Partial least squares structural equation modeling (PLS-SEM)

Purpose – The authors aim to present partial least squares (PLS) as an evolving approach to structural equation modeling (SEM), highlight its advantages and limitations and provide an overview of recent research on the method across various fields. Design/methodology/approach – In this review article, the authors merge literatures from the marketing, management, and management information systems fields to present the state-of-the art of PLS-SEM research. Furthermore, the authors meta-analyze recent review studies to shed light on popular reasons for PLS-SEM usage. Findings – PLS-SEM has experienced increasing dissemination in a variety of fields in recent years with nonnormal data, small sample sizes and the use of formative indicators being the most prominent reasons for its application. Recent methodological research has extended PLS-SEM's methodological toolbox to accommodate more complex model structures or handle data inadequacies such as heterogeneity. Research limitations/implications – While research on the PLS-SEM method has gained momentum during the last decade, there are ample research opportunities on subjects such as mediation or multigroup analysis, which warrant further attention. Originality/value – This article provides an introduction to PLS-SEM for researchers that have not yet been exposed to the method. The article is the first to meta-analyze reasons for PLS-SEM usage across the marketing, management, and management information systems fields. The cross-disciplinary review of recent research on the PLS-SEM method also makes this article useful for researchers interested in advanced concepts.