Combination of partial least square structural equation modeling scheme of principal component analysis with...

Area of Science:

• Social Sciences
• Public Health
• Statistics

Background:

• Structural Equation Modeling (SEM) has strict assumptions limiting its use.
• Partial Least Squares SEM (PLS-SEM) offers flexibility but faces challenges with weighting schemes.
• Improving PLS-SEM estimation methods is crucial for robust causal inference.

Purpose of the Study:

• Introduce a novel PCA-based weighting scheme for PLS-SEM.
• Integrate Importance-Performance Analysis (IPA) to identify intervention priorities.
• Enhance the accuracy and stability of PLS estimation.

Main Methods:

• Developed a Principal Component Analysis (PCA)-based weighting scheme for PLS-SEM.
• Applied Importance-Performance Analysis (IPA) to identify high-impact, underperforming indicators.
• Utilized PLS-SEM to analyze child malnutrition in East Java.

Main Results:

• The PCA-based weighting scheme improved PLS estimation stability and accuracy.
• Socio-economic conditions were identified as major drivers of food security, parenting, and health-environment services.
• Exclusive breastfeeding emerged as a critical area for intervention based on IPA.

Conclusions:

• The proposed methodological approach strengthens PLS-SEM.
• The study provides actionable insights for child malnutrition policy in East Java.
• Combining PCA-based weighting and IPA offers a powerful tool for applied research.