A new coefficient to measure agreement between continuous variables

TL;DR

This paper introduces rho1, a new robust coefficient based on L1 distances for measuring agreement between continuous variables, applicable to various distributions and outperforming existing methods in clinical and spatial studies.

Contribution

The paper presents a novel agreement coefficient, rho1, derived for elliptical distributions, with theoretical properties and validation, offering a robust alternative to existing measures.

Findings

Rho1 is robust to outliers due to its L1 basis.

Rho1 is applicable to normal and elliptically contoured distributions.

Numerical experiments confirm rho1's effectiveness in agreement assessment.

Abstract

Assessing agreement between two instruments is crucial in clinical studies to evaluate the similarity between two methods measuring the same subjects. This paper introduces a novel coefficient, termed rho1, to measure agreement between continuous variables, focusing on scenarios where two instruments measure experimental units in a study. Unlike existing coefficients, rho1 is based on L1 distances, making it robust to outliers and not relying on nuisance parameters. The coefficient is derived for bivariate normal and elliptically contoured distributions, showcasing its versatility. In the case of normal distributions, rho1 is linked to Lin's coefficient, providing a useful alternative. The paper includes theoretical properties, an inference framework, and numerical experiments to validate the performance of rho1. This novel coefficient presents a valuable tool for researchers assessing agreement between continuous variables in various fields, including clinical studies and spatial analysis.