Quality Assuring & Adding Value to the 2001 Census Output Area Classification Dan Vickers & Phil Rees Centre for Spatial Analysis and Policy School of Geography University of Leeds Paper presented at the Third International Population Geography Conference Liverpool 19 th -21 st June 2006
19
Embed
Quality Assuring & Adding Value to the 2001 Census Output Area Classification
Quality Assuring & Adding Value to the 2001 Census Output Area Classification. Dan Vickers & Phil Rees Centre for Spatial Analysis and Policy School of Geography University of Leeds. Paper presented at the Third International Population Geography Conference Liverpool 19 th -21 st June 2006. - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Quality Assuring & Adding Value to the 2001 Census Output Area
Classification Dan Vickers & Phil Rees
Centre for Spatial Analysis and PolicySchool of Geography
University of Leeds
Paper presented at the Third International Population Geography Conference
Liverpool 19th-21st June 2006
• Puts the 223,060 Output Areas from the 2001 UK Census into groups of similarity based on their attributes for 41 census variables.
• Output Areas are the smallest geography available from the 2001 Census, comprising from an average of 264 people and 110 households.
What is the Area Classification?
• The classification is available from multiple sources:
• From the ONS at: http://www.statistics.gov.uk/about/methodology_by_theme/area_classification/default.asp
• From my homepage: http://www.geog.leeds.ac.uk/people/d.vickers/OAclassinfo.html
Where is it available?
• The classification is available from multiple sources:
• Casweb http://census.ac.uk/casweb/ Can be added to any downloaded dataset
as a zone attribute.• UK Borders http://edina.ac.uk/ukborders/ Available to download with GIS boundary
files for mapping.
Where is it available?
• GeoView – Incorporation of OAC with Google Earth system, Paul Longley, Mike Batty and Richard Milton UCL.
• Uncertainty, Peter Fisher City University.• Expansion to a further three levels, Martin
Callingham Birkbeck College.• Access to Education Classification, Alex
Singleton UCL.• Changing patterns over time, Dan Vickers
University of Leeds/Sheffield.
What is being done with OAC?
• Camden LB, Tower Hamlets LB, Barking & Dagenham LB, Kensington & Chelsea LB,GLA, Bristol City Council, Derby City Council, Bradford Council, Suffolk CC, West of England local intelligence network, London Strategy Executive.
• London Health Observatory, Thames Valley Strategic Health Authority, Manchester joint Health Unit, North West Health Observatory, Diocese of Coventry.
• Woolworths, E-on (Powergen), Yell.com, Bell Hanson ltd, Pfizer, Abbey, Wegener DM, SAGA, Boots, Hospital Plan Insurance ltd, BusinessKen ltd, The Housing Vision, TNS.
• University of Surrey, UCL, City, UWE, Cornwall College, University of Nottingham.
Who else is using OAC?
• How do we know it’s right, or at least a reasonable representation of reality.
• Little none census data at OA scale to compare classification with.
• Can check against areas which I know well through local knowledge and groundtruthing, however this far from a comprehensive at the UK scale.
The Problem
• Groundtruthing exercises were conducted in locations across the UK.– Visited specific locations taking photos to
record the nature of each area– Photos available via my homepagehttp://www.geog.leeds.ac.uk/people/d.vickers/OAclassinfo.html
• Groundtruthing exercise also conducted using undergraduate students on fieldtrip to Bangor North Wales.
Groundtruthing
• A consultation exercise was devised to fill in the blanks left by the goundtruthing.
• Academics, local government officers and other professionals with a demographic background were invited to take part.
• Participants were asked to give postcode(s) of area(s) they know well e.g. home address, work address, previous home.
User Consultation
User Consultation• Shaded colours
represent each of the 7 super-group types
• Overlaid on OS 1:50,000 mapping
• Red triangle represents specified location
User Consultation
• Colours on the map to be linked to names.
• E.g. A4, B7, C1 etc.
Consultation Results
• 61 people specified 89 locations across the UK.
• Ranging between 4 and 12 locations in each region/country of the UK.
Consultation Results
The results
Where the respondents live
“There’s probably a lecture example on geodemographics and lifestyle classifiers in there
somewhere...” (Respondent qa38).
“I wonder if it would be possible to use this technique as a teaching tool, once the classifications are
finalised? I am sure it would make a great teaching resource for CHCC [Collection of Historical and
Contemporary Census Data and Related Materials] if you could automate the map production once a
postcode is entered?” (Respondent qa1).
Teaching Resource
“An interesting exercise, which tells my own life history - I grew up in 'typical traits'; went off to be a student; as a postgrad I rented a room also in 'typical traits' before as a young academic buying a small terraced house in 'city centre melting pot'. After a few years I moved on to a
house in 'typical traits' and a few years ago finally arrived in 'comfortable suburban estates'. You could probably
classify life-histories according to transition through these profiles!” (Respondent qa38).
“Can you create the same for 1991 and show us change?!” (Respondent qa13).
Life course and change over time
“Both areas on my piece of map should be group E as they are both ‘Idyllic
Countryside’” (Respondent qa36)
“OA [withheld] is classified as A and should really be E, this OA is just over the Bristol
Suspension Bridge and is definitely not average!” (Respondent qa39).
Preconceptions and Idealisation of Home
“While at such a level no mapping is going to be perfect, the map provided put very different land
covers together into single clusters. For example, on the map provided, large industrial areas were clustered together with comfortable suburban developments.” (respondent qa54).
“I’m surprised that Headingley [an area of Leeds heavily populated by students] appears to be in
the same category as the city centre” (Respondent qa44).
Distinguishing the Function of an Area from its residential Pattern
• The consultation had two functions, quality assurance of the product and solicitation of suggestions for improvement.
• The results of the quality assurance, we suggest, show the National Statistics Classification is fit for purpose.
• We have implemented a number of the suggestions of respondents in the final product and incorporated others in our research agenda. We believe that incorporating consultation into the research process has considerable value and should ensure a longer shelf life for the research product.