Systems | Development | Analytics | API | Testing

September 2019

Anscombe's Quartet

In this blog post I’m going to write about a famous piece in visualization history. How can we prove that a visualization is more worth than just looking at the data? That’s a question Francis Anscombe probably asked himself when he back in 1973 constructed the dataset that became known as Anscombe's quartet. A dataset he could use to show statisticians how wrong they were thinking that “numerical calculations are exact, but graphs are rough."