📊 Statistical Research: Mental Health vs Academic Performance

📌 Overview

This project analyzes the impact of mental health factors on students’ academic performance (CGPA) using a large dataset. It includes:

• 📊 Data visualization (scatter plots, heatmaps)
• 🤖 Predictive modeling (Linear Regression)
• 🔬 Statistical hypothesis testing (p-values)
• 📄 Automated PDF research report generation

---

🎯 Objective

To evaluate how mental health indicators such as:

• Stress
• Anxiety
• Depression
• Burnout
• Overall Mental Health Index

affect students’ academic performance (academic_performance as proxy for CGPA).

---

📂 Dataset

• File: student_mental_health_burnout_1M.csv

• Contains:

  • Mental health metrics
  • Academic performance indicators

---

🧠 Methodology

1. Data Analysis

• Identified relevant features:

  • Independent variables: mental health indicators
  • Dependent variable: academic performance

---

2. Visualization

Generated:

• Scatter plots for each feature vs CGPA
• Correlation heatmap

---

3. Predictive Modeling

• Model: Linear Regression

• Train-test split: 80/20

• Metrics:

  • R² Score
  • RMSE

---

4. Hypothesis Testing

• Pearson correlation test used

• Computed p-values for each variable

• Significance threshold:

  p < 0.05

---

5. Report Generation

• Generated a PDF research report

• Includes:

  • Model performance
  • Coefficients
  • Statistical results
  • Visualizations

---

🛠️ Tech Stack

• Python 🐍
• Pandas
• NumPy
• Matplotlib
• SciPy
• Scikit-learn
• ReportLab

---

📦 Installation

pip install pandas numpy matplotlib scipy scikit-learn reportlab

---

▶️ How to Run

1. Place dataset in project folder:

student_mental_health_burnout_1M.csv

2. Run the script:

python main.py

3. Output:

• 📊 Charts (PNG files)
• 📄 mental_health_research.pdf

---

📊 Results Summary

• Mental health variables show weak correlation with academic performance
• Predictive model shows low explanatory power
• Suggests other factors influence CGPA more strongly

---

⚠️ Limitations

• Correlation does not imply causation

• No time-series analysis

• Missing external factors:

  • Study habits
  • Sleep patterns
  • Socioeconomic background

---

🚀 Future Improvements

• Use advanced ML models (Random Forest, XGBoost)
• Feature engineering
• Add more behavioral data
• Time-series or longitudinal analysis
• Interactive dashboards (Plotly / Power BI)

---

📄 Output Example

• mental_health_research.pdf
• Scatter plots
• Correlation heatmap

---

🤝 Contributing

Feel free to fork and improve:

• Add new models
• Improve visualizations
• Enhance statistical rigor

---

📜 License

This project is for educational and research purposes.

---

👨‍💻 Author

Shivakumar Shivampeta