Estimating Census Health Geographies: using synthetic estimation with secondary survey and census data

With the 2011 Census probably being the last, this project focusses on testing an alternative approach to generating small area data. The aim is to create estimates for the four 'health' indicators presently provided in the census: self-assessed health, limiting long term illness, carer responsibilities, and absence of work due to disability.

The project will develop prior work using small area synthetic estimation based on multilevel modelling. This approach derives statistical models from large-scale surveys and uses the parameters of the models to develop estimates of the target variables. The project will cover the four constituent parts of the United Kingdom and will work with survey data from each country for the data collection period for the 2011 census. The resultant multilevel synthetic estimates for the four health indicators will be compared with the actual data for the indicators collected in the 2011 census. Research will focus on model development and effectiveness as well as the ability of the method to accomodate inter-censal change and spatial dependency.