Identification of Social Anxiety in High School: A Comprehensive Machine Learning Approach to Real-Time Student Behavior Analysis

Abstract

This research presents a comprehensive study on identifying social anxiety among high school students using data-driven techniques. The study integrates machine learning algorithms, clustering models, and correlation analysis to evaluate psychological and behavioral patterns from the Social Phobia Inventory (SPIN) questionnaire. Data collected from students of Little Scholars Matriculation Higher Secondary School, Tamil Nadu, forms the basis of this analysis. Results highlight distinct clusters of anxiety levels, their correlation with academic performance, and behavioral triggers. The paper further proposes Virtual Reality (VR) and Augmented Reality (AR)-based interventions to reduce anxiety, aiming to provide a modern, technology-assisted mental health support system for educational institutions.

Introduction

Social anxiety disorder (SAD) significantly impacts adolescents’ communication, confidence, and academic performance. Traditional assessments rely on subjective evaluations, which can be limited in scalability. This study introduces a machine learning framework to objectively assess, classify, and interpret anxiety traits in real time, enabling early identification of at-risk students and personalized interventions.

Objectives:

• Assess prevalence and intensity of social anxiety in high school students.

• Explore correlations between anxiety traits and student characteristics.

• Apply clustering and predictive analytics to identify patterns.

• Propose VR/AR tools to reduce anxiety through immersive experiences.

Methodology:
A mixed-method approach was used on 500 students aged 13–17 via the 17-item SPIN questionnaire. Responses were preprocessed for machine learning analysis. Supervised algorithms (Decision Trees, Random Forests) identified predictors, while K-Means clustering categorized students into anxiety-level groups. Ethical guidelines were strictly followed, including informed consent and counseling support.

Results:

• About 37% of students showed moderate to severe social anxiety.

• Strong correlations were found between self-esteem, fear of public performance, and avoidance behaviors.

• Five distinct anxiety clusters were identified.

• Machine learning models predicted anxiety with 91% accuracy.

Discussion:
Data-driven approaches effectively reveal emotional well-being patterns. High-anxiety students exhibited avoidance and physiological symptoms. Integrating VR-based exposure therapy can provide personalized interventions, highlighting the importance of proactive psychological support in schools.

Conclusion

The study successfully demonstrates the applicability of machine learning in identifying and analyzing social anxiety among adolescents. The model provides educators and mental health professionals with actionable insights into student behavior, enabling early intervention. By combining traditional psychology with technological innovation, this framework paves the way for data-assisted emotional health monitoring in schools.

References

[1] Hofmann, S.G. (2007). Cognitive factors that maintain social anxiety disorder. Cognitive Behavioral Therapy. [2] Furmark, T. (2002). Social Phobia: Overview of Community Surveys. Acta Psychiatrica Scandinavica. [3] Carpenter, K.L.H. (2016). Quantifying Risk for Anxiety Disorders in Preschool Children: A Machine Learning Approach.

Copyright

Copyright © 2025 Miss. Snehal Lakhe , Prof. Shafiya Sayyad, Prof. Mukul Jagtap . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.