Tips to use partial least squares structural equation modelling (PLS-SEM) in knowledge management

TLDR

Structural equation modelling combines latent variables and structural relationships, and the partial least squares approach is widely used to estimate complex cause‑effect models with latent variables across disciplines such as knowledge management. This paper examines how PLS‑SEM has been applied in KM research and offers new guidelines to improve PLS‑SEM reporting. The study conducted a systematic literature review of 63 KM journal publications from 2015‑2017 to objectively analyze PLS‑SEM usage. The review shows that PLS‑SEM is widely used in KM but that researchers commonly hold misconceptions about its purpose, data, model characteristics, and evaluation, underscoring the need for clearer reporting guidelines.

Abstract

Purpose Structural equation modelling (SEM) has been defined as the combination of latent variables and structural relationships. The partial least squares SEM (PLS-SEM) is used to estimate complex cause-effect relationship models with latent variables as the most salient research methods across a variety of disciplines, including knowledge management (KM). Following the path initiated by different domains in business research, this paper aims to examine how PLS-SEM has been applied in KM research, also providing some new guidelines how to improve PLS-SEM report analysis. Design/methodology/approach To ensure an objective way to analyse relevant works in the field of KM, this study conducted a systematic literature review of 63 publications in three SSCI-indexed and specific KM journals between 2015 and 2017. Findings Our results show that over the past three years, a significant amount of KM works has empirically used PLS-SEM. The findings also suggest that in light of recent developments of PLS-SEM reporting, some common misconceptions among KM researchers occurred mainly related to the reasons for using PLS-SEM, the purposes of PLS-SEM analysis, data characteristics, model characteristics and the evaluation of the structural models. Originality/value This study contributes to that vast KM literature by documenting the PLS-SEM-related problems and misconceptions. Therefore, it will shed light for better reports in PLS-SEM studies in the KM field.