Five Common Mistakes for Using Partial Least Squares Path Modeling (PLS-PM) in Management Research
Asyraf Afthanorhan, Zainudin Awang, Nazim Aimran
Abstract
The value of Partial Least Squares Path Modeling (PLS-PM) in management research has now been acknowledged, although the PLS-PM was developed for a reason. First, the PLS-PM was developed as an alternative to Covariance based Structural Equation Modeling (CBSEM) when exploratory research is conducted. As far as this method concerned, many researchers are misused or overuse the application of PLS-PM without understanding the basic knowledge in structural equation modeling. Thus, the purpose of this paper is to discuss the five common mistakes (data distributions, sample size limitations, unsatisfactory fitness index, misunderstanding between confirmatory and exploratory research, and poor factor loadings) for using PLS-PM over CB-SEM in management research. We concluded that the researchers should respect these methods and justify their use when conducting the research projects because some of the projects might be better for CB-SEM or PLS-PM.