Student Segmentation Based on Commuting Behaviour and Mental Health Using K-means Clustering
Year: 2026
Type: article
Study design: Observational Study
Abstract: Understanding the impact of commuting on students' mental health and academic performance is crucial in higher education research. This study employs the K-means clustering technique to classify students based on their commuting behaviors and the resultant psychological and academic effects. An analyzed dataset includes variables concerning travel characteristics, stress levels, sleep patterns, and academic impact, rigorously preprocessed and standardized for accurate results. The study identifies the optimal number of clusters using the Elbow Method and Gap Statistic, revealing a three-cluster solution. The analysis results in three significant student segments: Moderate Stress Commuters, High Stress Commuters, and Low Stress/Short Distance Commuters. The High Stress Commuters exhibit notably elevated levels of anxiety, sleep deprivation, fatigue, and academic struggles, while the Low Stress group enjoys comparatively better mental health with minimal academic disruptions. The clustering's validity is validated through silhouette analysis, which points to a moderately acceptable cluster structure. Additionally, oneway ANOVA results underscore a statistically significant difference in stress levels across the clusters (p < 0.001), affirmed further by Tukey's post hoc test demonstrating significant pairwise differences among the groups. These findings stress the essential influence of commuting conditions on students' mental well-being and academic achievements. This research offers educational institutions critical insights for the development of tailored interventions, such as enhanced transportation support and mental health resources, aimed at improving student wellbeing and academic performance.
Author: A. Poompavai
Language: English
Topic: Student Stress and Coping
Subfield: Developmental and Educational Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7223341954