Use of registers and population censuses as sampling frames in Great Britain

Article type: Research Article

Authors: Barnes, Robert¹

Note: [1] Robert Barnes is a graduate in economics from the University of Leeds. He is currently head of the Social Survey Division of the United Kingdom Office of Population Censuses and Surveys. He began work in survey research in 1963 and has worked both in Australia and the United Kingdom. Prior to his present position he was for two years head of the Census Division of OPCS.

Abstract: In Great Britain a wide range of government social surveys, using samples of households and individuals, are carried out on both a continuous and on an ad hoc basis. The great majority of these surveys use registers of various kinds as sampling frames. Some of these registers are intended to cover all addresses in the country in which there are private households. Other registers relate to specific sub-groups of the population. Each of the general registers has advantages and disadvantages but none is ideal for sampling purposes. In recent years nearly all samples required of the general population have been drawn, by automated means, from the Postcode Address File, although the balance of advantages of this system need to be continually monitored and evaluated. The decennial population census is not often used as a sampling frame, partly because it soon becomes out of date and partly because there are some public anxieties over its use for this purpose.

DOI: 10.3233/SJU-1986-4107

Journal: Statistical Journal of the United Nations Economic Commission for Europe, vol. 4, no. 1, pp. 81-92, 1986