George Udny Yule

February 18, 1871, on a small family farm in Morham, Scotland, Sir George Udny Yule and his wife, Henrietta Peach, gave birth to an ambitious little boy who would ultimately take many first steps in paving the way for the development of probability-based statistical methods. This little boy's name name was George Udny Yule.

Yule's parents allowed George to have an extensive childhood education. This education began first in a day school, and then he was sent to a preparatory school near Rugby. After prep school, Yule was off to the University where he would spend most of his life working and studying in the mathematics discipline.

Yule's university days as a student were spent at Winchester College and University College, London, where he studied civil engineering. He was unable to get a degree in this field because there were none available in the subject at the time. After turning away from civil engineering and leaving University College, Yule spent a year working with Heinrich Hertz. The investigations Yule did with Hertz on the passage of electric waves through dialects lead to Yule's first published paper. This was just the first of many of Yule's accomplishments.

In 1893, Yule again made a career change when he accepted a demonstratorshiip position under Karl Pearson. Among other things, some of his main duties under Pearson were to draw up class plans, prepare specimen diagrams, and to critique Pearson's work. After working at this new position for a while, it was clear that Yule himself was beginning to think like a statistician rather than a civil engineer.

As a statistician, Yule contributed significant examples to Pearson's studies before going to work on his own. One of Yule's primary contributions to Pearson's work dealt with Pearson's work on skew curves. Yule looked specifically at how to apply skew curves to pauperism. This application provided instruction to the use of Pearson's technique, and it opened his work up to a much larger audience. Yule also made Pearson's work more available to people by clarifying and breaking down some of Pearson's obscure points.

Yule separated from Pearson in 1897 when he accepted the position as the Assistant Professor of Applied Mathematics. Yule's work had a different focus than Pearson's. Whereas Pearson's approach in his work focused more on biology, Yule tried to apply statistical methods to deal with problems of the social sciences.

Three years after accepting this new Assistant Professor position, Yule married May Winifred Cummings, but their marriage was annulled in 1912. As a result of the annulment, Yule felt obligated to make more money. This obligation lead him to accept higher paying jobs which he didn't really enjoy. Although Yule didn't particularly enjoy some of his positions, he always found time to lecture on topics he enjoyed during the evenings. Most of his early lectures focused on topics which lead to his first book, An Introduction to the Theory of Statistics. This book was essential to the people of the time because it was the only comprehensive textbook on the subject until over 45 years later.

Yule made many career changes as a statistician. He held many positions in addition to those already mentioned. Yule left each of these positions for various reasons, but no matter where he was working or who he was working with, Yule was always accomplishing something that was significant. Some of Yule's major achievements in statistical theory deal with regression and correlation, association, time series, Mendelian inheritance, and epidemiology.

One major contribution Yule made began in 1912. That year Yule worked with Major Greenwood on the foundations of theory of accident distributions. In 1915 Greenwood and Yule published a paper in which they were two of the first people to study anti-typhoid and anti-cholera inoculations. Many of their results and much of their original work are still used today. Their paper to see if there was a connection between statistical and practical problems in infectious diseases.

Another contribution Yule and Greenwood made was their equation to figure out vaccine efficacy. This was significant in that it helped test vaccines more quickly so they could be used to treat the diseases that were so common in the early 1900's.

Some of the work Yule did on his own was very significant also. Yule was one of the first statisticians to work on unusual correlations. In 1926, Yule gave one of the first cohesive mathematical treatments of spurious correlations. He took a look at two variables which showed a very high correlation. He commented on the high correlation between the fall in proportion of Church of England marriages and the fall in mortality. Both the number of marriages and the mortality rate were affected by the Progress of Science since 1866. Because both variables were influenced by a common factory, it is reasonable to expect that they would be highly correlated.

See CHART!!!!

Yule concluded, however, that the correlation between the two is sheer nonsense and its meaning has no significance at all. Therefore the two variables are not in any sort of way, causally related to one another. This supports the fact that correlation does not imply causality.

Yule's work on regression was specifically built from Galton's original work. He further developed Galton's ideas about multivariate data. Although his formulation and use of regression analysis was not a full solution, it was a masterly step towards one.

Besides branching off from Galton's work, Yule's work was also a direct decent of Edgeworth and Pearson. Although Yule initially began his profession as a statistician under Pearson, the two mathematicians never saw eye-to-eye on a number of topics. One long controversy between the two resulted from Yule presenting the coefficient of association for the measurement of the degree of association in 2x2 contingency tables. This big disagreement created a bitter friendship between Yule and Pearson.

Yule was just one of a series of remarkable men who introduced many new ideas into statistical theory. After a near life-long commitment to mathematics, Yule settled quietly into retirement. A serious heart ailment may have been what lead to his death on June 26, 1951 in Cambridge, England. Without Yule's devotion and hard work, the use of probability-based statistical methods in the social sciences as well as in other disciplines would not be as advanced as it is today.