A statistician from the University of Chicago will join the Statistics Department, marking a key step in the expansion of one of Harvard's smallest departments.
The appointment of Donald B Rubin, a leading social science statistician, will increase the department's senior faculty to four.
He is only the second professor to receive tenure in the department since 1961, and the fifth in the department's 26-year history.
The arrival of Rubin is intended to rejuvenate the department's ties with the social sciences. In addition, professors said yesterday, the appointment is part of the long-term growth of the Statistics Department.
"We have argued again and again with the University that it is difficult to operate a successful Ph.D. program with a small department." Professor of Statistics Peter J. Huber said, adding. "The arguments are finally proving successful."
According to Frederick Mosteller, professor of Mathematical Statistics and the founder of the department, the Statistics Department needs to fill positions in mathematical statistics.
Computers are another key area, said Huber, adding that there are minor disagreements in the department as to priorities.
Because of today's strong job market, expansion of the department is going to be difficult, said Department Chairman Arthur P. Dempster.
Rubin's appointment "will offer an out-reach in the social sciences that we haven't had in the past few years," said Mosteller, who has worked primarily for the School of Public Health since 1976.
The departure of Mosteller and the retirement in 1976 of Professor of Statistics William Cochran weakened the department's links to social sciences.
William Kruskal, dean of social sciences at the University of Chicago, said yesterday that Chicago made several special offers in an attempt to retain Rubin, but to no avail.
"Sometimes the ball bounces one way, sometimes the other," he said, adding. "At that level of excellence, you can't talk about replacements."
Rubin, a Princeton graduate, said he feels "a fondness and an attraction for the department where I did my graduate studies."
"Because the department is so small, it should be a challenge to work there," he added.
Rubin taught at Harvard for a year in 1971, then left for the Educational Testing Service, where he served as chairman of the statistics group.
Rubin specializes in statistical applications, and is currently analyzing "missing values in census public-use files."
"Perhaps he alone of living statisticians has a good concept of this issue," said Steven M Strigler, professor of Statistics at the University of Chicago.
In 1981 Rubin was tenured at Chicago Harvard was interested in hiring him [in 1981] but we couldn't get our act together in time. Dempster said. "We finally got our act together.