Traffic Crashes and Smartphone Usage Analysis

A comprehensive data analysis project investigating the relationship between smartphone usage rates and traffic crash rates in the United States from 2006 to 2020.

🎯 Project Overview

This case study explores whether there's a correlation between the increasing adoption of smartphones and traffic crash rates, using statistical analysis and machine learning techniques.

📊 Dataset

The analysis uses two main datasets:

• traffic.csv: Monthly traffic crash data (2006-2020)
  • Date, crashes per 100,000 people, season
• crashes_smartphones.csv: Smartphone usage survey data (2011-2019)
  • Survey dates, smartphone usage percentages, corresponding crash rates

🔬 Analysis Methods

1. Exploratory Data Analysis

   • Time series visualization of crash rates
   • Seasonal pattern analysis
   • Smartphone usage trends

2. Statistical Analysis

   • Pearson correlation coefficient
   • Statistical significance testing (p-values)
   • Linear regression modeling

3. Predictive Modeling

   • Linear regression to predict crash rates
   • Model validation against 2020 actual data
   • Visualization of predictions vs. reality

📈 Key Findings

• Positive Correlation: Moderate positive relationship (r > 0.5) between smartphone usage and crash rates
• Statistical Significance: p-value < 0.05, indicating non-random correlation
• Seasonal Patterns: Winter and Fall show higher crash rates than Spring and Summer
• Prediction Accuracy: Model provides reasonable estimates but with notable deviation from actual 2020 data

🛠️ Technical Stack

• Python 3.8+
• Pandas - Data manipulation and analysis
• NumPy - Numerical operations
• Matplotlib - Data visualization
• Seaborn - Statistical plotting
• Scikit-learn - Machine learning and regression
• SciPy - Statistical analysis

📁 Project Structure

Case-Study/
├── CaseStudy1/
│   ├── notebook.ipynb          # Main analysis notebook
│   ├── traffic.csv              # Traffic crash data
│   └── crashes_smartphones.csv   # Smartphone usage data
├── CaseStudy2/
│   └── [additional analysis]
├── requirements.txt             # Python dependencies
├── README.md                   # This file
└── .gitignore                 # Git ignore file

🚀 Getting Started

1. Clone the repository

   git clone <repository-url>
   cd Case-Study

2. Install dependencies

   pip install -r requirements.txt

3. Run the analysis

   • Open CaseStudy1/notebook.ipynb in Jupyter Notebook or JupyterLab
   • Execute cells sequentially from top to bottom

📋 Requirements

See requirements.txt for detailed package versions. Key dependencies include:

• pandas >= 1.3.0
• numpy >= 1.21.0
• matplotlib >= 3.4.0
• seaborn >= 0.11.0
• scikit-learn >= 1.0.0
• scipy >= 1.7.0

🔍 Analysis Highlights

Correlation Analysis

• Pearson's r coefficient calculated between smartphone usage and crash rates
• Statistical significance testing to validate findings
• Visual correlation through scatter plots with regression lines

Time Series Analysis

• Monthly crash rate trends over 15-year period
• Identification of seasonal patterns and anomalies
• Comparison of pre- and post-smartphone era patterns

Predictive Modeling

• Linear regression model: Crashes_per_100k = 120.6637 + (0.6610 * Smartphone_usage)
• 2020 prediction vs. actual comparison
• Model performance evaluation

📝️ Notes

• Correlation vs. Causation: While statistical significance is found, correlation does not imply causation
• External Factors: Many variables contribute to traffic crashes beyond smartphone usage