Composite-based and covariance-based structural equations modeling: moving forward by changing the dialogue

Abstract

Purpose

The purpose of this paper is to integrate the findings of articles appearing in European Journal of Marketing’s special section on covariance-based versus composite-based structural equations modeling (SEM).

Design/methodology/approach

This is an editorial which uses literature review to draw conclusions regarding areas of agreement, areas for further research, and changing the discussion around composite-based SEM methods.

Findings

There are now four new areas of agreement regarding composite-based SEM. Researchers should adopt a toolbox approach to their methods and know the strengths and weaknesses of the research tools in their toolbox. Partial least squares (PLS) SEM and covariance-based SEM are not substitutes, and it is inappropriate to use the language of confirmatory factor analysis (CFA) in reporting measurement estimates from PLS SEM. Measurement matters and researchers need to devote effort to using reliable and valid multi-item measures in their investigations.

Originality/value

This postscript article outlines recommendations for authors, reviewers and editors regarding the analysis of data and reporting of results using structural equations models.

Citation

Voss, K.E. (2023), "Composite-based and covariance-based structural equations modeling: moving forward by changing the dialogue", European Journal of Marketing, Vol. 57 No. 6, pp. 1780-1792. https://doi.org/10.1108/EJM-02-2022-0137