Some Remarks on Nonlinear Relationships in Pls Path Modeling

Structural Equation Models (SEMs) are widely used to model complex causal relationships, especially in economic and social domains. In this framework, PLS Path Modeling algorithm is a widely used technique. However, PLS Path Modeling assumes linear relationships among the latent concepts, and between the observed variables and the corresponding latent variables. Recent studies in marketing research, have pointed out that this hypothesis may appear to be too restrictive. In this paper nonlinear approaches to PLS Path Modeling are discussed and an application to real data concerning the impact analysis of the European Union enlargement to 25 countries on Italian firms is presented.