Exploring the CHAID Method in Detecting the Strength of Relationships among Factors Associated with the Prevalence of Undernourishment Rahmat Trianto1, Dian Cahyawati1*, Eka Susanti1, Alfensi Faruk1
Universitas Sriwijaya, Faculty of Mathematics and Natural Sciences, Department of Mathematics
Abstract
Data from Statistic Indonesia (BPS) in 2024 show disparities in the Prevalence of Undernourishment (PoU) across regions in Indonesia. These disparities indicate that there are still challenges and obstacles in achieving the second goal of the Sustainable Development Goals (SDGs), namely Zero Hunger. This issue highlights the need for further studies to identify the factors influencing variations in PoU levels in Indonesia. This study aims to explore the factors that are strongly associated with PoU levels using the Chi-Square Automatic Interaction Detection (CHAID) method. The secondary data analyzed were obtained from the official websites of Statistics Indonesia (BPS) and National Food Agency (NFA). The observed variables include rice production, population density, percentage of poor population, percentage of per capita food expenditure, food consumption pattern score, Human Development Index , and Food Security Index across regions in Indonesia. The CHAID analysis, presented through a dendrogram, produced 23 nodes, consisting of 14 terminal nodes representing the final classification of PoU levels in Indonesia. The results show that five factors are closely related to PoU levels. The most influential factor associated with PoU is the Food Security Index, followed by Percentage of Poor Population, Population Density, Human Development Index, and Food Consumption Pattern score. These findings support the reference that the CHAID method can be used as an exploratory tool to describe the structure of relationships among variables, starting from the strongest to the weakest associations related to undernourishment. The results are expected to provide additional insights for policymakers in formulating national food security policies that align with the second goal of the SDGs.
Keywords: Chi-Square, Data Exploration, CHAID Dendrogram, Prevalence of Undernourishment.
