Urban Sub-Areas as Sampling Frameworks for Social Survey

Abstract

The acquisition of information on which decision-making can be based is a continuous need for planning departments and increasingly for other sections of local government such as the social services and education. These data requirements are not new but an increased need reflects both the problems of dealing with a society which has become more complex and the availability of processing techniques which make possible the effective use of large data sources. One basic source of data is the Census which is widely used in both central and local government and there is no doubt lat this constitutes a major single store of information. The main advantages of the Census include the fact that it is comprehensive in coverage, occurs at regular intervals and can be used as a benchmark for planning and research. Despite the importance of the Census, it has drawbacks. Its information is often too outdated to be useful even when it first becomes available and this disadvantage increases the further one is removed from Censustaking dates. Most regional breakdowns of the Census, for example, take at least 2 years to appear and in some circumstances the most recently available data may be up to 7 years out of date. Again, the Census office may diminish the value of data to users by adopting classification procedures which are over-generalized. Demographic details may be submerged in large age-groups whilst the territorial units by which information is recorded may be too gross for detailed physical planning. Another disadvantage is that the Census schedule may not cover relevant topics : it is necessarily limited and stands deficient in terms of several key social indicators. Although there are other official sources of statistics, the fact remains that for many basic information needs the locally-organized survey is the only alternative. Such surveys have become established practice in many local government departments; the case for them has been argued, and possible survey procedures have been described.1 The purpose of this present paper is to add to this list of procedural suggestions and to demonstrate ways in which small area statistics available from the Census can be utilized. Although it has been suggested that there are deficiencies in Census data which make local surveys necessary, the Census nevertheless provides the basic informational context and its value as a benchmark to which other statistics can be related cannot be overstated.