Cochran, W. G. (1909–1980)

William Gemmell Cochran was an early specialist in the fields of applied statistics, sample surveys, experimental design, observational studies, and analytic techniques. He was born in Rutherglen, Scotland, to Thomas and Jeannie Cochran on July 15, 1909, and he died on Cape Cod, Massachusetts, on March 29, 1980, at the age of 70. In 1927, Cochran participated in the Glasgow University Bursary competition and took first place, winning enough funds to finance his education. After taking a variety of classes, he was awarded an M.A. in mathematics and physics at the University of Glasgow in 1931. He then received a scholarship for a Cambridge University doctoral program, where he studied mathematics, applied mathematics, and statistics. He began his professional career at the Rothamsted Experimental Station in England after being persuaded by Frank Yates to leave Cambridge prior to the completion of his doctorate. Cochran remained at Rothamsted until 1939, working on experimental designs and sample survey techniques, including a census of woodlands with colleague and mentor Yates. During his years at Rothamsted, Cochran remained in touch with R. A. Fisher and was heavily influenced by Fisherian statistics. In his 5 years at Rothamsted (1934–1939), he published 23 papers. Also during his time at Rothamsted, Cochran met and married Betty I. M. Mitchell.

In 1939 Cochran accepted a post in statistics at Iowa State University, where he taught from 1939 to 1946. His task at Iowa was to develop their graduate program in statistics. During his years at Iowa he both served on and chaired the advisory panel to the U.S. Census and published a number of papers on experimental design. Cochran joined Samuel Wilks and the Statistical Research Group at Princeton University in 1943, examining probabilities of hits in naval warfare and the efficacy of bombing raid strategies. Shortly after World War II, he joined Gertrude Cox at the North Carolina Institute of Statistics, where he assisted in developing graduate programs in statistics. Cochran chaired the Department of Biostatistics at Johns Hopkins University from 1949 until 1957. During this time he authored two books, Sampling Techniques and (in collaboration with Gertrude Cox) Experimental Designs. In 1957 Harvard University established a Department of Statistics and appointed Cochran to head the department. Cochran remained at Harvard until his retirement in 1976.

During his career, Cochran was lauded with many honors. He was the president of the Institute of Mathematical Statistics in 1946, the 48th president of the American Statistical Association in 1953–1954, president of International Biometric Society 1954–1955, [Page 100]and the president of the International Statistical Institute from 1976 to 1981. Cochran was elected honorary fellow of the Royal Statistical Society in 1959, held a Guggenheim Fellowship in 1964, and won the S. S. WiTks medal of the American Statistical Association in 1967. He received honorary doctorate degrees from Johns Hopkins University and the University of Glasgow. From 1974 until his death in 1980, he worked with the National Academy of Sciences' National Research Council panel on incomplete data in sample surveys.

Cochran developed methods for including or excluding an independent variable in multiple linear regression. He also developed the Cochran Q-test, used to evaluate two variables measured on a nominal scale. Cochran was the statistical representative for the U.S. Public Health Service research on the effects of smoking on lung cancer. His work as part of the advisory committee provided the surgeon general with proof that lung cancer was directly related to smoking. He also worked on the Kinsey Report on human sexual behavior, on polio research, and on the effects of radiation on Hiroshima victims. He is well remembered for his many agricultural studies such as the yield of cereals, field counts of diseased plants, and the influence of rainfall.

...