Understanding Student Anxiety in the Digital Age: A Machine Learning Study of Social Media and Academic Stress

Abstract

Mental health is an issue affecting university students in India who are aged between 18-24 years. This is due to the fact that they are faced with a lot of school work, money issues, what other people think of them, and the fact that they are using social media all the time. Despite the fact that the university is becoming more aware of the issue, the students are too afraid to seek assistance due to the fact that they do not want to be judged. In addition, the university does not have people to assist the students in the issue of mental health. In this study, we were able to examine the amount of school work the students are faced with, the amount of media they are using, and the effect it has on them. We gave out questionnaires to 500 students in Jain University, and we were able to analyze the results using computers to determine what is stressing the students. Some students are at high risk, while others are at moderate risk, while others are at low risk of being affected mentally. We learned that if students don’t like the university they are in, then students will be more anxious. We learned that if students compare themselves with other people, then students will be more likely to use too much social media and aren’t getting enough help from the university they are in, then students will be more anxious. If students stop using too much social media, then students will feel better. We think that the university should help students with their health issues more. The university should hire more people that students could talk to, hire more teachers that students could get information from about how to stay healthy, and hire more teachers that students could get advice from, and hire more teachers that could help students that are more likely to have health problems with their mental issues.

Introduction

Mental health plays a crucial role in university students’ academic performance and overall well-being, but many students aged 18–24 face high levels of stress due to academic pressure, financial issues, and major life transitions. The problem is more severe in India due to stigma around mental health, limited institutional support, and lack of open discussion. Social media further contributes to anxiety and depression through constant comparison, fear of missing out, and excessive passive usage.

The study identifies a clear research gap: most previous work relies on surveys and descriptive analysis rather than machine learning–based data-driven approaches to understand student mental health. To address this, the research uses a dataset of 500 university students and applies machine learning techniques to analyze stress factors, social media behavior, and help-seeking patterns.

The methodology includes a structured survey with 27 variables, data preprocessing in Python, and application of K-Means clustering to group students into three mental health profiles. A Random Forest model is also used to predict anxiety levels, with evaluation based on accuracy and cross-validation. Ethical considerations such as anonymity and informed consent are ensured.

Key findings show that a large proportion of students frequently experience anxiety after using social media, with “always anxious” being the most reported category. Social comparison behavior strongly correlates with higher anxiety levels. The study also finds a significant gap between students recognizing the need for mental health support and actually seeking help, likely due to stigma and accessibility issues. Additionally, students who take intentional breaks from social media report lower anxiety levels, suggesting digital detox as a useful coping strategy.

K-Means clustering reveals distinct student mental health profiles, and the Random Forest model helps identify key contributing factors to anxiety. Overall, the study concludes that social media use, academic stress, and lack of support systems significantly impact student mental health, and it provides data-driven insights for universities to design better intervention and support strategies.

Conclusion

The present study has clearly proved that anxiety caused by social media is a prevailing and essentially chronic condition among Indian university students, with 70.0% of the surveyed student population experiencing anxiety \'Often\' or \'Always\' after using social media. With the use of KMeans clustering and Random Forest classification algorithms, this study has successfully been able to identify three different categories of student mental health risk profile and establish dissatisfaction with institutions as the strongest predictor for anxiety caused by social media among Indian university students (importance = 0.0871). It is important to realize and understand that for effective solutions to be developed and implemented to solve this problem, they need to be holistic in nature, considering not just behavioral aspects of anxiety caused by social media, but also institutional aspects related to dissatisfaction with support provision. The approach adopted in this study, using machine learning-based models for student mental health risk profiling, can be replicated in other universities in India, leading to a positive impact on policy reforms in those institutions.

References

[1] Z. H. Wang, A. L. Yang, N. N. Yang, Z. Liu, Q. Wang, L. Zhu, W. J. Hu, X. P. Liu, Y. C. Xiong, G. T. Cao, and C. Tao, \"Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus– infected pneumonia in Wuhan, China,\" Journal of Medical Internet Research, vol. 22, no. 9, p. e22822, Sep. 2020. DOI: 10.2196/22822 [2] S. Kakollu and A. A. Dar, \"Mental health challenges among university students in India: Academic pressure, digital exposure, and post-pandemic stressors,\" Asian Journal of Psychiatry, vol. 106, p. 104447, 2025. DOI: 10.1016/j.ajp.2025.104447 [3] S. M. Gaiha, T. T. Salisbury, M. Koschorke, U. Raman, and M. Petticrew, \"Stigma associated with mental health problems among young people in India: A systematic review of magnitude, manifestations and recommendations,\" BMC Psychiatry, vol. 20, p. 538, 2020. DOI: 10.1186/s12888-020-02937-x [4] B. Keles, N. McCrae, and A. Grealish, \"A systematic review: The influence of social media on depression, anxiety and psychological distress in adolescents,\" International Journal of Adolescence and Youth, vol. 25, no. 1, pp. 79–93, 2020. DOI: 10.1080/02673843.2019.1590851 [5] N. K. R. Sehgal, H. Kambhamettu, S. P. Matam, L. Ungar, and S. C. Guntuku, \"Barriers to AI-based mental health support adoption among Indian students: Stigma, privacy, and personalisation,\" in Proc. CHI Extended Abstracts (CHI EA \'25), 2025. DOI: 10.1145/3706599.3720137 [6] R. K. Chadda, \"Youth and mental health: Challenges ahead,\" Indian Journal of Medical Research, vol. 148, no. 4, pp. 359–361, Oct. 2018. DOI: 10.4103/ijmr.IJMR_1754_18 [7] T. Richardson, P. Elliott, R. Roberts, and M. Jansen, \"A longitudinal study of financial difficulties and mental health in a national sample of British undergraduate students,\" Community Mental Health Journal, vol. 53, no. 3, pp. 344–352, 2017. DOI: 10.1007/s10597-016-0052-0 [8] S. D. Hershner and R. D. Chervin, \"Causes and consequences of sleepiness among college students,\" Nature and Science of Sleep, vol. 6, pp. 73–84, 2014. DOI: 10.2147/NSS.S62907 [9] A. L. Rebar, R. Stanton, D. Geard, C. Short, M. J. Duncan, and C. Vandelanotte, \"A meta-meta-analysis of the effect of physical activity on depression and anxiety in non-clinical adult populations,\" Health Psychology Review, vol. 9, no. 3, pp. 366–378, 2015. DOI: 10.1080/17437199.2015.1022901 [10] B. Girase, R. Parikh, S. Vashisht, A. Mullick, V. Ambhore, and S. Maknikar, \"Mental Healthcare Act 2017 and its implications for rural youth mental health access in India,\" SSM – Mental Health, vol. 2, p. 100145, 2022. DOI: 10.1016/j.ssmmh.2022.100145

Copyright

Copyright © 2026 Vaishnavi Sridhar, Vasanth E, Ritika M, Ritvik Sasikumar, Rohan , Mr. Karthikeyan . 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.