Chaning Residentail Patterns 1991 - 2001
Dan Vickers,
Department of Geography, University of Sheffield
UPTAP Conference 21 -23 March 2007
Changing Residential Patterns 1991 - 2001 Background
Project scheduled to run 1st January 2006 – 31st December 2006, in the School of Geography, University of Leeds.
PhD handed in Friday 13th January 2006.
Viva started a year ago to this minute across the corridor.
1st September 2006 Appointed ‘Lecturer in Social and Spatial Inequalities’ in the Department of Geography, University of Sheffield.
Therefore only able to complete 8 out of the scheduled 12 months of the project.
Rearranged schedule to complete the research one day a week, to be completed by the 1st September 2007.
Changing Residential Patterns 1991 - 2001 Research
Original idea: To attempt to examine the change in residential patterns over time using geodemographic techniques.
Why So?My PhD created the national classification of output areas for the ONS http://www.statistics.gov.uk/about/methodology_by_theme/area_classification/default.asp
The classification show distinct and clear residential patens.Click for map: http://www.casa.ucl.ac.uk/googlemaps/OAC-super-EngScotWales.html
Changing Residential Patterns 1991 - 2001 Research
Variable selection.
Can the variables be compared across time?
Can the whole UK be used?
Matching the geography.
Methodology.
Two years separately or all together.
Transformation?
Standardisation.
Method of Clustering.
How many clusters?
Results so far
Changing Residential Patterns 1991 - 2001 Research
Variable Selection
Social Class a/b
Social Class c1
Social Class d
Social Class e
Non White
Changing Residential Patterns 1991 - 2001 Research
Matching the Geography
1991 data assigned to 2001 geography using population weighted centriods.
Changing Residential Patterns 1991 - 2001 Research
Methodology
Data ranked then standardised between 0 -1.
K-means clustering used.
Both years clustered together.
Looking at between 2 – 7 clusters.
Changing Residential Patterns 1991 - 2001 Research
Results so far Cluster Number of Case * Cluster Number of Case Crosstabulation
Count
34281 8417 15679 58377
10190 30546 18206 58942
6231 7015 35100 48346
50702 45978 68985 165665
1
2
3
Cluster Numberof Case
Total
1 2 3
Cluster Number of Case
Total
1991
2001
Final Cluster Centers
.67 .59 .23
.63 .50 .29
.29 .37 .70
.21 .23 .70
.70 .17 .48
s_ab
s_c1
s_d
s_e
s_non_white
1 2 3
Cluster
Number of Cases in each Cluster
109079.0
104920.0
117331.0
331330.0
.000
1
2
3
Cluster
Valid
Missing
In the same group: 99,927 (60.3%)
In different group: 65,738 (39.7%)
Cluster 1: -7,675
Cluster 2: -12,964
Cluster 3: +20,639
Changing Residential Patterns 1991 - 2001 Research
1991
k3_91 * k3_01 Crosstabulation
Count
540 374 449 1363
329 1719 1921 3969
73 359 2835 3267
942 2452 5205 8599
1
2
3
k3_91
Total
1 2 3
k3_01
Total
2001North East
k3_91 * k3_01 Crosstabulation
Count
2268 1292 1929 5489
1279 4722 3474 9475
461 1022 6263 7746
4008 7036 11666 22710
1
2
3
k3_91
Total
1 2 3
k3_01
Total
2001
1991
North West
k3_91 * k3_01 Crosstabulation
Count
1459 724 1396 3579
941 3586 2630 7157
365 881 4810 6056
2765 5191 8836 16792
1
2
3
k3_91
Total
1 2 3
k3_01
Total
2001
1991
Yorkshire & the Humber
Cluster 1: -421 (-31%)
Cluster 2: -1,517 (-38%)
Cluster 3: +1,938 (+59%)
Cluster 1: -1,481 (-27%)
Cluster 2: -2,439 (-26%)
Cluster 3: +3,920 (+51%)
Cluster 1: -814 (-23%)
Cluster 2: -1,966 (-27%)
Cluster 3: +2,780 (+46%)
k3_91 * k3_01 Crosstabulation
Count
1708 609 1262 3579
737 3037 2021 5795
429 771 3531 4731
2874 4417 6814 14105
1
2
3
k3_91
Total
1 2 3
k3_01
Total
k3_91 * k3_01 Crosstabulation
Count
2356 756 2343 5455
952 2948 1727 5627
613 613 5150 6376
3921 4317 9220 17458
1
2
3
k3_91
Total
1 2 3
k3_01
Total
k3_91 * k3_01 Crosstabulation
Count
3681 1350 1701 6732
1412 3457 1931 6800
734 935 2999 4668
5827 5742 6631 18200
1
2
3
k3_91
Total
1 2 3
k3_01
Total
Changing Residential Patterns 1991 - 2001 Research
1991
2001East Midlands
2001
1991
West Midlands
2001
1991
East
Cluster 1: -705 (-20%)
Cluster 2: -1,378 (-24%)
Cluster 3: +2,083 (+44%)
Cluster 1: -1,534 (-28%)
Cluster 2: -1,310 (-23%)
Cluster 3: +2,639 (+41%)
Cluster 1: -905 (-13%)
Cluster 2: -1,058 (-16%)
Cluster 3: +1,963 (+42%)
k3_91 * k3_01 Crosstabulation
Count
1504 1010 720 3234
1307 5864 2363 9534
393 1265 2590 4248
3204 8139 5673 17016
1
2
3
k3_91
Total
1 2 3
k3_01
Total
k3_91 * k3_01 Crosstabulation
Count
6610 2111 1854 10575
3010 5134 2018 10162
1356 1134 3418 5908
10976 8379 7290 26645
1
2
3
k3_91
Total
1 2 3
k3_01
Total
k3_91 * k3_01 Crosstabulation
Count
14155 191 4025 18371
223 79 121 423
1807 35 3504 5346
16185 305 7650 24140
1
2
3
k3_91
Total
1 2 3
k3_01
Total
Changing Residential Patterns 1991 - 2001 Research
1991
2001London
2001
1991
South East
2001
1991
South West
Cluster 1: -2,186 (-12%)
Cluster 2: -118 (-28%)
Cluster 3: +2,304 (+43%)
Cluster 1: +401 (+4%)
Cluster 2: -1,783 (-18%)
Cluster 3: +1,382 (+23%)
Cluster 1: -30 (-1%)
Cluster 2: -1,395 (-15%)
Cluster 3: +1,425 (+34%)
Changing Residential Patterns 1991 - 2001 Research
20011991
Changing Residential Patterns 1991 - 2001 Research
2001
1991
Changing Residential Patterns 1991 - 2001 Research
Changing Residential Patterns 1991 - 2001 Outputs
Papers
Vickers, D.W. and Rees, P.H. (2007). Creating the National Statistics 2001 Output Area Classification. Journal of the Royal Statistical Society, Series A 170(2).
Vickers, D.W. and Rees, P.H. (2006). Introducing the National Classification of Census Output Areas, Population Trends, 125.
Conferences and Meetings
June 2006, 3rd International Population Geography Conf, Liverpool.
August 2006, Regional Science, Jersey.
2nd November 2006 Exploring the use and value of the ONS output area classification, seminar at the RSS in London.
Tomorrow, Geodemographics workshop, ICOSS University of Sheffield.
2nd April 2007 repeat of the RSS seminar by popular demand.
April 2007 AAG San Francisco.
Changing Residential Patterns 1991 - 2001 Outputs
Other activities
The formation of a Classification user group as part of the RSS
Website will hopefully be formally lunched on 2nd April
Other meetings and events to follow
http://areaclassification.group.shef.ac.uk/