Biostatistics

Abstract

After our Faculty Seminar at the School of Public Health had continued for a few years, Dean Howard Hiatt decided to strengthen the quantitative side of his health services effort. Biostatistics had had a serious presence since the School was established in 1922. The founding chair of the department was E. B. Wilson, a mathematician who had earlier been chair of the Department of Physics at MIT. After he retired from Harvard, he was the project supervisor for Bill Cochran’s research contract with the Office of Naval Research, and I got to know him more professionally than I had through the American Academy of Arts and Sciences, where he had been the Executive Officer. In the course of his visits with Cochran, he and I talked about a research problem in factor analysis whose results I still have not published. He checked over my approach and used it on a practical problem he was helping someone with.

Wilson was polite and formal, rather in the way people are sometimes portrayed as being at the turn of the century or earlier. In spite of this, I found him easy and enjoyable to talk with. In a 1927 paper in theJournal of the American Statistical Association, he had developed the idea of confidence limits for the binomial situation, possibly before others had done so. When I mentioned this to him, he made no claim to priority. He seemed unduly modest about it.