ABSTRACT
The most fundamental problem with all cartographic based displays of spatial information is the strong (if not complete) dependency of the results on the nature of the spatial data and the inherent coding, spatial distortion, and generalisation filters that are unavoidably involved. These problems are greatest when socioeconomic data are being mapped because the underlying objects of interest (people, society) are invisible in GIS terms but the spatial representation that is used to make them visible are essentially static physical map features that are really describing and representing something quite different. Furthermore, there is a major generalisation issue. Dorling (1995) explains part of the problem as follows: “Imagine that in your town or village a symbol is painted on every roof top showing the ages, occupations, wealth and political opinions of the people who lived in each home, and that you were given a detailed aerial photograph of the area. It would not take long to see where the most and least affluent areas are and what the people living there
tended to do and how they vote. If, however, you were interested in the whole country or rather than just part of a town, this method would no longer work. The roof signs would not longer be visible…a more subtle picture needs to be created, a picture which is not necessarily directly related to the physical geography of the country” (p. xiv). The question is really what should this picture show and how should it be designed? Indeed this is the most basic and fundamental of all the questions that could be asked.
