230 likes | 721 Views
Chapter 3. Methodological context. Spatial analysis as a process. Problem formulation Data gathering Exploratory analysis Hypothesis formulation Modelling and testing Consultation and review Reporting and implementation. Analytical methodologies.
E N D
Chapter 3 Methodological context www.spatialanalysisonline.com
Spatial analysis as a process • Problem formulation • Data gathering • Exploratory analysis • Hypothesis formulation • Modelling and testing • Consultation and review • Reporting and implementation www.spatialanalysisonline.com
Analytical methodologies • Mitchell (2005) Draper et al (2005) www.spatialanalysisonline.com
Analytical methodologies - PPDAC • After Mackay & Oldford (2002) www.spatialanalysisonline.com
PPDAC • Views spatial analysis in a broader context • Model is highly iterative • Analysis is just part of the process • Considers the process flow in reverse • Spatial is special – extending the PPDAC methodology … www.spatialanalysisonline.com
PPDAC • Extended PPDAC for spatial analysis • Spatial focus – spatial relationships and spatial dependence • Spatio-temporal focus • Failure to meet classical statistical criteria • Pattern and process relationship • Spatial data sourcing and quality www.spatialanalysisonline.com
PPDAC – Problem phase • Problem definition • Clear definition of problem/project and (client) expectations is paramount • Desk research as part of initial process • Identification of specific spatial issues: • spatial (and statistical) scale factors • spatial arrangement factors • data availability and quality • spatial grouping issues • ecological fallacy, atomistic fallacy www.spatialanalysisonline.com
PPDAC – Plan phase • Nature of project • Costing requirements • Decision-support requirements • Public participation issues • Operational requirements • Timing and critical dates • Funding and resourcing • Feasibility and risk analysis • Client expectations www.spatialanalysisonline.com
PPDAC – Plan phase – cont. • Specifications and standards • Data components required • Data sourcing and augmenting • Sampling, populations and inference • Adopting multiple approaches • Seeking robustness • Testing/simulating the process flow • Handling exceptions & rare events • Re-examining the problem specification www.spatialanalysisonline.com
PPDAC – Data phase • Data quality & sourcing issues • Quality • Cost • Licensing • Availability • Completeness & consistency • Timeliness • Detail/resolution • Intangible/qualitative data www.spatialanalysisonline.com
PPDAC – Data phase • Data quality & GIS tools - examples • Boundary definition and density computation • Missing data handling and masking • Classification • Data transformation • Breakline handling • Error mapping • Simulation www.spatialanalysisonline.com
PPDAC – Data phase • Changing context • Spatial Data Infrastructures • Geoportals • Metadata • New data capture tools • Changes in data visualisation & usage • Interdisciplinary nature of geospatial analysis www.spatialanalysisonline.com
PPDAC – Analysis Phase • Simplicity and parsimony • Exploratory analysis & visualisation • Interpretation of spatial patterns • Construction of hypotheses • Build-test-criticise ‘model’ cycle • Developing preliminary conclusions • Match to expectations? www.spatialanalysisonline.com
PPDAC – Analysis Phase • Types of model • Mathematical/Statistical models • Cartographic modeling • Flow models • Micro-scale modeling... etc www.spatialanalysisonline.com
PPDAC – Analysis Phase • Simple dynamic cartographic model www.spatialanalysisonline.com
PPDAC – Analysis Phase • Sample risk assessment model output www.spatialanalysisonline.com
PPDAC – Conclusion Phase • Expectations on delivery met? • Simplicity again • Clear/high quality visualisation • Identify and report strengths, weaknesses and possible sources of error • Recommendations • Future actions www.spatialanalysisonline.com
PPDAC – Final comments • Work through and specify each model step before commencing – test out ideas • Identify risk areas and revisit PPDAC elements as required • Confirm overall programme and deliverables with ‘client’ • Manage risks and changes www.spatialanalysisonline.com