An Innovative Application of Composite-Based Structural Equation Modeling in Hospitality Research With Empirical Example

Abstract

Partial least squares path modeling (PLS-PM) and generalized structured component analysis (GSCA) are two key estimators derived from a full-fledged composite-based structural equation modeling (SEM). The analyses of PLS-PM and GSCA have been recently extended to mimic factor-based SEM, and the extended approaches are called PLS_C and GSCA_M, respectively. Simulation studies have confirmed that the relative performance of PLS-PM is comparable with that of GSCA. Similarly, GSCA_M, PLS_C, and the traditional factor-based SEM perform equally well in parameter recovery. Although composite-based SEM perfectly fits into the current research landscape that focuses on a prediction-oriented approach, empirical research in the hospitality context that uses PLS-PM, GSCA, PLS_C, and GSCA_M estimators is extremely rare. To encourage hospitality researchers to adopt these methodologies, we demonstrate an illustrative example using PLS-PM, GSCA, PLS_C, and GSCA_M based on the confirmatory composite analysis (CCA) procedure. Measurement and structural invariances, applications of model fit, PLS_predict, and importance-performance map analysis are incorporated into our example. Finally, practical management in the hospitality field based on this methodology is discussed.