#OTD 1851 Jacques Bertillon b (d 4 July 1922) 🇫🇷Physician, statistician demographer. Introduced the Bertillon Classification of Causes of Death, precursor of the International Statistical Classification of Diseases (ICD) /4
#OTD 1945 John Gaddum (1900-1965) 🇬🇧 professor of pharmacology, coined the term 'lognormal' when the “distribution of log x is normal". In 1879 Donald MacAlister🏴had shown log transformation of non-normally distributed data could result in a normal distribution 1/2
Oct 1906 was a busy month for Karl Pearson publication-wise. He had 7 papers (6 single-authored) plus an obituary tribute to Frank Weldon in Biometrika. Of course it was 'his' journal. Topics ranged from correlation of intelligence with skull size (negligible), 1/2
#OTD 1924 John A Nelder b (d 7 Aug 2010) FRS 1976, Guy Silver 1977 & Gold Medal 2005; @RoyalStatSoc President 1985. Best known for contributions to design of experiments, esp block designs & generalized linear models 1/6
In Sept 1949 Monte Carlo sampling method was formally introduced for solving functions without known solutions by computer-intensive probability-based methods. Stanislaw Ulam (1909-1984) got the idea from playing solitaire during recovery from surgery. Nicolas Metropolis named it
#OTD 1759 William Playfair b (d 11 Feb 1823) 🏴 “One of the leading 19th cent British statisticians” Founder of statistical graphics methods & #dataviz, & inventor of several including line, area, pie, & bar charts
#OTD Constance van Eeden d (b 27 Apr 1927) 🇳🇱🇨🇦@AmStatNews Fellow 1973, @SSC_stat SSC Gold Medal. Best known for important work on nonparametric statistics, maximum likelihood estimation, robust statistics, & parameter spaces
#OTD 1915 Olive Jean Dunn b (d 12 Jan 2008) USA @AmStatNews Fellow 1968 @AAAS 1965, Professor Biostatistics UCLA; 1974 UCLA Woman of Science. Best known for developing the Bonferroni' (or Dunn) correction for constructing simultaneous confidence intervals 1961.
#OTD 1890 Louis-Dominique-Jules Gavarret d (b 28 Jan 1809) 🇫🇷An early pioneer of the 'statistical method' in medicine, his book 'Principes de Statistique Médicale' (1840) is the one of the first to outline statistical principles for rigorous clinical trial conduct. 1/11🧵
In an unusual departure for @AmStatPresident addresses, in Aug 1971 Churchill Eisenhart (@AmStatNews Fellow 1943, 66th ASA President 1971) gave a comprehensive historical overview of a central problem of statistical inference, the origin of the mean. /2
OTD 1921 Charles Dunnett b (d 18 May 2007) MBE 🇨🇦President Statistical Society of Canada @SSC_stat 1982, SSC Gold Medal 1986, @AmStatNews Fellow 1965. Best known for Dunnett’s test for multiple comparisons. He has 2 papers among the top 25 most-cited statistics papers /2
In 1926 RA Fisher introduces the idea of confounding in experimental designs. However, in 1843 John Stuart Mill was actually first to define the concept for experiments in his book System of Logic /3
In Aug 1955 the term “artificial intelligence” is coined. It shows up in a proposal for a summer workshop submitted by Claude Shannon (Bell Labs), John McCarthy (Dartmouth), Marvin Minsky (Harvard), & Nathaniel Rochester (IBM). The workshop took place in July and August 1956
#OTD 1960 Carlo Bonferroni d (b 28 Jan 1892)🇮🇹 Although best known for the Bonferroni adjustment for family-wise error rate, the method was actually developed by Olive Dunn. /4
In Aug 1934 ‘likelihood’ was first used in a Bayesian context by Harold Jeffreys. He wrote what he called the "the theorem of Inverse Probability" as the Posterior Probability proportional to (Prior Probability × Likelihood)
Jeffreys H 1934. Proc Royal Soc Lond A, 146, 9-16
Classic data sets: To demonstrate linear hierarchical models, Cochran & Cox (1957) used data from Frances Cook’s 1938 thesis work: a 3 x 6 trial on 'breakage angle' of chocolate cake. There were 15 replicates, which is a lot of cake. /4
#OTD 1895 Egon Pearson b (d 12 Jun 1980) 🇬🇧 FRS, CBE, Guy Gold Medal 1955, President @RoyalStatSoc 1955. Hon member ASA @AmStatNews. Best known for his collaboration with Jerzy Neyman & development of the Neyman-Pearson theory 1933 for statistical hypothesis testing /6🧵