Dataset Scope
Multi-Decade, Global
Core Method
EDA & Correlation
Skills Highlighted
Data Storytelling, Hypothesis Testing
The Complex Inquiry
The objective was to move beyond simple statistics and use a multi-variable dataset to statistically confirm or reject common hypotheses about migration (e.g., is unemployment a stronger driver than political stability?). This required a disciplined approach to handle complex, often highly correlated, features.
The Statistical Solution
I employed Python-based statistical tools to analyze and communicate complex relationships in the data.
- Correlation Matrix: Generated a matrix and heatmap to identify multicollinearity between potential driving factors (e.g., GDP vs. Education Spend).
- Regional Subsetting: Focused the analysis on specific regions (like Sub-Saharan Africa) to provide targeted, meaningful insights instead of generic global observations.
- Hypothesis Refinement: Used visualization (scatter plots, box plots) to test and refine initial hypotheses, ensuring conclusions were backed by statistical evidence.
Key Findings & Data Story
The analysis provided a clear narrative: **Environmental factors, while often secondary, showed the sharpest correlation increase** with migration intent over the last decade. This conclusion provided a new, actionable dimension for humanitarian and economic policy discussion.
Visualization: Correlation Heatmap highlighting key relationships.
Explore the Full Data Story
Review the methodology and source code for this strategic analysis.
View GitHub Repository