The application of inner weight PCA in SEM-PLS for stunting model development

This study aims to determine the factor score in the stunting model using the Structural Equation Modeling-Partial Least Squares (SEM-PLS) approach with the inner weight Principal Component Analysis (PCA) scheme. The main objective of this research is to optimise factor estimation in complex stunting models, taking into account the high intercorrelation between indicator variables. The method used in this research is SEM-PLS with inner weight PCA scheme. The data used is empirical data on malnutrition status obtained from the Southeast Sulawesi Health Office. This study shows that the use of the inner weight PCA scheme in SEM-PLS is able to produce a more stable and accurate factor score in the stunting model compared to other more conventional schemes. This finding suggests that the integration between SEM-PLS and PCA can improve the validity and reliability of stunting models. The Q-square results of the Path weighting scheme are better than the factor and PCA schemes. However, there are some limitations in the application of this method, such as the need for a deep understanding of PCA techniques and complexity in the interpretation of results. The originality of this study lies in the new approach in determining factor scores in stunting models, which can be a reference for other researchers in the field of public health and nutrition.