An information statistics approach to zone design in the geography of health outcomes and provision
Social scientists and policy makers are usually faced with boundary changes in administrative areas over time. The control of spatial issues deriving from boundary changes is even more important when they affect the organisation and allocation of resources in a national health system. In addition, the problem becomes more acute when the health organisations analyse sensitive data using geographies constructed to serve other administrative purposes. In recent literature, the modifiable nature of areas is reflected in the modifiable areal unit problem (MAUP) and widely acknowledged frameworks for geographical analysis are developed targeting to overcome this problem. The aim of this research is to suggest and develop methodologies supporting the health related studies to provide valuable decisions. In order to achieve this aim the following research objectives have been developed. In this thesis, the crucial objective is to identify how geographical problems are related to health policies exploring available methodologies and suggesting solutions derived from informative statistic measures to unresolved practical issues. Consequently, an automated computer system developed formulating these problems in graph theory context and utilising their components through object oriented algorithms. The test and evaluation of the system is applied in a series of case studies investigating the effects of MAUP in various geographies and aggregation levels. The overall objective provides strategies and valuable practice for using the system as well as suggesting areas of health research that may benefit from the methodology. In the final chapter, the thesis concludes with a summary of findings and limitations for the suggested methodology providing an outline of the research directions for further work into the spatial issues in relation to health research.