To develop a generic statistical framework to enable the optimal combination of complex spatial and temporal data from survey and non-survey sources for the estimation of indicators of the socio-economic performance of urban and regional system. An important principle is that the framework will be generic and not tied to any specific application domain. It will however, take account of the particular characteristics of urban and regional systems.
To ensure, in particular, that this framework can accommodate (i) data sources that provide both direct and indirect information on the relevant population parameters (ii) data at different levels of aggregation (iii) the issues raised by the aging of sample survey data and the consequent need for updating (iv) the effect of sampling and non-sampling errors (including survey non-response and other sources of missing data) and (iv) the opportunities presented by new data streams from IST systems themselves.
The work will comprise two main activities. The first activity will be a wide ranging literature review that will concentrate on identifying the state-of-the-art and the state-of-practice in relation to the combination of data sources as it exists in a number of relevant fields including environmental monitoring, epidemiology and public health, earth observation, geomatics and navigation systems, transport and logistics and economic and social statistics. The review will identify the nature of the data combination problems encountered in these domains, the existing approaches adopted and the empirical performance of these approaches.
This appraisal of the existing literature will be an important input to the second activity, which is the development of the new theoretical framework. We anticipate that this will adopt a broadly Bayesian perspective on the problem.
| Back to Top | Workplan | Home |
Page last updated: 28 February 2008 |