests from the literature, which were derived differently and are therefore generalized by the permutation-based tests.