Online Retail Statistical Analysis

📌 Project Overview

This project performs statistical analysis on the Online Retail II dataset (Kaggle).
The goal is to explore customer purchasing patterns, identify high-value segments, and test hypotheses regarding sales behavior across regions and time periods.

✨ Features

• Data cleaning and preprocessing
• Exploratory Data Analysis (EDA)
• Statistical hypothesis testing (t-tests, ANOVA, Chi-Square, etc.)
• Visualization of sales trends and customer behavior
• Business insights from data-driven results

📂 Repository Structure

online-retail-statistical-analysis/
│
├── data/                   # Dataset (not included due to license)
├── scripts/                # R scripts for analysis
├── images/                 # Generated plots
├── requirements.txt        # Dependencies (R packages)
└── README.md               # Documentation

🚀 Getting Started

1. Clone the repository

git clone https://github.com/erenbg1/online-retail-statistical-analysis.git
cd online-retail-statistical-analysis

2. Install dependencies

In R, install required libraries:

install.packages(c("tidyverse", "ggplot2", "dplyr", "lubridate", "car"))

3. Run analysis

Open the scripts in scripts/ folder and run them in RStudio or terminal:

source("scripts/data_cleaning.R")
source("scripts/eda_visualizations.R")
source("scripts/hypothesis_testing.R")

📊 Example Results

• Top 5 Countries by Sales Volume: United Kingdom, Netherlands, Germany, France, Ireland
• Hypothesis Test: Significant difference found between sales volume across weekdays vs weekends.
• Visualization: Monthly sales trend shows strong seasonality (peaks in Q4).

📈 Future Work

• Perform advanced customer segmentation (RFM analysis, clustering).
• Introduce predictive modeling for sales forecasting.
• Add anomaly detection for unusual purchasing patterns.
• Deploy interactive dashboards (Shiny).

📝 License

This project is released under the MIT License.