ESDS has moved to the UK Data Service. Users registered with ESDS - what you need to know
| Home | A-Z index | Site map | Contact | Login | Search: 
ESDS Government logo - link to ESDS Government home page

ESDS Government

Introducing small area estimation techniques and applying them to the Health Survey for England using Stata

Wednesday 26 January 2011

Basement Computer Lab, Humanities Bridgeford Street, University of Manchester

Course tutor – Alan Marshall

Pre-requisites – Attendees should have a working knowledge of regression techniques and be familiar with the use of the Stata statistical analysis package.

Many of the surveys supported by ESDS Government are either not available for sub-national areas such as districts and wards or are based on such small sample sizes that local survey estimates are not reliable. For many users of government surveys this represents an information gap that hinders research and policymaking with a local dimension. This workshop introduces statistical methods that combine survey data with other locally available information to generate estimates for geographical areas at which survey data is unavailable or unreliable.

The workshop is structured around a number of practical sessions where participants will gain hands on opportunities of using different models to generate district estimates of disability. The Health Survey for England will be used to provide detailed data on disability and the Census (and other sources) will provide the local information that is combined with the survey data to generate district estimates of disability.

Get the data:

Materials from this HSE workshop, including the unrestricted access teaching dataset, are available for reuse. Just enter your email address below and click on ‘get the data’ .

Your email address:


9.00 Coffee and registration
9.30 Session 1: Introduction
Illustration of the general small area estimation problem and solution
Geography in ESDS Government surveys
Local sources of information
Introducing the Health Survey for England and its disability measures
10.00 Practical 1, Part 1: Generating schedules of disability rates
10.45 Refreshments
11.00 Session 2: Estimating disability – Mathematical representation
Example of the mathematical representation technique
11.30 Practical 1, Part 2: Curve fitting
12.15 Lunch
1.00 Session 3: Relational models theory
Example of the relational model approach
1.30 Practical 2 – Relational models
2.15 Refreshments
2.30 Session 4: Synthetic regression theory
Example of the synthetic regression technique
3.00 Practical 3 - Synthetic regression
3.45 Session 5: Conclusions and possible next steps
Evaluation of the models fitted and their estimates
Which model to use
Assessing the quality of the models
Introducing more complex modelling strategies
4.00 End

ESDS Home Page > Government
© Copyright 2003 - 2014 Universities of Essex and Manchester. All rights reserved.
Contact   |    Copyright and Disclaimer    |    Accessibility    |    Send us comments on this page
Link to University of Essex Link to University of Manchester Link to JISC Link to ESRC