Why identical summary statistics can hide wildly different visual patterns in your data
In short: Data visualization often relies on summary statistics, but these can be deceptive. This video explores the Datasaurus Dozen, a collection of datasets that share nearly identical statistical properties despite having vastly different visual shapes, demonstrating why looking at the raw data is essential.
The Datasaurus Dozen project, developed by Justin Matejka and George Fitzmaurice, challenges the assumption that summary statistics provide a complete picture of a dataset. By using a technique called simulated annealing, they generated multiple datasets that maintain the same statistical profile—such as mean and variance—while producing distinct, recognizable shapes when plotted.
The original concept was inspired by Alberto Cairo's 'Datasaurus' plot. This work highlights the critical importance of visualization in data analysis, as relying solely on numerical summaries can obscure underlying patterns or outliers that are immediately apparent when the data is graphed.
Opens the Intellect+ feed with this card first. Swipe up for the next one.
Sources
- Summary by Intellect · source: The Datasaurus Dozen
To cite this card, link to https://intellectplus.co.uk/app/science-the-datasaurus-dozen.
How we source, check and correct cards: Editorial standards.
General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.
Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?
Related
More in Science · Latest · A–Z · All topics