Principles of automated zone design
Presenter(s): David Martin & Sam Cockings
This series of short videos provides an overview of the principles and techniques of automated zone design. The videos introduce both the use of automated zone design in the creation of contemporary official statistics and techniques for researchers wanting to implement their own zone design.
An overview of zones and zone design
In this video David explains the importance of zone design, firstly by looking at the design features in some contemporary census mapping of London, then turning to consider the implications of zone design more broadly in terms of the Modifiable Areal Unit Problem. The impact of zone design is illustrated with reference to the placement of electoral boundaries to achieve different results, known as Gerrymandering, and also by consideration of the way in which statistical relationships observed in geographically aggregated data may affected by zone design considerations.
Principles of automated zone design methods
This video introduces the principles of automated design methods, explaining the principles of zone design by iterative recombination of building block zones in order to optimise specific design crtieria. The importance of the specific building block zones and the specification and measurement of zone design criteria are explained.
Introduction to AZTool software
This video introduces the operation of the AZTool zone design software, illustrating the input files, program control by use of the parameter file, and output files produced. The import and export of data to and from Geographic Information System (GIS) software is illustrated and a simple zone design scenario is described using a test dataset.
Examples of automated zone design in practice
This video outlines the major UK statistical zones and associated datasets which have been created using automated zone design tools during the period 2001 to 2016 many of which are associated with the 2001 and 2011 censuses in England, Wales and Northern Ireland. The implications of the zone design process are highlighted in order that researchers intending to use these data should fully understand how they have been generated.
About the author
David Martin is currently Professor of Geography at the University of Southampton. He was Coordinator of ESRC's Census Programme from 2002-12 and is currently involved in the leadership of three major ESRC initiatives: the UK Data Service, the Administrative Data Research Centre for England and the National Centre for Research Methods. Since 1996 he has worked closely with the Office for National Statistics (ONS), the National Statistical Organization for England and Wales, having developed algorithms and software for the automated creation of census output areas which was adopted as the basis for all 2001 small area census outputs by ONS and the Northern Ireland Statistics and Research Agency (NISRA). These techniques were further developed as the geographical basis for the Neighbourhood Statistics System used for the government?s Indices of Deprivation 2004-15 and ONS statistical geography more broadly. The same methodology was implemented for 2011 census outputs and further developed to create a new UK-wide system of workplace zones for the publication of census workplace data. David's other research interests include spatiotemporal population modelling and statistical disclosure control of geographically referenced data.
- Published on: 25th April, 2016
- Event hosted by: Co-Director, National Centre for Research Methods
- Keywords: small areas | automated zone design | disclosure control | geographical referencing | census | population |
- To cite this resource:
David Martin & Sam Cockings. (2021) Principles of automated zone design. National Centre for Research Methods online learning resource. Available at https://www.ncrm.ac.uk/resources/online/all/?main&id=20416 [accessed: December 8, 2021]