Rethinking Partial Least Squares Path Modeling: Breaking Chains and Forging Ahead

Rigdon (2012) argued that researchers should embrace PLS path modeling as an explicitly composite-based technique, and in connection with this proposal sketched a “concept-centric” approach to measurement as an alternative to the dominant “factor-centric” measurement paradigm. Bentler and Huang (2014) reassert the classical linkage between factor analysis and measurement—but this linkage depends on the implausible assumption that indeterminate factors are identical to the conceptual variable in researchers' theoretical models. Dijkstra (2014) correctly notes that assessing measurement validity within Rigdon's framework is difficult—but the ease of validity assessment within factor-centric measurement framework is an illusion. Sarstedt et al. (2014) endorse Rigdon's call to divorce PLS path modeling from factor analysis, and the current paper offers some further thoughts about the costs and challenges of embracing PLS path modeling as a composite-based technique.