A transdisciplinary view of measurement error models and the variations of X = T plus E

摘要

The purpose of this paper is to formally outline a sequence of propositions that describe the connections between five linearly additive measurement error models commonly used in disciplines from psychometrics and test theory to economics to epidemiology, and one new model formerly proposed in Kroc & Zumbo (2018). We show that although these models are deceptively similar in their general algebraic form, X = T + E, they have different error structures that both connect and distinguish them. (c) 2020 Elsevier Inc. All rights reserved.