Student Mental Health Analysis: Understanding the Silent Struggle

In today’s fast-paced and competitive academic world, mental health among students has become a critical issue—yet it remains under-discussed and stigmatized. In this blog, we dive into a data-driven analysis of student mental health, exploring the underlying factors, patterns, and actionable insights uncovered through our research project.

Overview

Objective:
To analyze various demographic and lifestyle factors affecting student mental health using data analytics and machine learning techniques.

Tools & Technologies:

  • Python (Pandas, Matplotlib, Seaborn)
  • Jupyter Notebook
  • Machine Learning (Logistic Regression, Random Forest)

Dataset:
The dataset was sourced from Kaggle and includes attributes such as age, gender, course, year of study, GPA, sleep habits, and reported stress or mental health conditions.

Exploratory Data Analysis (EDA)

The first step in our process was cleaning and visualizing the data to uncover patterns. Key findings include:

  • Gender distribution: Female students were slightly more represented.
  • Stress prevalence: A significant number of students reported high stress levels and poor mental health.
  • Sleep hours: Strong correlation between sleep deprivation (<5 hours) and increased mental distress.
  • Academic Pressure: Final-year students reported the highest stress levels.

Machine Learning Models

We implemented two classification models to predict the likelihood of students experiencing mental health challenges:

1. Logistic Regression

  • Accuracy: ~79%
  • Highlights the linear relationship between features like sleep, GPA, and academic year.

2. Random Forest Classifier

  • Accuracy: ~85%
  • Captures non-linear relationships and provides feature importance ranking.

Top Features Identified:

  • Sleep Hours
  • Academic Year
  • Daily Screen Time
  • GPA

Key Takeaways

  • Sleep deprivation is a significant factor in poor student mental health.
  • Stress increases considerably in later academic years.
  • Lower GPA correlates with higher mental distress.
  • High screen time, though less dominant, contributes to mental fatigue.

Recommendations

Based on the analysis, the following interventions are recommended:

  • Mental Health Workshops: For awareness and emotional resilience.
  • Accessible Counseling: On-campus or virtual psychological support.
  • Sleep & Screen Hygiene: Promoting healthy digital habits and rest.
  • Academic Mentorship: Pairing senior mentors with juniors to ease academic pressure.

Conclusion

Mental health is not a standalone issue—it is a complex interplay of lifestyle, academic, and emotional factors. By leveraging data analysis and machine learning, we can uncover the hidden patterns that affect student well-being and design targeted interventions.

This project is one step toward fostering a more empathetic and proactive academic environment.

Project Repository

You can explore the full code and analysis in the Jupyter Notebook:
Download Notebook

Acknowledgments

Special thanks to the data contributors and mentors who supported this project. Let’s continue driving conversations and policies that prioritize student well-being.

If you found this post helpful, feel free to share or reach out for feedback or collaboration.