PLS FAC-SEM: an illustrated step-by-step guideline to obtain a unique insight in factorial data

Abstract

Purpose The purpose of this paper is to provide an illustrated step-by-step guideline of the partial least squares factorial structural equation modeling (PLS FAC-SEM) approach. This approach allows researchers to assess whether and how model relationships vary as a function of an underlying factorial design, both in terms of the design factors in isolation (i.e. main effects) as well as their joint impact (i.e. interaction effects). Design/methodology/approach After an introduction of its building blocks as well as a comparison with related methods (i.e. n-way analysis of variance (ANOVA) and multi-group analysis (MGA)), a step-by-step guideline of the PLS FAC-SEM approach is presented. Each of the steps involved in the PLS FAC-SEM approach is illustrated using data from a customer value study. Findings On a methodological level, the key result of this research is the presentation of a generally applicable step-by-step guideline of the PLS FAC-SEM approach. On a context-specific level, the findings demonstrate how the predictive ability of several key customer value measurement methods depends on the type of offering (feel-think), the level of customer involvement (low-high), and their interaction (feel-think offerings×low-high involvement). Originality/value This is a first attempt to apply the factorial structural equation models (FAC-SEM) approach in a PLS-SEM context. Consistent with the general differences between PLS-SEM and covariance-based structural equation modeling (CB-SEM), the FAC-SEM approach, which was originally developed for CB-SEM, therefore becomes available for a larger amount of and different types of research situations.