# Proof, tea and Fisher: the origin of experimental design

In the design of experiments in statistics, the lady drinking tea is a random experiment reported by Ronald Fisher in his 1935 book The Design of Experiments.

One day, Ronald Aylmer Fisher (the inventor of experimental design) witnessed a scene in which a lady was outraged:

He was served tea with milk first!

Fisher, amused, doubted, however, that the order in which the tea and milk were added would affect the taste of the drink. The lady said she could taste a difference.

Fisher therefore challenged it. He devised a test that is still known as Fisher's Exact Test. All you have to do is compare 4 cups for each configuration. If all 8 cups are recognised, then there is statistical proof of the difference.

The lady had no trouble recognising the order in which far more cups were prepared than was necessary to prove her claim. Fisher's description in his book is less than 10 pages long and is notable for its simplicity and completeness in terms of terminology, calculations and experimental design.

Even if the Y is not measurable, we can easily provide statistical proof of a difference. Six Sigma is sometimes as simple as that! To find out more about experimental design, you can follow our training.

Retour en haut