Correlation Studio

Correlation Studio is a web-based SaaS statistics application for identifying correlations in data by running experiments on your own or AI-discovered datasets, then compiling findings into shareable portfolios.

What is Correlation Studio?

Correlation Studio is a statistics platform that takes time-series or tabular datasets as input, runs correlation experiments on them, and outputs discoveries expressed as correlation coefficients (like r=0.9939) and visual portfolio pages. It runs as a web application, available at https://correlationstudio.com. The site includes a chatbot named Corrie who acts as curator and guide.

Key Features

• Remote dataset wizard AI — Search for and ingest datasets from external sources (e.g., FRED, DataHub, Our World in Data) directly into the workspace without writing code.
• Experiments — Run correlation tests between any two datasets, with results shown as Pearson r values; for example, US imports of passenger cars vs. crude oil prices shows r=0.9939.
• Discoveries — Automatically surface high-correlation pairs across your datasets; the site lists discoveries like the 0.9983 correlation between world population and US employment.
• Portfolios — Combine experiments, discoveries, datasets, and notes into shareable project pages with AI-generated analysis.
• Corrie chatbot — An on-site assistant that answers questions, curates content, and helps you navigate datasets and workflows.
• Trending topic hubs — The homepage organizes content around current events (e.g., US-Canada tariff dispute, Europe wildfire heatwave) with relevant datasets and analyses.
• CSV export/import — Datasets are presented as CSV files with row/column counts, e.g., 176 rows for ICT tariffs, 2,504 rows for NASA FIRMS MODIS Canada.

Who is it for?

• Data analysts — Quickly test hypotheses about economic indicators, such as the correlation between the PCE price index and CPI-U (r=0.9984).
• Researchers — Explore environmental datasets like NASA FIRMS active fire detections or global wildfire burnt-area series without coding.
• Students and educators — Use real-world datasets (S&P 500, gold prices, unemployment levels) to teach correlation analysis and statistics.
• Policy wonks — Analyze tariff policy impacts using datasets like the Kiel Tariff Deltas CAN-USA or ICT tariff indicators for Mauritius.

What can you do with Correlation Studio?

• Track market relationships: Correlate S&P 500 daily returns with oil prices (WTI) or compare NASDAQ Composite with NASDAQ100, getting r values and charts.
• Monitor environmental phenomena: Upload NASA FIRMS active-fire hotspot CSV data for Canada or global detections and find correlations with temperature or GDP.
• Build a data portfolio: Combine your experiments and datasets into a clean portfolio page (e.g., "Getting Started" or "S&P 500 vs Oil") and share it with colleagues.

How does Correlation Studio work?

1. Bring or find data — Upload your own CSV files or use the remote dataset wizard AI to pull datasets from sources like FRED or DataHub.
2. Run experiments — Select two datasets and compute correlations; the platform returns statistical coefficients instantly.
3. Explore discoveries — Let the system surface unexpected high correlations across your dataset collection.
4. Compile a portfolio — Organize your findings, add AI analysis, and publish a shareable page with a custom URL.