How Beer Brewing Revolutionized Modern Statistics
If you’re bringing home a pint of Guinness this St. Patrick’s Day, don’t forget to thank the brewery for its contribution to science – and no, I’m not talking about nitrogen pouring. Without Guinness, we might not have one of the most important statistical tools in modern science.
William Sealy Gosset was Guinness’ chief experimental brewer at the turn of the 20th century and he had a problem. Guinness wanted to increase production, and Gosset needed to find a way to test large, heavy populations by examining a smaller sample. If successful, it would allow the company to streamline its processes in everything from quality control to purchasing ingredients in bulk.
For example, if Guinness only wanted hop flowers with a specific resin content, how would it be sure that a small sample of the harvest was representative of the entire harvest? Testing the entire crop would be impossible, testing even a large sample would be economically infeasible, and testing a small sample could introduce statistical noise into the equation. Unfortunately, at the time, statisticians were concerned about larger sample sizes and had not studied the issue.
Read more: “The man domesticated by beer”
To find the answer, Guinness sent Gosset to University College London in 1906 so he could learn from biostatistician Karl Pearson. By comparing small sample means to larger sample means, Gosset determined how much the two differed, which allowed him to develop the t distribution. While larger sample sizes produced a normal bell curve with a tight peak and flat tails, the Gosset t distribution had a shallower peak and larger tails, reflecting the noise inherent in small samples.
Published in Biometrics In 1908, under the pseudonym “Student” (potentially to disguise the fact that Guinness was conducting statistical research), Gosset created what is today known as the “Student’s t-test”. The test, later perfected by famed mathematician Ronald Fisher, allowed statisticians to determine whether sample means differed in a statistically significant way from population means. Today it is an essential tool used by researchers in all scientific fields, from astronomy to zoology.
The T test represented a huge scientific leap forward, and it was not developed for war, or even out of basic necessity – simply to brew a better pint of stout.
Enjoy Nautilus? Subscribe for free to our newsletter.
Main image: sweeticons / Shutterstock
