Doug Altman and Martin Bland jointly published more than 100 articles, letters, and notes, included one of the 30 most highly cited papers ever. Here I describe how this came about, as a result of mutual interest in the quality of medical research.

The statistical question was how to quantify agreements between two different methods of measuring the same thing. The usual approach in the medical literature was to estimate the product moment correlation. We pointed out that correlation does not measure agreement; two methods of measurement could theoretically give results where one was ten times the other but the correlation was perfect. We suggested a different approach based on the distribution of the differences between measurements by the two methods. This included a graphical check on the assumptions, which users found to be a useful way of presenting the agreement between the methods of measurement.

Bio: Martin Bland was born 6 March 1947, Stockport, England. He attended Imperial College of Science and Technology, University of London, 1965-9, where he was awarded a B.Sc. in Mathematics, and an M.Sc., Mathematical Statistics and Operational Research, Imperial College, London. In 1980 he was awarded a Ph.D. in Epidemiology in the Faculty of Medicine, University of London. He worked in agricultural research for what is now Astra Zeneca, then moved to public health and epidemiology, first at St. Thomas’ Hospital Medical School, then St. George’s, both University of London, then at the University of York, from which he retired in 2015 and became Emeritus Professor of Health Statistics. In 1975, he spent two terms as Visiting Lecturer in Biostatistics, Faculty of Medicine, University of Dar Es Salaam, Tanzania. He has published two textbooks, An Introduction to Medical Statistics and Statistical Questions in Evidence-Based Medicine, both Oxford University Press.