Predicting Mental Health Outcomes for Native and Non-Native Language Students Through ICT-Based Learning Analytics

Abstract

This paper presents the impact and assessment of Information and Communication Technology (ICT)-based learning techniques employed by native and non-native language medium schools and predicts the factors that affect the mental health outcomes of students in these medium schools. The implementation of this study has been executed into five stages: (i) implication of reliable & aligned questionnaire associated with this study, (ii) targeting a group of students for data collection, (ii) performing the data reliability and data analysis tasks, (iv) deriving the key factors associated with the objective of this study, and (v) finally, building the prediction models as prototypes based on the textual analysis of the employed questionnaire for the mental health predictions among students. This study is being carried out using questionnaire-based data collected from 400 students in the Indian education system. Here, the medium of schools is Bengali (native) and English (non-native) languages. After data collection, rigorous statistical and machine learning-based data analysis techniques are employed. In the findings, Statistical tests (e.g., Mann–Whitney U, Kaiser-Meyer-Olkin (KMO), and Bartlett) revealed that English-medium students, particularly girls, reported higher levels of perceived stress compared to their Bengali-medium counterparts. Factor analysis confirmed more complex stress structures in English-medium groups, and predictive models achieved over 81% accuracy in classifying stress profiles. These findings suggest that language background is a significant factor in shaping students' psychological well-being in ICT-based education systems. The outcome of this study will be used to derive the key factors for mental health problems among students and build the prediction model that will be derivable from this study.

Received: 7 April 2025 | Revised: 19 August 2025 | Accepted: 13 November 2025

Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.

Data Availability Statement

Data are available from the corresponding author upon reasonable request.

Author Contribution Statement

Kunal Ghosh: Conceptualization, Methodology, Software, Formal analysis, Resources, Data curation, Writing — original draft, Writing — review & editing, Visualization, Project administration. Saiyed Umer: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Writing — original draft, Writing — review & editing, Visualization, Supervision, Project administration.