Egon Sharpe Pearson


Quick Info

Born
11 August 1895
Hampstead (near London), England
Died
12 June 1980
Midhurst, Sussex, England

Summary
Egon Pearson was an English statistician who worked on statistical hypothesis testing. He was the son of Karl Pearson.

Biography

Egon Pearson's father was Karl Pearson, whose biography is given in this archive, and his mother was Maria Sharpe. Egon was the middle child of three in the family; Sigrid Loetitia was three years older and Helga Sharpe three years younger than Egon. Even as a child at the age of five he was aware of his father's work and his efforts to bring the journal Biometrika into existence. Later in life Egon recalled creating his own journal as a five years old:-
... which was all scrawls with chalk.
He attended school, first at Dragon School Oxford from 1907 to 1909, then going to Winchester College from which he graduated in 1914. He had been accepted to study at Trinity College Cambridge in June of that year.

World War I began in 1914 before he was due to matriculate at Cambridge, and had Pearson's health been good he would have found himself in military service. However, his health had never been good as a child and he had a heart murmur which now prevented him from enlisting. He therefore went to Trinity College, Cambridge, to begin his university studies. These studies were interrupted by influenza which hit him hard and he was unable to study from August 1914 until the end of that year. At the end of one year of study, Pearson left Cambridge in 1915 determined to make a contribution to the war effort, and he worked for the Admiralty and the Ministry of Shipping.

Pearson never took up his undergraduate studies at Cambridge again after the war but was awarded his B.A. in 1920 after taking the Military Special Examination in 1919 which had been set up to cope with those who had their studies disrupted by the war. He then began research at Cambridge, but not on statistics as one might have expected, rather on solar physics. However, even the astronomy lectures he attended, which were given by Eddington, involved him in statistics since Eddington was lecturing on the theory of errors. Pearson also attended astronomy lectures by F J M Stratton and undertook work with F L Engledow and G U Yule.

In 1921 Pearson joined his father's Department of Applied Statistics at University College London as a lecturer. However, despite being a lecturer, his father seems to have kept him away from lecturing. Instead Pearson attended all of his father's lectures and began to produce a stream of high quality research publications on statistics. In 1924 Pearson became an assistant editor of Biometrika but perhaps one of the most important events for his future research happened in the following year.

Neyman had a two year Rockefeller Research Fellowship for the years 1925-27. He spent the first of these two years in University College, London, and the second of them in Paris. When Neyman met Pearson in his father's Department in 1925 he did not realise that he was going through a sort of crisis. Karl Pearson's work had been under attack from R A Fisher for a number of years and Egon later explained (see for example [3]):-
... I had to go through the painful stage of realising that K.P. [his father] could be wrong ... and I was torn between conflicting emotions:
  1. finding it difficult to understand R.A.F. [Fisher],
  2. hating him for his attacks on my paternal 'god',
  3. realising that in some things at least he was right.
In [3] the friendship that developed between Neyman and Pearson during 1926 is described. It paints a picture of Pearson, and his difficulties, at this time:-
Pearson was an introverted young man who felt inferior for a number of reasons. He had grown up in the great shadow of K.P. [Karl Pearson], "lovingly protected" in his childhood and kept out of the war in his youth. At Cambridge he had felt cut off from classmates of his own generation, all veterans of the conflict. He suspected that K.P. was disappointed in him, for he had not gone on to his second mathematics tripos but had taken his degree on the basis of work he had done during the war. After joining the staff of K.P.'s laboratory, he had continued to live at home and to have almost all his social contacts with relatives.
It is perhaps worth noting that 1926 was the year when Karl Pearson allowed his son to begin lecturing at University College. Even then it took place simply because Karl Pearson's health prevented him teaching, rather than for positive reasons.

Pearson and Neyman agreed to undertake a joint research project in June 1926, just before Neyman left for Paris. Their joint research was carried on by letters, but there were meetings such as in the spring of 1927 when Pearson visited Neyman in Paris. During the visit they mapped out their first joint paper and planned their future research. Neyman describes their collaboration in [9]:-
The initiative for cooperative studies was Egon's. Also, at least during the early stages, he was the leader. Our cooperative work was conducted through voluminous correspondence and at sporadic get-togethers, some in England, others in France and some others in Poland. This cooperation continued over the decade 1928-38.
Neyman also describes in [9] the aims of their work:-
My joint work with Egon was concerned with problems of testing hypotheses. One aspect of this work was philosophical ... The decade-long joint work of Egon and myself was aimed at building a mathematical theory of tests the use of which could minimise the frequency of erroneous conclusions regarding the hypotheses considered ...
As H A David writes in [6]:-
The Neyman-Pearson theory of testing statistical hypotheses has become an integral part of every statistician's education and vocabulary.
There was a second major correspondence which Pearson carried out over much the same period as the Neyman-Pearson collaboration. This was one which Pearson carried out with Gosset, and Gosset's ideas played a big role in the discussions between Pearson and Neyman.

Pearson visited the United States in 1931 and, in addition to lecturing in Iowa, he held discussions with Shewhart in the Bell Telephone Laboratories in New York. The following year Shewhart visited Pearson in London and their discussions on quality control in industry led to the creation of the Industrial and Agricultural Research Section of the Royal Statistical Society.

In 1933 Karl Pearson retired from the Galton Chair of Statistics which he had held in University College London. Against his wishes the University authorities decided to split the Department into two separate departments; the Galton Chair and Head of the Department of Eugenics went to Fisher, while Egon Pearson was appointed Reader and became Head of the Department of Applied Statistics. If Karl Pearson did not like that arrangement, then certainly neither did Fisher.

Neyman came to work in Pearson's Department in 1934 and in that same year Pearson married Eileen Jolly, the daughter of a solicitor. Egon and Eileen Pearson had two daughters. Family commitments, further administrative duties and assuming the role of Managing Editor of Biometrika on his father's death in 1936 all reduced the time that Pearson could devote to research. He had been awarded the Weldon prize and medal in 1935, mainly for his work with Neyman, but despite having Neyman as a colleague, Pearson's efforts began to be directed towards revising his father's two volume work Tables for Statisticians and Biometricians. He worked on this project with H O Hartley, but it was some considerable time before the revision was completed; volume 1 appeared in 1954 and volume 2 in 1972. H A David [6] writes:-
With their attractive layout, easy means for interpolation, and extensive, helpful introductory material, these tables have been widely recognised as models of their kind.
The advent of World War II in 1939 led to a shift in Pearson's work. Neyman had already left London in 1938 for a post in Berkeley and with the outbreak of war Pearson began to undertake war work for the Ordinance Board undertaking statistical analysis of the fragmentation of shells hitting aircraft and similar work. In fact, Pearson enjoyed this period:-
... despite bombings, V1 flying bombs, and V2 rockets.
In 1946 he was awarded a C.B.E. for his war service.

Pearson had found the years working in the same institution as Fisher very difficult. Karl Pearson had attacked Fisher aggressively and had, in return, been attacked aggressively by Fisher. Egon Pearson had a very different personality from his father and held Fisher's work in high regard. However, even after Karl Pearson died, Fisher kept up his attack on him in print and Egon Pearson must have found it difficult to share a building with Fisher. After Fisher moved away from London in 1939, especially after he began to work at Cambridge from 1943, Pearson must have found the atmosphere in London a much happier one.

He did suffer a great tragedy in 1949 however when his wife Eileen died of pneumonia. He continued to contribute greatly to statistics, however, retiring from University College in 1961, and retiring as Managing Editor of Biometrika in 1966. It was in this year that he was elected a Fellow of the Royal Society, rather belatedly as some of his biographers have noted. In 1967 he married Margaret Theodosia and moved to Cambridge until the death of his second wife in 1975 when he moved to West Lavington near Midhurst in Sussex.

In addition to the honours which we have mentioned above, Pearson was awarded the Gold Medal of the Royal Statistical Society in 1955. He was elected president of the Society in 1955-56.

A former student, then colleague, wrote in [1]:-
... he was always an august, respected and indeed Edwardian, figure. It was only after many years as a junior colleague that one came to a fuller appreciation of his personal qualities. These were consonant with those so much in evidence in his published work: namely; clarity, combined with a studied simplicity, considered judgement and restraint, never pursuing conclusions further than the evidence warranted.
He is described in the Dictionary of National Biography in the following terms:-
Pearson had a quiet disposition, but his shy and rather diffident manner hid an independent and pertinacious spirit which had enabled him to surmount both the controversies surrounding his father Karl and contemporaries like Fisher and Neyman, and some health problems, such as his delicate health when an undergraduate, a heart condition of long standing, and occasional back trouble due to his considerable height.


References (show)

  1. Obituary in The Times
    See THIS LINK
  2. E S Pearson, The selected papers of E S Pearson (Berkeley, Calif., 1966).
  3. C Reid, Neyman (New York, 1997).
  4. M S Bartlett and L H C Tippett, Egon Sharpe Pearson, 1895-1980, Biometrika 68 (1) (1981), 1-11.
  5. N L Johnson and S Kotz, Egon Sharpe Pearson, in N L Johnson and S Kotz (eds.), Leading personalities in statistical sciences (New York, 1997), 146-149.
  6. H A David, Egon S Pearson, 1895-1980, Amer. Statist. 35 (2) (1981) 94-95.
  7. E L Lehmann, The Neyman-Pearson theory after fifty years, in Proceedings of the Berkeley conference in honor of Jerzy Neyman and Jack Kiefer, Berkeley, Calif., 1983 I (Belmont, Calif., 1985), 1-14.
  8. P G Moore, A tribute to Egon Sharpe Pearson, J. Roy. Statist. Soc. Ser. A 138 (2) (1975), 129-130.
  9. J Neyman, Egon S Pearson (August 11, 1895-June 12, 1980) : An appreciation, Ann. Statist. 9 (1) (1981), 1-2.

Additional Resources (show)


Honours (show)

Honours awarded to Egon Pearson

  1. RSS Guy Medal in Gold 1955

Cross-references (show)


Written by J J O'Connor and E F Robertson
Last Update October 2003