The potentials of partial least squares structural equation modeling (PLS-SEM) in leisure research

Abstract Partial least squares structural equation modeling (PLS-SEM) is a multivariate statistical technique that helps examine complex relationships among a number of variables. Although its use has increased over decades, PLS-SEM remains underutilized in leisure research. The purpose of this methodological paper is to offer a primer on PLS-SEM for leisure researchers and to present a critical review of PLS-SEM’s strengths and limitations, while identifying potential applications of PLS-SEM across different sub-fields and theories in leisure research. Specifically, as to strengths, we discuss PLS-SEM’s sample size requirements, accommodation of formative and reflective measures, ability to model many variables and relationships, and statistical prediction capacity. In terms of its limitations, we review criticisms regarding PLS-SEM’s biased estimates as well as the lack of measurement error estimation and model fit assessment tools. Lastly, we provide recommendations for leisure researchers who wish to use PLS-SEM and journal editors and reviewers who assess PLS-SEM articles.