Integrating Path Analysis and Kendall’s Tau-based Principal Component Analysis to Identify Determinants of Child Health

Abstract

This study develops a latent variable path analysis model using a Mixed-Scale Principal Component Analysis (PCA) approach based on Kendall’s Tau correlation to identify key determinants of child health in Batu City, Indonesia. Primary data were collected from 100 mothers with children under five years old through questionnaires. The variables examined include Family Demographics, Nutritional Consumption, and Child Health Condition, each measured using mixed-scale indicators (ordinal and numerical). Kendall’s Tau-based PCA was applied to reduce data dimensionality and construct latent variables, which were then integrated into a path analysis model. The results show that maternal age is the most dominant indicator in shaping the Family Demographics construct, while balanced nutritional food is the strongest indicator forming the Nutritional Consumption construct. Path analysis further reveals that Family Demographics significantly affect Child Health Condition both directly and indirectly through Nutritional Consumption, with a coefficient of determination of 77.62\%. These findings underscore the critical role of demographic and nutritional factors in determining child health outcomes and highlight the methodological advantage of Kendall’s Tau-based mixed-scale PCA for analyzing heterogeneous indicator data within a structural path framework.