John William HOPKINS, 1908-1989

Dr. Hopkins, like Dr. Goulden, started out as a bioscientist. He received an M.Sc. degree in plant biochemistry from the University of Alberta in 1931. His work at the University, more or less as a graduate assistant, involved him in the then current hot topic of Canadian biology: the search for a strain of wheat which is resistant to rust. Having an enquiring mind and a basic bent to look for “the story that the data are trying to convey”, he soon realized that in the field experiments with alternative varieties of wheat the variability between plots was typically greater than that between varieties.

The new book Statistical Methods for Research Workers by a British biologist called R.A. Fisher, was just recently published (First Edition, Oliver & Boyd, Edinburgh, 1925; Fourteenth Edition, 1970, xiii + 362 pp.). John Hopkins became fascinated by the power of the new methodology to deal with his problem. So he went to London and obtained the Ph.D. degree there in 1934 under Fisher. Upon his return to Canada, he joined the National Research Council where he established a fully-fledged statistics laboratory — one of the earliest, perhaps the earliest, in Canada. He organized NRC’s facilities for the analysis of biometric data and the evaluation of experiments. Over the years the fields of application multiplied and became increasingly complex to encompass full-scale models of bio- and ecosystems: these methods allowed NRC to study the sensitivity of individual coefficients and the robustness of the entire models in the presence of sampling and measurement errors.

Dr. Hopkins remained with NRC until his retirement in 1973, except that during World War II he was seconded (one of the very few active statisticians in the government service) to participate in the operations research effort. He served first in Halifax with the Eastern Air Command, then with the Canadian Heavy Bomber Group overseas. In recognition of his outstanding service he became an MBE.

Nowadays, most federal agencies operating in the biosciences have well developed statistical services, but this was certainly not the case in the 1930s, 1940s, or even the early 1950s. As a result, Dr. Hopkins was repeatedly asked to collaborate with other departments when they had problems which appeared to be statistical. In a very real sense, therefore, he was recognized as the dean of the small but gradually growing statistical community. Before the birth of the ASA chapter in Ottawa, the regular seminar series on statistics hosted at NRC and largely organized by Dr. Hopkins served as the major professional communication link among federal statisticians.

Dr. Hopkins was an elected Fellow of the Royal Society of Canada, the American Statistical Association, and the Royal Statistical Society; a member of the International Statistical Institute, and a charter member and officer of the Biometric Society.

In Dr. Hopkins we are honoring a pioneer of applied statistics in Canada.

Ivan P. Fellegi, 1983