Combining survey and aggregate data
by
David Steel
University of Wollongong

The next major step in official statistical methodology will be the development of general, reliable, robust methods of combing data from different sources with associated indicators of accuracy and reliability and proper inference procedures. Relationships between variables can be analysed using survey data. There is often a large amount of information also available in the form of aggregated data for geographical areas from administrative and other sources that could potentially be used in the analysis. However, statistical analysis based solely on data aggregated over spatial units will often produce results very different from those obtained from analysing the corresponding data for individuals. In this talk methods that make use of the aggregated data, possibly from different sources, and the survey data in the analysis. The roles of the data from each source are considered.

Date received: April 1, 2002