#OTD 1876 William Sealy Gosset b (d 16 Oct 1937) 🇬🇧Best known for Student's t-statistic & distribution, developed 1908 while working for Guinness Brewery. He is probably the first modern industrial statistician. 1/
2/ Guinness had almost doubled beer production between 1887 & 1914 so consistent quality was a concern. The problem was determining how representative a small sample might be of the whole batch. Gosset was assigned to the problem because he had taken maths at Oxford
3/ so was “less scared of this kind of problem than the other brewers”. Fisher greatly admired Gosset, calling him “one of the most original minds in contemporary science”.
4/ Fun fact: Gosset published under the pseudonym “Student”. Guinness employees were not allowed to publish under their own names. One story has it C.D. LaTouche, the managing director of Guinness gave him the choice of “A Student” or “A Pupil’.
5/ Ziliak suggests it originated from his 1906–1907 notebook which had "The Student's Science Notebook" on the cover. The notebook contained data on counting yeast cells with a haemacytometer
Ref Student 1908 The probable error of a mean Biometrika 4:1-24
Ziliak ST 2008. The great skew: R A Fisher and the copyright history of ‘Student’s’ t. faculty.roosevelt.edu/Ziliak
• • •
Missing some Tweet in this thread? You can try to
force a refresh
WS Gosset was interested in the practical application of statistics when the “problem is to get results as quickly and as cheaply as possible”. In his famous 1908 paper he proposed a criterion for the practical significance of a result 1/
2/ He proposed 3x the probable error of the normal curve. This would be roughly a one-sided p-value of 0.02 (1 in 50 odds), because probable error is ~0.67456 of a standard deviation.
3/In 1927 he weighed in on what he called “the vexed question of the repetition and rejection of results”. He works out correlated errors in a sequential small-n data series, makes recommendations for identification of discordant observations & suggests some procedural rules.
"Student" Part 2: The paper now described as “path-breaking” received little if any notice at first. Gosset was a chemist, not a mathematician, so he struggled with proofs. He “guessed” (his words) the correct form of the distribution he called z based on the 1/9 🧵
2/“properties of correlation coefficient & Professor Pearson's types of frequency curves".
Still a student at Cambridge, RA Fisher was the first to recognise the importance of Gosset’s 1908 paper, not just for its practical importance but b/c it was central
3/ to understanding an entire family of sampling problems. He went to his tutor FJM Stratton to discuss discrepancies between Gosset’s results & his own. Stratton had met Gosset on the latter’s previous visits to Cambridge. He suggested Fisher write to Gosset which he did.
The first population #census in North America was performed in 1665-6 by Jean Talon (“Canada’s first official statistician”) 🇫🇷. He was appointed by Louis XIV of France to improve the management of French colonies in what is now 🇨🇦. 1/
2/ He needed to measure the population to “gauge the progress of European colonization” & implement policies to diversify the economy & strengthen governance.
3/The census occurred during a hard winter, which actually was an advantage as most people had to stay home & could not travel. Talon proudly reported 3,418 people in total. However, there were many omissions -accidental & otherwise - duplications, & mistakes in adding up.
#OTD 1881 Hilda Hudson b (d 26 Nov 1965)🇬🇧 OBE. Best known to statisticians for developing the classic SIR model of epidemic infectious disease with Ronald Ross 1916-17 she also pioneered application of sophisticated mathematics to aeronautical engineering. 1/5
2/ She was the first ever female invited speaker at International Congress of Mathematicians 1912 where she read a paper on curved surfaces. Semple called her “a distinguished mathematician of great erudition and integrity”.
3/ During WWI, she joined the Admiralty to head the Structural Analysis section. With Letitia Chitty (1897-1982) & Beatrice Cave-Browne-Cave (1874-1947) they revolutionised fixed-wing aircraft design. After the war, they wrote the classic Handbook of Strength Calculations.
In June 1905 Karl Pearson introduced the terms ‘kurtosis’, ‘leptokurtic’, ‘platykurtic’ & ‘mesokurtic’ to describe shapes of skewed frequency distributions. The paper was otherwise a huffy (& lengthy) rejoinder to some German critics of his 1899 paper on skew variation 1/4
2/ As well as highly entertaining criticisms of his foes, the paper also presents derivations & distribution data for a variety of biological data, including human skulls, crab ‘foreheads’, shell lengths, & organ weights.
3/ In 1927 WS Gosset ("Student") provided a humourous aide memoire: 'platykurtic' = platypus w short tails vs 'leptokurtic' = 2 kangaroos ‘lepping’ (I don't make this stuff up you know)
#OTD 2012 Genuchi Taguchi d (b 1 Jan 1924) 🇯🇵 founder of the Taguchi method for quality product improvement - reduction of process variation through robust design of experiments. His methods revolutionized manufacturing quality control practices & mindsets. 1/4
2/ He collaborated throughout the 1950s with other notable statisticians such as CR Rao Fisher Shewart & was sponsored by Tukey at Princeton
3/ his methods differed from the conventional specifications based on tolerances alone but developed the concept of quality loss (rather than just quality)