Cuthbert Daniel: Industrial Statistician

Abstract

Cuthbert died in New York City on August 8, 1997, at age 92. All professions have outstanding individuals. The entertainment business calls them stars. Certainly Cuthbert was a star among statisticians: original in his contributions, unique in personality, well known and popular throughout the profession. If you ever met Cuthbert Daniel, however tangentially, you remembered. His speech was crisp, witty, and occasionally devastating. His seminars were always well attended and he enjoyed teaching. Conversations with his sister, Maude Corser, suggest he was always lively and self-confident. As a young man in high school he found joy in countering conventional dress codes by wearing khaki woolen shirts with white ties. He would not conform. He was then, and clearly remained, a singular person. Two excellent personal interviews of Cuthbert are available: the first, An Interview with Cuithbert Daniel, is a videotape in the Distinguished Statistician Series of the ASA (1987) hosted by J. S. Hunter, sponsored by IBM, and introduced by Herman Friedman. The second interview, Conversation with Cuthbert Daniel by Ed Tufte appears in Statistical Science, (1988). To honor Cuthbert's 80th birthday in 1984 Colin Mallows edited a book of papers contributed by his many friends: Design, Data & Analysis (Mallows 1987). It contains an excellent biography and a bibliography complete to that date. An appreciation of Cuthbert's contributions to applied statistics and the design of experiments evokes the need to describe the world of industrial statistics near the end of WWII. Waging war required masses of quality equipment, and American industry quickly took up the challenge. As part of the war effort W. Edwards Deming, in a letter to W. Allen Wallis at Stanford University in April 1942, suggested establishing a broad educational effort to teach Shewhart methods. Soon an extensive educational program was started and by the end of the war Shewhart charts and Dodge-Romig acceptance sampling procedures were in wide use throughout industry. During the war period very little is found descriptive of the industrial applications of experimental design. L. H. C. Tippett's book Statistical Methods in Industty was published by the British Iron and Steel Federation in 1943, and K. A. Brownlee's pamphlet Indutstrial Experimen?tation, used during the war, was published by His Majesty's Stationery Office in 1946. In the final traumatic year of WWII Walter Shewhart was President of the ASA, the first industrial statistician to so serve. It is interesting to read in his Presidential address, statistics is not simply a tool as is so often stated but a scientific way of looking at the universe: statistical method is not something apart from the scientific method but is the scientific method.. . (The italics are Dr. Shewhart's.) Of course, an essential part of the scientific method is the planning of experiments, and it was here that Cuthbert was about to make his many contributions. An early interest in the sciences led Cuthbert to MIT as an undergraduate. A maverick, he took heavy concentrations in English and history to go along with the usual engineering requirements. He received both a BS and MS in chemical engineering (1925, 1926). An interest in physics led him to the University of Berlin for a year but he soon abandoned physics to return to the U.S. as an instructor at Cambridge School, Kendall Green, MA. Later, for four years, he taught physics and the sciences to teachers in New York City. A concern with good teaching led to an appointment as a research associate in Evaluation of School Broadcasts at Ohio State 1939-1940, and later in New York City at the Office of Radio Research, Columbia University. His very first paper appears in the Journal of Applied Psychology and deals with a problem of comparisons of popular songs. His first book, Radio Listener Panels, co-authored with H. Gaudet, was published by the Federal Radio Education Committee in 1941. In the early 1940s Cuthbert's wife, Janet, a PhD biochemist, happened to bring home R. A. Fisher's Statistical Methods for Research Workers. That event led to a reading of Fisher's Design of Experiments and a career in statistics was catalyzed. He found a focus for his education in mathematics, physics, and engineering. From then forward he be-