Edwin D. Goldfield, 87, a retired statistician with the U.S. Census Bureau and a leader in the field of census statistics, was found dead Sept. 27 at his home in Temple Hills. Mr. Goldfield, who lived alone, was discovered by friends when he failed to show up for an appointment. According to the Maryland medical examiner's office, he died of cardiovascular disease.

He began his career at the U.S. Census Bureau in 1940 with what he thought would be a temporary appointment to work on the statistical processing of the decennial census. That temporary assignment stretched into to a 35-year government career. During that time, he was program coordinator for the 1950 Census, chief of the bureau's Statistical Reports Division, assistant director of the Census Bureau and chief of its International Statistical Programs Center.

He also served as staff director of the House of Representatives Subcommittee on the Census, as a consultant to the Social Science Research Council and the Mutual Security Agency and as a member of the editorial advisory board of the International Encyclopedia of the Social Sciences. He led several U.S. delegations to international statistical conferences.

Mr. Goldfield was born in Brooklyn, N.Y., and graduated from the City University of New York in 1938, majoring in mathematics, statistics and economics. He became interested in a career as a statistician while working as an intern in the office of New York Mayor Fiorello H. LaGuardia. He received a master's degree in statistics from Columbia University in 1940 and intended to get his doctorate but instead went to work. He did graduate work in statistics and economics at American University from 1940 to 1946.

He took a federal examination to qualify as a junior statistician and received the second-highest grade in the country. He took the exam a second time and did even better.

A faculty adviser at Columbia urged him to move to Washington and take a temporary job at the Census Bureau, which would allow him, as he recalled in an oral history, "to get in on the 'ground floor' and have that experience, which was available once every 10 years." Except for the temporary assignment with the House of Representatives, he stayed with the Census Bureau until his retirement.

During his years at the bureau, Mr. Goldfield was involved with the changeover from manual tabulation of information to computer tabulation, relying initially on vacuum-tube computers the size of a city block, then optical scanning devices and finally digital computers capable of large-scale data processing.

He also helped implement statistical sampling.

After his retirement in 1975, he joined the Committee on National Statistics at the National Academy of Sciences. He served as study director for a panel that produced the report "Privacy and Confidentiality as Factors in Survey Response" (1979), and he also served as director of the committee.

Mr. Goldfield retired again in 1987 but remained active in federal statistics. He maintained an office at the Census Bureau, visited the Committee on National Statistics regularly and directed the Census Alumni Association. He was a past president of the American Statistical Society and a fellow of the American Statistical Association.

There are no immediate survivors.

Edwin D. Goldfield's temporary job at the U.S. Census Bureau in 1940 stretched into a 35-year career.