Key to Reporting PLS-SEM Results

Abstract

Structural Equation Modelling (SEM) is predicated on the establishment of a causal relationship between endogenous and exogenous latent variables by measuring them with the help of indicators. These causal relationships, established on the basis of extant literature, are represented by a proposed research model, and hypotheses are designed and tested simultaneously. This method is usually analysed using covariance-based (CB-SEM) and variance-based (PLS-SEM) approaches. In numerous academic studies, the dominant approach for fitting and hypothesis testing of research models describing the relationships between latent variables has been the use of LISREL and AMOS software for covariance-based structural equation modelling. In recent years, there has been increase in the number of papers in which SEM analyses are performed with the PLS-SEM approach compared to CB-SEM. The present study commences with a discussion of CB-SEM and PLS-SEM, followed by the provision of straightforward and pragmatic guidance on the presentation of results in the context of PLS-SEM analysis, which constitutes the primary focus of the study. The paper initiates with an examination of the rationale behind the utilisation of PLS-SEM, subsequently offering a concise overview of the fundamental principles for the reporting of PLS-SEM outcomes. The paper also provides a brief overview of the basic concepts of sample size, assumptions, model predictive power, Importance-Performance Map (IPMA) and moderating in PLS-SEM.