William S. Gosset
Brewing Better Beer, Brewing New Statistics
What would cause the head brewer of the famous Guinness
brewery in Dublin. Ireland, not only to
use statistics but to invent new statistical methods? The search for better
beer, of course. William S.
Gosset (1876-1937), fresh from Oxford University, joined Guinness as a brewer
in 1899. He soon became involved in
experiments and in statistics to understand the data from these
experiments. What are the best
varieties of barley and hops for brewing?
How should they be grown, dried, and stored? The results of the field experiments, as you can guess,
varied. Statistical inference can
uncover the pattern behind the variation.
The statistical methods available at the turn of the century ended with
a version of the z test for means-even confidence intervals were not yet
available.
Gosset faced the problem we noted in using the z test to
introduce the reasoning of statistical tests:
he didn’t know the population standard deviation s. What is more,
field experiments give only small numbers of observations. Just replacing s by s in the z statistic and calling the result roughly
normal wasn’t accurate enough. So
Gosset asked the key question: what is
the exact sampling distribution of the statistic (x-u)/s?
By 1907 Gosset was brewer-in-charge of Guinness’s experimental brewery. He also had the answer to his question and had calculated a table of critical values for his new distribution. We call it the t distribution. The new t test identified the best barley variety, and Guinness promptly bought up all the available seed. Guinness allowed Gosset to publish his discoveries, but not under his own name. He used the name “Student” and the t test is sometimes called “Student’s t” in his honor. Gosset’s statistical work helped him become head brewer, a more interesting title than professor of statistics.
From Moore, D.S. The Basic Practice of Statistics, 2nd Edition. W.H. Freeman, 2000.