I was saddened to hear of the death of Norm Breslow last week, at the age of 74.
I first met Norm 25 years ago when I started working at the Medical Research Council Biostatistics Unit in Cambridge. Norm came to visit the Biostatistics Unit on sabbatical. He was given a small office just off the library and installed a PC that could be seen churning away for hours on end running programs in Gauss. It turned out that these calculations were for the landmark paper that he wrote with David Clayton, Approximate Inference in Generalized Linear Mixed Models, which has now been cited over 2000 times. Examples from this paper were later re-used by the BUGS project to illustrate MCMC methods and are more commonly known to BUGS users as the Seeds, Epil, Oxford, and Ice examples.
In the field of cancer epidemiology, Norm is most famous for the two volumes of Statistical Methods in Cancer Research that he co-wrote with Nick Day while visiting the International Agency for Research on Cancer (IARC) in the late 1970s. By the time I started in cancer research, these two books, one on case control studies and one on cohort studies, were considered textbooks but Norm was always at pains to point out that they were never intended to be. They were in fact research monographs surveying the state of the art in statistical methodology. It should be no surprise that these books went on to be so successful. They were rigorous, comprehensive, and illustrated with real data from studies conducted at IARC. I have no doubt that they made a major contribution to the scientific profile of IARC in the 1980s.
I myself moved to IARC in the mid 1990s. I continued to see Norm regularly as he remained a friend of IARC and visited several times over the years. He was part of a generation that inspired me when I first started working in biostatistics and it saddens me to think that this generation is now starting to pass away. My condolences to his family.