Geodemographics: creating a classification at the level of the individual

This research challenges the existing geodemographics ethos by investigating the benefit to be gained from a move away from conventional areal unit categorisation to systems capable of classifying at the individual level. This research will present a unique framework through which classifications can be developed at this level of resolution. Inherently methodological, a local classification for Leeds (UK) will be presented plus further examples of this applied framework. Issues such as ecological fallacy, Modifiable Areal Unit Problem and generalisation are aspects to be considered when interpreting spatially aggregated data. A move away from such problems is one of the central objectives of this research. Data variables from the UK’s 2001 Small Area Microdata file underpin this research. These variables undergo transformation from categorical states into scale variables based on gross monthly income data present in the British Household Panel Survey therefore enabling effective clustering. Micro-simulation is then employed to create an individual-level population. The framework presented comprises entirely census variables but also demonstrates a linkage capability to other non-census datasets, such as the British Household Panel Survey (now Understanding Society), for deeper profiling, classification validation and enrichment.