The Terms "Multivariate" and "Multivariable" Are Used Incorrectly and Interchangeably in Orthopaedic Publications: Should We Care About the Distinction?

Abstract: Orthopaedic surgery research increasingly utilizes statistical models to adjust for confounding, provide additional precision, and describe complex relationships. Traditionally, a statistical model is termed "multivariable" if it examines the relationship between multiple independent variables (e.g., covariables). The term "multivariate" is used to describe statistical models with multiple dependent variables (e.g., outcomes). The precise use of statistical terminology is important for study appraisal and reproducibility. The term "multivariate" is frequently used incorrectly in the orthopaedic literature to describe statistical models with a single dependent variable. In our analysis, we found that the term "multivariate" was used frequently, and in >90% of cases, the term was used to describe models that contain only a single dependent variable. Consistent with our perception, the terms "multivariable" and "multivariate" are not used with optimal precision. Below, we discuss the impact of the misuse of statistical terminology and present suggestions for improving statistical reporting.
Competing Interests: Disclosure: The Disclosure of Potential Conflicts of Interest forms are provided with the online version of the article (http://links.lww.com/JBJS/H378).
(Copyright © 2022 by The Journal of Bone and Joint Surgery, Incorporated.)