Mean Squared Error Estimation for Small Area Means, Quantiles and Poverty Indicators: A nonparametric bootstrap approach

This paper describes a group of online services which are designed to support social survey research and the production of statistical results. The 'Grid Enabled Specialist Data Environment' (GESDE) services constitute three related systems which offer facilities to search for, extract and exploit supplementary data and metadata concerned with the measurement and operationalisation of survey variables. The services also offer users the opportunity to deposit and distribute their own supplementary data resources for the benefit of dissemination and replication of the details of their own analysis. The GESDE services focus upon three application areas: specialist data relating to the measurement of occupations; educational qualifications; and ethnicity (including nationality, language, religion, national identity). They identify information resources related to the operationalisation of variables which seek to measure each of these concepts - examples include coding frames, crosswalk and translation files, and standardisation and harmonisation recommendations. These resources constitute important supplementary data which can be usefully exploited in the analysis of survey data. The GESDE services work by collecting together as much of this supplementary data as possible, and making it searchable and retrievable to others. This paper discusses the current features of the GESDE services (which have been designed as part of a wider programme of ‘e-Science’ research in the UK), and considers ongoing challenges in providing effective support for variable-oriented statistical analysis in the social sciences.