Session 4: Exploratory Spatial Data Analysis (ESDA) Course on Spatial Econometrics with Applications Profesora: Coro Chasco Yrigoyen Univeridad Autónoma de Madrid Lugar: Universidad Politécnica de Barcelona 12-13, 18-20 de junio, 2007 ©2007, Coro Chasco Yrigoyen All Rights Reserved Course Index S1: Introduction to spatial econometrics S2: Spatial effects, spatial dependence S3: Spatial autocorrelation tests S4: Exploratory Spatial Data Analysis (ESDA) S5: Specification of spatial dependence models S6: Spatial regression models: OLS estimation and testing PS1: GeoDa: introduction and ESDA S7: Spatial dependence models: estimation and testing S8: Modelling strategies in spatial regression models PS2: SpaceStat: SpaceStat: confirmatory spatial data analysis S9: Specification of spatial heterogeneity models S10: Spatial heterogeneity models: estimation and testing PS3: Practical exercise and evaluation @ 2007, Coro Chasco Yrigoyen All Rights Reserved 2 . CHASCO, C. (2003), “Econometría espacial aplicada a la predicción-extrapolación de datos microterritoriales”. Comunidad de Madrid; pp. 28-48. Session 4 Overview and Goals Introduction to ESDA Spatial distributions plots: 1. Choropleth maps 2. Box-map 3. Cartogram Spatial association plots: 1. Moran’s scatterplot 2. LISA maps 3. Bivariate & space-time autocorrelation plots Readings @ 2007, Coro Chasco Yrigoyen All Rights Reserved 3 1 Session 4 4.1. Introduction to ESDA (I) ESDA is a subset of EDA. EDA is a collection of descriptive and graphical statistical tools intended to discover patterns in data and suggest hypotheses. EDA allow the user to directly manipulate various “views” of the data: histograms, box plots, q-q plots, dot plots, and scatterplot matrices. @ 2007, Coro Chasco Yrigoyen All Rights Reserved Statistica (StatSoft) 4 Session 4 4.1. Introduction to ESDA (II) Geographic maps: a way of illustrating spatial data. ESDA: the explicit consideration of the map as an integrated “view” of the data in a dynamic graphics framework. @ 2007, Coro Chasco Yrigoyen All Rights Reserved 5 Session 4 . ANSELIN, L. (1998),“Exploratory spatial data analysis in a geocomputational environment”. Conference in GeoComputation’98, Bristol (UK), pp. 17-19. 4.1. Introduction to ESDA (III) ESDA techniques: 1. Describe and visualize spatial distributions. 2. Linking & brushing: allows the selection of locations in different views or screens. 3. Identify atypical locations or spatial outliers. 4. Discover patterns of spatial association, clusters or hot spots. 5. Suggest spatial regimes or other forms of spatial heterogeneity. Geostatistical data: point data as a sample of an underlying continuous distribution Lattice data: discrete spatial locations @ 2007, Coro Chasco Yrigoyen All Rights Reserved 6 2 Session 4 4.2. Spatial distributions plots 4.2.1. Choropleth maps 4.2.2. Box-map 4.2.3. Cartogram @ 2007, Coro Chasco Yrigoyen All Rights Reserved 7 4.2. Spatial distributions plots 4.2.1. Choropleth maps @ 2007, Coro Chasco Yrigoyen All Rights Reserved QUANTILE MAPS 8 4.2. Spatial distributions plots 4.2.1. Choropleth maps (II) @ 2007, Coro Chasco Yrigoyen All Rights Reserved Session 4 Session 4 PERCENTILE MAP 9 3 4.2. Spatial distributions plots Session 4 4.2.1. Choropleth maps (III) STANDARD DEVIATION MAP @ 2007, Coro Chasco Yrigoyen All Rights Reserved 10 4.2. Spatial distributions plots Session 4 4.2.2. Box-map @ 2007, Coro Chasco Yrigoyen All Rights Reserved 11 4.2. Spatial distributions plots Session 4 4.2.3. Cartogram @ 2007, Coro Chasco Yrigoyen All Rights Reserved 12 4 Session 4 4.3. Spatial association plots 4.3.1. Moran’s scatterplot 4.3.2. LISA maps 4.3.3. Bivariate & space-time autocorrelation plots. 13 @ 2007, Coro Chasco Yrigoyen All Rights Reserved Session 4 4.3. Spatial association plots 4.3.1. Moran’s scatterplot Visualizes I as the slope of the regression line in a scatterplot with Wz on Y-axis and z on X-axis. . ANSELIN, L. & S. BAO (1997), “Exploratory Spatial Data Analysis”. In “Recent developments in spatial analysis” (Eds. Fischer y Getis), Springer-Verlag, Berlín; pp. 35-59. 14 @ 2007, Coro Chasco Yrigoyen All Rights Reserved Session 4 4.3. Spatial association plots 4.3.1. Moran’s scatterplot (II) Moran scatterplot map IV I (-) (+) II III (+) (-) Moran scatterplot @ 2007, Coro Chasco Yrigoyen All Rights Reserved 15 5 Session 4 4.3. Spatial association plots 4.3.1. Moran’s scatterplot (II) @ 2007, Coro Chasco Yrigoyen All Rights Reserved 16 4.3. Spatial association plots Session 4 4.3.1. LISA maps Local spatial autocorrelation statistics are useful to identify hot spots: Spatial concentration of high/low values or Spatial outliers Local autocorrelation is always present in global spatial autocorrelation, but it can also exist in the absence of it. @ 2007, Coro Chasco Yrigoyen All Rights Reserved 17 Session 4 4.3. Spatial association plots 4.3.1. LISA maps (II) Non-significant Moran’s I over the whole area of the Spanish provinces @ 2007, Coro Chasco Yrigoyen All Rights Reserved 18 6 4.3. Spatial association plotsSession 4 4.3.1. LISA maps (III) 19 @ 2007, Coro Chasco Yrigoyen All Rights Reserved . LÓPEZ, F. & C. CHASCO (2004), “Space-time lags: Specification strategy in spatial regression models ”. REAL 04-T17, Session 4 http://www2.uiuc.edu/unit/real/d-paper/real04-t-17.pdf . 4.3.3. Bivariate & space-time plots (II) Multivariate spatial correlation Bivariate Moran spatial autocorrelation Wartenberg, 1985 Moran space-time autocorrelation Anselin et al., 2002 mkl = zk′W s zl I kl = zk = [Yk − µk ]/ σ k z k′Wzl z ′z k Our proposal I t − k ,t = zt′− kWzt zt′− k zt − k zt = [Yt − µ t ] / σ t z l = [Yl − µ l ] / σ l z t − k = [Yt − k − µ t − k ] / σ t − k Ws is a doubly standardized W @ 2007, Coro Chasco Yrigoyen All Rights Reserved 20 Session 4 4.3.3. Bivariate & space-time plots (III) It-k,t: Moran space-time autocorrelation coefficient Moran’s I value coincides with the slope Employment rate (E) of the regression line of Wzt on zt-k p-value=0.001 p-value=0.427 Wzt Zt-k t=2002 k=4 W:contiguity matrix (0-1) Population (P) @ 2007, Coro Chasco Yrigoyen All Rights Reserved 21 7 4.3.3. Bivariate & spacetime plots (IV) Session 4 Bivariate LISA @ 2007, Coro Chasco Yrigoyen All Rights Reserved 22 Session 4 4.4. Readings 1 2 @ 2007, Coro Chasco Yrigoyen All Rights Reserved 23 Session 4 4.4. Readings (II) 3 4 @ 2007, Coro Chasco Yrigoyen All Rights Reserved 24 8 Session 4 4.4. Readings (III) 5 @ 2007, Coro Chasco Yrigoyen All Rights Reserved 25 9