Sir Ronald A. Fisher, FRS, 17 February 1890 - 29 July 1962
Ronald Fisher, usually regarded as the "Father of Statistics", was born 125 years ago today in London, one of seven children and the second of twins, but the other was still-born. In 1904, at age 14, Ronald entered a renowned school, Harrow, but it was a difficult time for him as his Mother passed away. Despite this he went on to excel at the school, leading to his matriculation at Cambridge in 1909.
Although he formally studied mathematics and astronomy at Cambridge, he was also keenly interested in biology. He graduated in 1912 with a BA in Astronomy, and continued to focus on George Biddell Airy's manual Theory of Errors. (G.B. Airy was a major contributor to Math and Astronomy.) It was Fisher's interest in such that eventually led him to investigate statistical problems.
After leaving Cambridge, Fisher had no means of financial support and so worked for a few months on a Canadian farm. Returning to London, he took up a post as a "statistician" in the Mercantile and General Investment Company. When war broke out in 1914 he enthusiastically tried to enlist in the army, but his A1 health was done in by his poor eyesight, and so he was rejected. [Perhaps Fisher's myopia was the main factor in his developing a deep multidimensional geometric-space sense.] He became a teacher of mathematics at various schools between 1915 and 1919.
Fisher married, in 1917 at age 27, Ruth Eileen Grattan-Guiness, who was only a few days past her 17th birthday at the time. The Fishers went on to have two sons and seven daughters, one of whom died in infancy and one who lost his life in World War II. (Fisher's daughter Joan married well-known statistician George E.P. Box and wrote a biography of her father, R.A. Fisher: The Life of a Scientist, which was published in 1978.)
Fisher left secondary-school teaching in 1919 when he was offered two positions simultaneously. Karl Pearson offered him the post of Chief Statistician at the Galton Lab and he was also offered the post of Statistician at the Rothamsted Agricultural Experiment Station, which appealed to Fisher's interest in farming. He accepted the latter, also in part because of anticipating conflicts with Pearson. While at RAES, he went on to make many big contributions to both statistics and genetics, such as in the design and analysis of experiments, resulting in the big concepts of randomization and ANOVA. (Such is covered in Chapter 5 of David Salsburg's book The Lady Tasting Tea.)
In 1921, Fisher introduced the concept of likelihood, leading to his maximum-likelihood estimation the following year. His first book, in 1925, Statistical Methods for Research Workers, ran to many editions extending throughout his life, and for many years was the top-selling book in science in the world. (It was translated into French, Italian, Spanish, German, Japanese, Russian, and a few other languages.) As SMRW did not contain derivations, proofs, and other theoretical math, it was accessible to the biologists and others in the scientific community, becoming a standard reference work.
Ten years after the initial publication of SMRW Fisher added another classic reference, the expected The Design of Experiments.
When Pearson retired as Galton Professor at University College London in 1933, Fisher was appointed his successor, with such split in two between Fisher and Pearson's son Egon, who became close to Jerzy Neyman, whose new approach to significance testing was anathema to Fisher.
Ten years after the appointment Fisher went back to Cambridge to become Arthur Balfour Professor of Genetics. He investigated the linkage of genes for different traits and developed methods of multivariate analysis to deal with such.
L.J. Savage recalled, "I occasionally meet geneticists who ask me whether it is true that the great geneticist R.A. Fisher was also an important statistician"! In reviewing Salsburg's book The Lady Tasting Tea [the title referring to a cute experiment by Fisher], Bradley Efron, Professor of Statistics at Stanford and creator of the Bootstrap, writes, "Particularly well told [in the book] is the story of Ronald Fisher, the double genius who founded both mathematical statistics and mathematical genetics. If scientists were judged by their influence on science then Fisher would rank with Einstein and Pauling at the top of the modern ladder."
The accolades started late for Fisher. He was elected Fellow of the Royal Society in 1929, awarded its Royal Medal in 1938, its Darwin Medal in 1948, and its Copley Medal in 1955. In addition, he was elected to the American Academy of Arts and Sciences in 1934 and the U.S. National Academy of Sciences in 1948. He received the Guy Medal from the Royal Statistical Society in 1946. (As Salsburg points out near the beginning of his book, appropriately titled "Fisher Triumphant", Fisher, despite his profound statistical contributions in the 1920s, was in "virtual isolation" then, with the RSS finally first recognizing him in December 1934 by inviting him to present a paper, "an honor reserved for only the most prominent workers in the field". Interestingly, this occurred six months after Neyman presented his paper to the RSS, titled "On the Two Different Aspects of the Representative Method: The Method of Stratified Sampling and The Method of Purposive Selection", which was probably the groundbreaking paper in probability sampling and gave us the term "confidence interval".)
Also, the R.A. Fisher Lectureship is sponsored by the ASA and given at the Joint Statistical Meetings. [William G. Cochran gave the first one in 1972.]
Various institutions awarded him an honorary degree, including Harvard (in 1936), University of Calcutta (1938), University of London (1946), University of Glasgow (1947), University of Adelaide (1959), University of Leeds (1961), and the Indian Statistical Institute (1962).
Fisher was knighted by the Queen in 1952.
Twenty-Five Terms Due to Sir Ronald A. Fisher
Term First Used
variance 1918a
analysis of variance 1918b
likelihood 1921
maximum likelihood 1922a
parameter 1922a
statistic 1922a
consistency 1922a
efficiency 1922a
sufficiency 1922a
degrees of freedom 1922b
Student's t, t 1924
z-distribution 1924
test of significance 1925a
level of significance 1925a
sufficient statistic 1925b
randomization 1926
randomized blocks 1926
confounding 1926
interaction 1926
sampling distribution 1928
covariance 1930
Greco-Latin square (with F. Yates) 1934
null hypothesis 1935
Yates's correction for continuity 1936
normal score (with F. Yates) 1938
Notes: 1. Credit for the above should go to H.A. David, Distinguished Professor Emeritus of Statistics, Iowa State University, from whose two articles in the May1995
and February 1998 issues of The American Statistician the above were compiled.
2. The a and b with a year indicates different papers by Fisher within the year.
3. The "z-distribution" in the 1924 paper is not the normal distribution but likely either the (now-named) F dist. or some pre-F distribution.
[Note: Originally posted in the APStat Community.]
-------------------------------------------
David Bernklau
(David Bee on Internet)
-------------------------------------------
Original Message:
Sent: 02-15-2015 21:43
From: David Bernklau
Subject: Quasquicentennial of Note
This message has been cross posted to the following eGroups: Statistical Education Section and ASA Connect .
-------------------------------------------
Tuesday marks the 125th anniversary of the birth of Sir Ronald A. Fisher (17 February 1890 - 29 July 1962)
-------------------------------------------
David Bernklau
(David Bee on Internet)
-------------------------------------------