A zone design approach A zone design approach for investigating for investigating inequalities inequalities in infant mortality in infant mortality Konstantinos Daras (University of East Anglia) Seraphim Alvanides (University of Newcastle) Ian Gregory (Queen's University Belfast) 3 rd International Population Geographies Conference 19 -21 June 2006, Liverpool, UK
24
Embed
A zone design approach for investigating inequalities in infant mortality
A zone design approach for investigating inequalities in infant mortality. Konstantinos Daras (University of East Anglia) Seraphim Alvanides (University of Newcastle) Ian Gregory (Queen's University Belfast). 3 rd International Population Geographies Conference - 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
A zone design approach A zone design approach for investigating inequalities for investigating inequalities
in infant mortalityin infant mortality
Konstantinos Daras (University of East Anglia)
Seraphim Alvanides (University of Newcastle)
Ian Gregory (Queen's University Belfast)
3rd International Population Geographies Conference
19 -21 June 2006, Liverpool, UK
2
The Geography of Infant Mortality The Geography of Infant Mortality in England & Wales (1911-1971)in England & Wales (1911-1971)
Was there a north-south divide?
Looking at inequalities: Between urban/rural LGDs Between/within counties Between/within automated zones
Infant mortality: Deaths before 1st birthday per 1,000 live births
•Social class •Urban/rural
•Low income•Overcrowded•Unhygienic
Infant mortality is related to:
4
Infant Mortality Rates Infant Mortality Rates in England and Wales (1840-1970) in England and Wales (1840-1970)
Wo
rld
Wa
r I
Wo
rld
Wa
r II
(Source: Beaver (1973): p. 244)
5
Decline of Infant Mortality Decline of Infant Mortality
Conventional explanations: Cities had high rates because of rapid urban growth Lower rates were driven by improvements in
sanitary conditions in towns and by better health care.
Improvements started simultaneously across the whole country in the 1880s.
After Woods et al (1988 & ’89); Lee (1991); Williams and Galley (1995); Woods (1997)
6
Data sources (1)Data sources (1)
The Great Britain Historical GIS (Gregory and Southall, 1998)
1911: 2061 Local Gov. Districts 1928: 2037 Local Gov. Districts 1951: 1527 Local Gov. Districts 1971: 1419 Local Gov. Districts
(Data aggregated to 1971 level)
Infant deaths (before 1st birthday)
Infant births Estimated population Urban/rural classification
7
Data sources (2)Data sources (2)
Wo
rld
Wa
r I
Wo
rld
Wa
r II
(Sou
rce:
Bea
ver
(197
3):
p. 2
44)
1881
1928
1931
1911
1951
1971
Available datasets
Phases of inequality change according Lee (1991)
8
Measuring InequalitiesMeasuring Inequalities
Gini coefficient Coefficient of variation Variance of logarithms Theil’s entropy index
•Between areas inequalities: Gini coefficient of areas•Within areas inequalities:
N
Ga Ga: Gini coefficient of area aN : number of areas
9
Zone DesignZone Design
Aggregation of areal units into zones, so that certain criteria are satisfied.
Available zone design systems AZP (no GIS): Openshaw (1977, 1984) ZDES: Alvanides et al. (1999, 2000) AZM: Martin (2002) A2Z: Daras (2005), O-O & Graph theory
10
How A2Z worksHow A2Z works
ZONE 1ZONE 2
AREAL UNIT
NEW ZONE 2
NEW ZONE 1
OB
JEC
TIV
E F
UN
CT
ION C
ON
TIG
UIT
Y C
HE
CK
• Optimises an objective functionOptimises an objective function• Controls the contiguity of zoneControls the contiguity of zone
11
A2Z componentsA2Z components
Initial AggregationInitial Aggregation ( (Random or userRandom or user))
Shape ConstraintsShape Constraints (Weak or Strong compactness(Weak or Strong compactness))
Boundary ConstraintsBoundary Constraints ( (Railway, A road, WardsRailway, A road, Wards))
Objective FunctionObjective Function ( (HomogeneityHomogeneity))