Determination of characteristics of stunting risk families in Subulussalam, Aceh in 2022 using k-modes cluster analysis

¹ The Regional Development Planning Agency, Subulussalam City, 24781 Subulussalam, Indonesia
² Statistics Department, Universitas Syiah Kuala, 23111 Banda Aceh, Indonesia
³ Mathematics Department, Universitas Syiah Kuala, 23111 Banda Aceh, Indonesia

^* Corresponding author: marzuki@usk.ac.id

Abstract

Stunting is a disruption in the growth and development of children caused by nutritional deficiencies and recurrent infections. Stunting can lead to a decline in intelligence levels and lower educational achievements. The aim of this research is to determine the characteristics of families at risk of stunting in the Subulussalam City. This study utilizes data from the Family Census conducted by the National Family Planning Coordination Board (BKKBN) in 2021. The data consists of 8,954 households at risk of stunting, using 10 variables for stunting risk calculation, such as households with children under two years old, under five years old, households with a fertile-aged wife, households without access to safe drinking water, and others. This study employs the k-modes cluster analysis method, which involves grouping a dataset based on specific variables into k clusters. The k-modes algorithm randomly selects k objects as cluster centers and calculates the distance of each object to the cluster modes. Households at risk of stunting in Subulussalam City mostly exhibit characteristics of having a fertile-aged wife and having too many children. Most households at risk of stunting already have access to safe drinking water and proper toilets. Clustering using k-modes resulted in an optimal k model with k =2 (SW/SB ratio) and k =5 (R-Squared). The model with k =2 revealed significant differences in the variables of households with children under two years old and ownership of proper toilets.