Remembering Oscar Kempthorne (1919-2000)

On November 15, 2000 the statistics community was saddened by the death of one of its most prominent members and leaders, Oscar Kempthorne, who had given over 50 years of his life to statistical science as an educator and researcher. Obituaries and other accounts detailing aspects of and achievements during his personal and professional life have appeared elsewhere (IMS Bulletin 30 (2), 2001; Bancroft, 1984; David, 1984). The purpose of this paper is different: it is to highlight his major contributions to statistical science, and to indicate how these ideas are still guiding statistical thinking today. Oscar Kempthorne contributed largely to three major areas: to experimental design, to genetic statistics, and to the philosophy and foundations of statistics. These seem to be rather distinct areas, but his research shows a common thread in the form of his concern for acquiring scientifically sound data and interpreting such data. In this context he considered the analysis of variance as one of the most powerful statistical techniques, and it is therefore not surprising that much of his research, certainly in experimental design and genetic statistics, centers around this technique. This work established him very early on as one of the leading statisticians of our time.