#OTD 1857 John Snow shows again the thinking that made him one of the founders of epidemiology. In a letter to the Lancet, he suggests that the routine adulteration of bread with alum might be a cause of rickets, a mechanistic hypothesis to explain it, then a case-control design to test it. 1/5🧵

#OTD 1913 Anders Hald b (d 11 Nov 2007). ASA Fellow 1952. RSS Fellow. The father of modern Danish statistics he revolutionized statistics teaching to include mathematical statistics & theory, quality control, survey sampling, & design of experiments.

#OTD 1621 Thomas Hariot d (b c1560) Founder of the English school of algebra, he pioneered use of several math symbols such as dot ∙ for multiplication, & invented binary (base 2), 50 yr before Leibnitz. His work was almost unknown until it was made available online 2012-15 1/5

Karin Kock-Lindberg b #OTD 2 July 1891 (d 28 Jul 1946). ASA Fellow 1956. Director of Statistics Sweden and Chair of the Swedish Statistical Society she was an economist working on credit trade cycles and interest rates.

William Fielding Ogburn (29 Jun 1885- 27 Apr 1959) ASA Fellow 1920; ASA President 1931; JASA editor 1920–1926. One of most prolific sociologists of his generation, he is best known for the concept of “cultural lag” and social evolution through invention, accumulation, diffusion, and adjustment.

#OTD 1893 Prasanta Chandra Mahalanobis b (d 28 Jun 1972) OBS FRS, Hon RSS Fellow 1954, ASA Fellow 1961. The Father of statistics in India & founder of the Indian Statistical Institute & journal Sankhya. he developed the Mahalanobis distance measure & Feldman–Mahalanobis model.

#OTD 1858 Alice Lee b (d 5 Oct 1939). Showing that "negative” studies can have major impact, her doctoral research demonstrated no correlation between skull volume & intelligence. As craniometry was used to “scientifically prove” male genius & female inferiority, the fallout was massive.

#OTD 1884 President Chester Arthur signed a bill into law (23 Stat. 60) creating the US Bureau of Labor Statistics, one of the oldest statistical agencies of the US government. BLS published the first consumer expenditure survey in 1888 & wholesale prices (later called Producer Price Index) in 1902

#OTD 1883 Mary van Kleek b (d 8 Jun 1972) ASA Fellow 1945; Chair ASA Committee on Governmental & Labor Statistics. Pioneering research on industrial labor conditions, wage disparities, worker exploitation, workplace safety, & female employment in industry.

In June 1676 in a letter to Henry Oldenburg, secretary of the Royal Society, Isaac Newton uses modern notation to present negative integers as exponents. They had been used by Nicolas Chuquet in 1484 in Le Triparty en la Science des Nombres but he used an older notation.

Happy birthday to the International Statistical Institute ISI. It was formally founded 24 June 1885 during the London Statistical Society Jubilee celebration meetings. The first president was Sir Rawson W. Rawson. The first session was held in 1887 then every 2 years

#OTD 1912 Alan Turing b (d 7 Jun 1954). OBE FRS. Father of AI & pioneering work in theoretical computer science & information theory. His contributions to probability & statistics are almost unknown, as most of these papers were classified until 1980s, & 2 were only declassified in 2012.

#OTD 1885 the Royal Statistical Society publishes their Jubilee Volume featuring the groundbreaking ‘Methods of Statistics’ by Francis Ysidrio Edgeworth. The journal had nothing in mathematical statistics until this paper. From the discussion it seems nobody knew what he was talking about

Happy🎂to Siméon Poisson b #OTD 1781 (d 25 Apr 1840) His name is most associated with the Poisson distribution, developed to model jury decision-making. It gained traction in 1898 with Bortkiewicz’s analysis of Prussian horse kick data. But it is another example of Stigler’s Law of Eponomy 1/3

#OTD 1901 Karl Pearson reads a paper to the Royal Society describing experiments quantifying the ‘personal equation’ now called inter-observer measurement error. He used a measuring device constructed by Horace Darwin.

#OTD 1875 Reginald Punnett b (d 3 January 1967) He developed the Punnett square used to predict probabilities of genotypes in breeding experiments. With William Bateson he was a proponent of Mendelian genetics, becoming the target of attacks from the Biometric school -Karl Pearson and WFR Weldon.

Happy 🎂Blaise Pascal b #OTD 1623 (d 19 Aug 1662). In a series of letters, he and Pierre Fermat developed foundational concepts for modern probability after an enquiry on gambling strategy from the Chevalier de Méré. Pascal's doctor had advised him to take up gambling for health reasons. Yes. 1/5🧵

#OTD 1895 Charles P Winsor b (d 4 Apr 1951) ASA Fellow 1949. Developer of 'winsorization', the clipping of outliers by setting them to prespecified percentiles of the data. He also did important work on small samples, Gompertz growth curves & linear regression with error in both variables

#OTD 1910 Maurice S Bartlett b (d 8 Jan 2002). ASA Fellow 1951; Guy Silver & Gold Medals; RSS President 1966. Known for Bartlett’s test for homogeneity of variances, he worked on stochastic processes, time series, & multivariate analyses & introduced Monte Carlo methods to ecology & epidemiology 1/

#OTD 1937 Jerzy Neyman introduces the confidence interval in a talk given at the Royal Statistical Society. Listeners are not enthusiastic: in the vote of thanks, AJ Bowley wonders aloud if confidence intervals are just a “confidence trick”, doubts their validity & if they were any possible use.

Happy🎂Samuel S. Wilks b #OTD 1906 (d 7 Mar 1964) ASA Fellow 1940, ASA President 1950. Professor of mathematical statistics at Princeton he made major contributions to multivariate analysis (think Wilks's λ distribution), confidence interval estimation and quality control. 1/2

Two 🎂of celebrated statisticians #OTD: John Tukey and David Duncan 1/3🧵

Happy🎂 to Peter Armitage CBE b #OTD 1924 (d 14 Feb 2024) One of the few winners of all 3 Guy Medals & elected president of 3 statistical societies, he revolutionised the practice and profession of medical statistics, especially his work on statistical stopping rules for sequential medical trials

#OTD 1822 Charles Babbage reads a 1-page note to the Royal Astronomical Society in which he proposes a computational "Difference Engine", precursor of the modern computer. Twenty years later, after massive expenditure & nothing to show for it, the project was scrapped (sounds familiar tbh) 1/x🧵

Happy 🎂William Sealy Gosset b #OTD 1876 (d 16 Oct 1937). The first modern industrial statistician he is best known for Student’s t-statistic & distribution developed while working for Guinness Brewery. Fisher said he was “one of the most original minds in contemporary science.” 1/2

In 1766 Daniel Bernoulli published one of the first attempts to analyse censored data by including individual lifetimes with counts. For a formal approach we had to wait until June 1958 for Ed Kaplan and Paul Meier & another 12 years for their method to become established for medical applications.

#OTD 1881 Hilda Hudson b(d 26 Nov 1965). OBE. A distinguished mathematician & pioneer in aeronautical engineering, she is better known to statisticians for developing the classic SIR model of epidemic disease with Ronald Ross 1916-17. 1/2

In June 1948 William Feller published a unified derivation for what is now known as the Kolmogorov-Smirnov goodness of fit test. He combined & simplified the original proofs of Andrei Kolmogorov (published in Italian 1933) and Nikolai Smirnov 1939. Frank Massey popularized it in his 1951 JASA paper

Isador Lubin b #OTD 1896 (d 6 Jul 1978) ASA Fellow 1944; ASA President 1946. Commissioner Bureau of Labor Statistics 1933–1946. Statistical adviser to FDR & member of the "Brain Trust" he was a major architect of the Marshall Plan, the US foreign aid program for rebuilding Europe after WWII

Doomed to mediocrity or regression to the mean? In June 1934 Harold Hotelling (ASA Fellow, 1937) writes a polite but devasting critique on the 1933 book The Triumph of Mediocrity in Business by Horace Secrist (ASA Fellow, 1918). 1/2

Alos #OTD 1934 Michael Hills b (d 7 Jan 2021) Coauthor with David Clayton of the classic text 'Statistical Models in Epidemiology' he worked at the London School for Hygiene and Tropical Medicine & was Head of Biometrics and Computing, Natural History Museum

Charlotte Kipling b #OTD 1919 (d 9 Aug 1992) UK ichthyologist. RSS Fellow. Pioneered statistical methods for long-term population studies of freshwater fish (Lake Windermere Perch and Pike Project), showing how fish populations react to habitat change and overfishing

#OTD 1900 Samuel A Stouffer b (d 24 Aug 1960) ASA Fellow 1940. Developed a method combining p-values from independent statistical tests to determine if the joint null should be rejected, a compromise between methods devised by Fisher & Egon Pearson 1/2

#OTD 1833 17-yr old Ada Byron (later Ada King Countess of Lovelace) meets Charles Babbage for the first time. She was introduced by her tutor Mary Somerville, a friend of Babbage. They corresponded for many years: the British Library holds a collection of letters dating between 1836-1851. 1/2

In June 1949 Maurice Kendall (ASA Fellow, 1950 RSS President 1960) uses the term “frequentist” for the first time, referring to differences in the statistical approach of Venn, Neyman, and von Mises and the non-frequentist Bayesians represented by Harold Jeffreys

Raymond Pearl b#OTD 1879 (d 17 Nov 1940), ASA Fellow 1920; ASA President 1939; Chief Statistical Division US Food Administration 1917–1919. One of the first to combine mathematics and statistics to interpret population genetics, he is best known for the Pearl-Reed-Verhulst logistic growth model.

Charles Davenport b #OTD 1866 (d 18 Feb 1944). ASA Fellow 1921. Introduced Pearson's biometric methods to America. A keen eugenicist, his statistical 'justifications' for identifying & removing 'degenerates' influenced US immigration policy for decades & was a model for Nazi extermination programs.

#OTD 1780 Pierre Laplace reads his "Mémoire sur le calcul aux suites appliqué aux probabilités" to the Royal Academy of Sciences describing his theory of inverse probability, “one of the most significant [papers] in development of mathematical statistics”. (In 1950 Fisher renamed it "Bayesian”)

Happy 🎂Beatrice Cave-Browne-Cave b #OTD 1874 (d 9 Jul 1947) She first worked with Karl Pearson at the Galton Lab, then in 1916 joined the Admiralty to perform pioneering aeronautical research for which she later received an MBE. Pearson was incensed when she left... 1/4

Louis Leon Thurstone b #OTD 1887 (d 29 Sep 1955) ASA Fellow 1949. A pioneer in psychometrics, he is best known for developing standardized mean and SD of IQ scores as an alternative to the Binet test system & for his contributions to factor analysis (designed to model intelligence)

#OTD 1923 William Seeley Gosset (‘Student’) reads his paper ‘On Testing Varieties of Cereals’ to the Society of Biometricians and Mathematical Statisticians. The paper is published Dec 1923 & presents the first formal analysis of variance table in a footnote

Happy belated 🎂 to Sir William Petty (26 May 1623-16 Dec 1687) FRS, physician, professor of Music, scientist, economist. His “Political Arithmetick” was the foundation for modern macroeconomics. With John Graunt he developed early census methods & the first published books on vital statistics. 1/6

Happy🎂to Abraham de Moivre b #OTD 1677(d 27 Nov 1754) FRS. One of the pioneers of classical probability theory, he is best known for his large-sample approximation to the binomial by the normal distribution & the first occurrence of the normal probability integral 1/5 🧵

From the Department of “Statistics/Maths Terms That Fortunately Never Caught On”: The Welsh physician Robert Recorde (c. 1510 – 1558), inventor of the equals sign =, was sometimes less apt at coining mathematical terminology. Like 'zenzizenzizenzike ' 1/4

Gauge Repeatability & Reproducibility was developed at General Motors in the 1960s for statistical quantification of precision, variation & measurement error. However, Daniel Gabriel Fahrenheit b #OTD 1686, pioneer of exact thermometry, had the same basic idea almost 300 years before. 1/6 🧵

There would be no Big Data without Really Big Numbers. In 'Le Triparty en la Science des Nombres’ Nicolas Chuquet (1445-1488) developed his system of exponents for powers of one million (‘byllions’ ‘tryllions’). Alert readers will notice that he got a bit muddled with numbers this size...

Accountant as statistician: #OTD 1875 George O. May b (d 25 May 1961) ASA Fellow 1941. Senior partner at Price Waterhouse, AICPA Gold Medal 1944, Accounting Hall of Fame 1950 for key developments in stronger corporate auditing & accounting principles.

#OTD 1728 Gabriel Cramer first uses the term ‘expectation’ ["l'espérance mathématique"] in its modern mathematical sense. Later Laplace runs with it in his ‘Théorie Analytique des Probabilités’. In 1838 'expectation’ first appears in English in Augustus de Morgan's ‘Essay on Probabilities’

Mindel Sheps b #OTD 1913 (d 13 Jan 1973) ASA Fellow 1970. Physician, political healthcare activist, biostatistics professor, advisor to WHO, she made substantial contributions to mathematical demography but is best known for her highly influential NEJM 1958 paper on the asymmetry of relative risk.

The first formal use of random subject allocation is credited to Austin Bradford Hill & the 1947 streptomycin trial. However, he was scooped by ~2000 yr. In ancient Athens the klèrôtèrion was a randomized allotment machine for fair & secure selection of candidates for juries and public offices 1/3