Standard indexes of poverty and deprivation are rarely sensitive to how the causes and consequences of deprivation have different impacts depending upon where a person lives. More geographically minded approaches are alert to spatial variations but are also difficult to compute using desktop PCs.The aim of the ESRC sponsored project was to develop a method of spatial analysis known as 'geographically weighted regression' (GWR) to run in the high power computing environment offered by 'Grid computation' and e-social science. GWR, like many other methods of spatial analysis, is characterised by multiple repeat testing as the data are divided into geographical regions and also randomly redistributed many times to simulate the likelihood that the results obtained from the analysis are actually due to chance. Each of these tests requires computer time so, given a large dataset such as the UK Census statistics, running the analysis on a standard machine can take a long time! Fortunately, the computational grid is not standard but offers the possibility to speed up the process by running GWR's sequences of calibration, analysis and non-parametric simulation in parallel.An output is a model of the geographically varying correlates of car non-ownership fitted for the 165,665 Census Output Areas in England. Specifically, a geographically weighted regression of the relationship between the proportion of households without a car (or van) in 2001 (the dependent variable), and the following predictor variables: proportion of persons of working age unemployed; proportion of households in public housing; proportion of households that are lone parent households; proportion of persons 16 or above that are single; and proportion of persons that are white British.
