Defining Locality Boundaries with Synthetic Data

Abstract

Boundaries can be defined by applying a number of alternative techniques to various types of information; the choices made on these decisions should reflect the purpose for which the boundaries will be used. In this paper I report research to define a set of localities intended to provide an up-to-date and relevant definition of local communities across Britain for the very varied purposes of academic social scientists. First, the multidimensional nature of modern localities are outlined. Three established types of regionalisation procedure are then reviewed, leading to the identification of an appropriate method for analysing a suitable dataset. At the same time, it is concluded that a broadly based set of boundary definitions requires a more innovative approach in order to collate the many strands of evidence which are relevant to locality definitions. The response here is the development of synthetic data which codifies the critical information in any boundary set into a form which can then be combined with other, similar evidence. This leads to the first empirical challenge, which is to collect or create boundary sets which each provide a relevant strand of evidence for locality definitions. I apply the preferred regionalisation method to the synthetic data which has been created, and illustrate the localities which have been defined on this basis. I end by suggesting some other ways in which synthetic data might be analysed to provide insights into patterns of spatial association at the local scale.