Reliability assessment and approaches to determining agreement between measurements: Classic methods paper

Abstract 

This classic methods paper (Bland and Altman, 2010) considers the assessment of agreement between measures, an often overlooked aspect of assessing measurements taken for use in research and practice and (re) introduces the ubiquitous ‘Bland Altman’ procedures for assessing agreement. The importance of these procedures is high and they address issues that are not always considered in research which uses measurement scales or describes the characteristics of scales developed for use in clinical practice. Many widely used approaches for reliability assessment can fail to consider the agreement between measures at all and can give an entirely misleading impression of an instrument's suitability for use in research or practice.

Keywords: Reliability, Test–retest reliability, Inter-rater reliability, Measurement, Reproducibility of results, Validation studies, Observer variation