Clustering Undergraduate Students Based on Their Self-Esteem and Academic Achieve- ment Via The K-Means Approach

Abstract

Introduction: Self-esteem is one of the key foundations of human personality and is known as an important component in mental health and social psychology. Students with high self-esteem tend to be more engaged and persistent in areas of achievement. This study was devoted to cluster undergraduate students of Shahrekord University of Medical Sciences based on their self-esteem and academic achievement.

Methods: The multi-stage cluster sampling method (three faculties and three departments in each) was used to select 260 undergraduate students from various fields in 2022. The data collection tool included a background information checklist, a 10-item self-esteem questionnaire, and a 39-item academic achievement questionnaire. The elbow method was used to estimate the number of optimal clusters. The NbClust package in R 4.2.1 software was used for clustering analysis based on the k-means approach.

Results: In this study, out of 260 participating students, 176 (67.7%) were girls and 84 (32.3%) were boys. The overall mean ± standard deviation of academic achievement was 105.2 ± 10.3. There is a positive and significant correlation (r = 0.44) between academic achievement and self-esteem (P-value <0.001). The optimal number of clusters was estimated as four based on the elbow method. Self-esteem in cluster number 1 with 35 students was at the lowest level at -2.6±2.9. The academic achievement was significantly different among the obtained clusters (P<0.001). Cluster number 4 with 48 students had less academic progress with 94.0 ± 6.1 than the other three clusters.

Conclusion: Based on the obtained findings, performing effective interventions for promoting self-esteem and academic achievement seems necessary.