Project 4 — COVID-19 Data Analysis¶
Analyse one of the most consequential datasets of our time — and learn why analytical rigour and honest communication matter most when the stakes are high.
Overview¶
The COVID-19 pandemic generated an unprecedented volume of public data. This project uses global case, death, and vaccination data to practise time-series analysis, per-capita normalisation, and the careful communication that public health data demands. It's also a lesson in data ethics — how the same numbers can inform or mislead depending on how they're presented.
Business Objective¶
Build an analytical report and dashboard that accurately tracks the pandemic's progression across countries, normalises for population to enable fair comparison, and communicates findings without sensationalism or distortion.
Tools and Skills¶
| Tool | Used for |
|---|---|
| Python (Pandas) | Time-series cleaning, per-capita calculations, rolling averages |
| matplotlib / seaborn | Trend charts, country comparisons, small multiples |
| SQL | Aggregations, peak detection, wave analysis |
Difficulty¶
Intermediate — heavy time-series work, requires careful handling of cumulative vs daily values and per-capita normalisation.
Dataset¶
- Source: Our World in Data COVID-19 dataset
- Download: OWID COVID-19 Data
- Update frequency: Was updated daily during the pandemic; historical snapshots available
- Schema details: dataset-guide
Project Files¶
| File | Description |
|---|---|
| business-problem | Analytical objectives and questions |
| dataset-guide | Dataset schema and key columns |
| data-cleaning | Time-series cleaning, per-capita engineering |
| sql-analysis | Wave and peak analysis queries |
| dashboard-design | Dashboard design |
| insights-and-recommendations | Findings and communication guidance |
| interview-questions | Project interview questions |
Learning Outcomes¶
- Work with cumulative and daily time-series data correctly
- Normalise metrics per capita for fair cross-country comparison
- Apply rolling averages to smooth reporting noise
- Detect waves and peaks in time-series data
- Communicate sensitive public-health data responsibly