Presentation on theme: "11 Pre-conference Training MCH Epidemiology – CityMatCH Joint 2012 Annual Meeting Intermediate/Advanced Spatial Analysis Techniques for the Analysis of."— Presentation transcript:
11 Pre-conference Training MCH Epidemiology – CityMatCH Joint 2012 Annual Meeting Intermediate/Advanced Spatial Analysis Techniques for the Analysis of MCH Data Tuesday, December 11, 2012
Session Leaders Russell S. Kirby, PhD, MS, FACE Department of Community and Family Health, College of Public Health, University of South Florida Marilyn O’Hara, PhD Director of GIS and Spatial Analysis Lab Department of Pathobiology University of Illinois 2
Acknowledgement: This presentation based on a Powerpoint lecture by Professor Dante Verme, George Washington University 4
6 GIS Integrates databases, graphics with digital maps. Geographic display of information
7 What is GIS?
10 What is GIS?
11 Hot Spot Analysis
12 Hot Spot Analysis Identify Statistical Significant Spatial clusters of high (hot) or low (cold) from a particular event (areas of high counts from an event). It works with number of events summarized in a point. Based on the Getis-Ord test statistic
13 Hot Spot Analysis 911 Calls in Portland
14 Hot Spot tool is located in the Mapping Clusters toolset in the Spatial Statistics tools. Hot Spot Analysis
15 Hot Spot Analysis To work properly it would require as input a feature class from a geodatabase. Populate its dialog.
16 Hot Spot Analysis
17 Hot Spot Analysis Distance Bands Between Neighbor Counts Illustration
18 Hot Spot Analysis
19 Hot Spots
20 Hot Spots
21 Weighting- Distance
22 Hot Spots
23 Spatial Regression
24 Spatial Regression Regression: Regression establishes a relationship among a dependent variable and a set of independent variable(s) Purpose: better understand patterns of spatial relationships between attributes. Objective: predictions
25 Spatial Regression Multiple Regression Model
26 Spatial Regression
27 Spatial Regression Usually follows hot-spot analysis
28 Spatial Regression Spatially Join the 911 Calls in Portland to a census tract layer to determine how many calls were made from each tract. Why? Demo and SES information is available.
29 Spatial Regression A spatial ordinary least square (OLS) regression model is going to determine if the number of 911 calls (dependent variable) from a Portland, OR, census track is a function of the population in each tract (independent variable).
30 Spatial Regression
31 Spatial Regression
32 Spatial Regression
33 Spatial (OLS) Regression
34 Spatial (OLS) Regression
35 Spatial (OLS) Regression
36 Spatial (OLS) Regression
37 Spatial Regression Thematic Map of Residuals
38 Spatial (OLS) Regression Moran’s Test for Residual Spatial Autocorrelation We would like the residuals to be randomly distributed over the study area
39 Spatial Regression What to do next? Identify more predictors to be included in the model. Could be done graphically. Generate a scatter plot matrix. Check next two slides.
40 Spatial Regression
41 Spatial Regression What to do next? Identify more predictors to be included in the model. Generate a matrix scatterplot.
43 Simpson’s paradox House density House Price Spatially aggregated dataSpatially disaggregated data House density Source: Yu and Wei, Geography Department UW Source: Yu and Wei, Geography Department UW
44 GWR Associations vary spatially and are not fixed. GWR constructs separate equations by including the dependent and explanatory variables of features that are within the bandwidth of each target feature.dependentexplanatory variables Bandwiths are preferable chosen to be adaptive. It generates a local regression model for each feature. It is truly a spatial analytical technique.
45 OLS vs GWR GLOBALModel LOCALModel
46 Fixed weighting scheme Bandwidth Weighting function Source: Yu and Wei, Geography Department UW Source: Yu and Wei, Geography Department UW
47 Adaptive weighting schemes Bandwidth Weighting function Source: Yu and Wei, Geography Department UW Source: Yu and Wei, Geography Department UW
48 Weight Matrix
49 Weighting Scheme I
50 Weighting Scheme II d ij = distance between two features i and j d ij = distance between two features i and j h i = nearest neighbor distance from feature i h i = nearest neighbor distance from feature i
51 Weighting Scheme II
52 Spatial GWR Regression
53 GWR Are the regressions coefficients varying across the study area. – –F-tests based on the variability of the individual regression coefficients Surface map of the local regression coefficients over the study area.