A Comprehensive Partial Least Squares Approach to Component-Based Structural Equation Modeling
PLS Path Modeling (PLS-PM) is generally meant as a component-based approach to structural equation modeling that privileges a prediction oriented discovery process to the statistical testing of causal hypotheses. Nowadays, formative relationships are more and more used in real applications but pose a few problems for the statistical estimation and interpretation. It seems quite natural to introduce a PLS Regression (PLS-R) external estimation mode within the PLS-PM algorithm so as to overcome the mentioned problems, preserve the formative relationships and still remain coherent with the component-based and prediction-oriented nature of PLS-PM. Here, the main issues concerning the use of formative indicators in PLS-PM are investigated.
TRINCHERA, L. and ESPOSITO VINZI, V. (2008). A Comprehensive Partial Least Squares Approach to Component-Based Structural Equation Modeling. In: 32nd Annual Conference of the German Classification Society (GfKl).
Mots clés : #Multicollinearity