Daniele Trogu and Michele Campagna on "Spatial Statistics and Composite Indicators: a review of existing case studies and open research issues on Spatial Composite Indicators"
DSPy a system for AI to Write Prompts and Do Fine Tuning
Trogu & Campagna - input2012
1. Spatial Statistics and Composite Indicators:
a review of existing case studies
and
open research issues
on Spatial Composite Indicators
by
Daniele Trogu, Michele Campagna
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
2. Summary
Composite indicators and the spatial dimension
Critical analysis of recent advances
Open issues for further research
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
3. Composite indicators
Multidimensional measure
Description of complex characteristics of reality
Synthesises the effects of a certain set of sub-factors
Provide performance ranking of spatial units
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
4. CI: application domain
Socio-economic systems
Policy-making support tool
Usually based on a-spatial data
CI reference to spatial units (administrative boundary)
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
5. CI: application domain
Usually based on a-spatial data
CI reference to spatial units (administrative boundary)
1
INDICATOR is AVG. MEASURE
? ? ?
2 ?
LARGE SMALL-SCALE SPATIAL UNIT
NO SPATIAL MEASURES
?
? ?
? ?
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
6. Building CIs
Theoretical framework Weighting and aggregation
Data selection Robustness and sensitivity
Imputation of missing data Back to the real data
Multivariate analysis Links to other variables
Presentation and visualization
Normalisation
(Source: OECD and JRC, 2008)
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
7. CI and Spatial statistics:
understanding the spatial structure
CI values may be not random in space
CI value patterns may be spatial dependent
Thus, spatial dependence must be take into account in forecast
models
In policy-making
accounting for the spatial structure allows to better understand
phenomena in spatial units
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
8. Spatial Dimension of CI
State of Art
Recent growing interests towards the spatial dimension of CI due to:
Advance in technologies (computing power & software)
Advance in spatial data availability (e.g. Spatial Data Infrastructures)
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
9. Analysis of recent advances:
Spatial CI Matrix
Taxonomy of three main domains:
Indicators
Sub-domains
Methods
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
10. Spatial CI matrix structure:
Spatial CI
Spatial CI Matrix
Matriix
Indicators Sub-domains Methods
Indicator Economy Spatial features
Authors
Social Spatial Units
Composite
Environmental Study of spatial
Domain
dependence
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
11. Matrix(1): Indicators
Indicators give meta-information about CI:
Indicator (name)
Authors
Composite
Domain
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
12. Matrix (2): Sub-domain
Sub-domain shows which are the sub-factors considered to define the
CI:
Economy
Social
Environment
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
13. Matrix (3): Methods
Describes the methodology and technical aspects used for analysing
the spatial dependence of the CI:
Spatial features
Spatial units
Study of Spatial Dependence
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
14. Results
CI are often used only in the socio-economic field
Spatial dimension in composite indicator is still a rather unexplored
field
Spatial units are usually administrative boundary
Spatial data are seldom used in CIs construction
Spatial location may influence the performance of a given spatial
units
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
15. Open research issues
More research to understand the implication for CIs in terms of spatial
dependence
New methods to extend CIs design methodologies to the case of
spatial data (i.e. GIS based spatial composite indicators)
Possibility to choose spatial units on the basis of particular spatial
features of phenomena (e.g. Landscape value or Env. Sustainability)
Understand how spatial analysis can help in the definition of Spatial
Composite Indicator
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it
16. Thank you
for your attention!
Questions, comments, remarks, suggestions are
welcome!
D. Trogu daniele.trogu@unica.it
M. Campagna campagna@unica.it