SpaceStat is organized into four modules, each concerned with a specific set of functionality:

• Data Module: data input, manipulation and transformation
• Tools Module: spatial weights creation, manipulation and analysis
• Explore Module: descriptive spatial statistics
• Regress Module: spatial regression analysis

Data Module

Item

Functionality

Input

Creation of SpaceStat data sets and spatial weights files from ascii input

Merge/Select

Manipulation of SpaceStat data sets (merging data sets, adding and deleting observations or variables, subsetting); subsetting spatial weights

Variable Create

Constructing constants, observation numbers and dummy variables, random variates, relabeling variables

Variable Transform

Standard data transformation functions (log, exp, standardization, etc.)

Spatial Transform

Spatial lag, spatial moving average, spatial filter and spatial transformation (moving average and autoregressive), non-contiguous random resample

Rate Transform

Constructing proportions, standardizations (Freeman-Tukey, arcsin, Anscombe, Empirical Bayes), smoothing (EB, spatial window, spatial EB)

Variable Algebra

Addition, subtraction, etc. of variables, trend surface polynomials, regime variables, expansion variables, principal components

Matrix Algebra

Element by element manipulations, matrix multiplication and inverse, determinant and trace

List

Summary of contents of SpaceStat data set and weights files, listing of contents or selected variables/observations of SpaceStat data sets, listing of contents of spatial weights files

Tools Module

Item

Functionality

Weight Characteristics

Connectivity structure (most/least connected, unconnected, frequency distribution of links), dominant root, eigenvalues, traces

Weight Transform

Row-standardization and higher order contiguity, element by element manipulation of spatial weights, boundary shares over distance weights, dissolve areal units in weights file

Weight Conversion

Conversion between three spatial weights formats (full matrix, sparse contiguity, sparse general), relabeling and sorting elements of weights files

Distance Weights

Computing distance matrices and construction of spatial weights based on distance (contiguity, inverse distance, inverse distance power, k-nearest neighbors)

Sparse Distance

Weights Same as distance weights but using sparse formats rather than full matrix

Access Measures

Computation of origin-destination pair distance and various measures of accessibility (potential, travel cost, covering)

Raster Weights

Construction of contiguity weights for regular grids using rook, bishop or queen criterion, resampling based on coding approach

GIS Functions

Generic functions to construct spatial weights and centroids from ascii input files (e.g., Arc/Info AAT files, boundary files)

explore module

Item

Functionality

Descriptive Statistics

Non-spatial descriptive statistics, quartiles, percentiles, outliers, correlations, principal components; multivariate spatial autocorrelation

Join Count Statistics

Binary and multinomial join count statistics with inference based on a normal approximation (non-free sampling) and a permutation approach

Moran

Moran’s I statistic for global spatial autocorrelation with inference based on normal approximation, randomization and permutation, spatial correlogram, Moran scatterplot, local Moran

Geary

Geary’s c statistic for global spatial autocorrelation with inference based on normal approximation, randomization and permutation, spatial correlogram

G-statistics

Global G statistic for spatial autocorrelation, local Gi and Gi* statistics

QAP

Combinatorial statistics for Moran’s I, Geary’s c and Sokal absolute difference, generic matrix comparision

Regress module

Model Specifications

• generic regression
• trend surface regression
• spatial regimes
• spatial expansion
• spatial analysis of variance (ANOVA)

Model Estimation

Model

Estimation Methods

Classic Model

Ordinary Least Squares (OLS); OLS Robust (White, Jackknife); Weighted Least Squares

Spatial Error Model

Spatial Autoregressive Error (SAR), maximum likelihood estimation (ML); SAR Error with groupwise heteroskedasticity (e.g., spatial regimes), ML; SAR Error with weighted regression, ML; Spatially weighted least squares (interactive); SAR Error, generalized moments (GM) estimator (two-step); SAR Error, GM estimator (iterated); SAR Error with groupwise heteroskedasticity, GM estimator

Heteroskedastic Error Model

Generic heteroskedasticity (user-specified), feasible generalized least squares (FGLS); Generic heteroskedasticity (user-specified), ML; Groupwise heteroskedasticity (FGLS); Groupwise heteroskedasticity (ML); Random coefficients (FGLS); Random coefficients (ML)

Spatial Lag Model

Spatial Autoregressive Lag (SAR), ML; SAR with groupwise heteroskedasticity, ML; SAR, two stage least squares (2SLS); SAR with groupwise heteroskedasticity, 2SLS; SAR, robust 2SLS; SAR, bootstrap

Systems Model

Endogenous variables, 2SLS; Endogenous variables with groupwise heteroskedasticity, 2SLS-GM; Endogenous variables, robust 2SLS; Endogenous variables with SAR error autocorrelation, GM-2SLS; Endogenous variables with SAR error autocorrelation and groupwise heteorskedasticity, GM-2SLS

Specification Tests and Diagnostics

Model

Tests and Diagnostics

Classic Model

Multicollinearity condition number; Bera-Jarque normality; Breusch-Pagan, Koenker-Bassett, White heteroskedasticity; Moran's I, Kelejian-Robinson, LM-error, LM-lag, robust LM-error, robust LM-lag, LM-SARMA spatial autocorrelation tests
Where appropriate: Chow test on regimes, test on spatial expansion coefficients

Spatial Error Model

Spatial Breusch-Pagan heteroskedasticity; LR test on spatial error coefficient; LR and Wald test on spatial common factor hypothesis; LM test on remaining spatial lag dependence Where appropriate: Chow test on regimes, test on spatial expansion coefficients, LR and Wald test on groupwise heteroskedasticity

Heteroskedastic Error Model

Wald and LR test on heteroskedastic coefficients; LM test on spatial lag and spatial error Where appropriate: Chow test on regimes

Spatial Lag Model

Spatial Breusch-Pagan heteroskedasticity; LR test on spatial autoregressive coefficient; LM test on remaining spatial error dependence
Where appropriate: Chow test on regimes, test on spatial expansion coefficients; LR and Wald test on groupwise heteroskedasticity

SpaceStat Extension for ArcView

The SpaceStat Extension adds two menu items to the View GUI in ArcView, Data and SpaceStat:

The Data menu item is also available in the Table GUI. The functionality of the two menus is as follows:

• Data Menu: data interchange between ArcView and SpaceStat
• SpaceStat Menu: visualization of SpaceStat data analysis results

Data Menu

Item

Functionality

Add Selected Features Dummy

Adds a new variable to an ArcView table with value equal to one for selected features (selected locations) and zero for the others.

Add Centroid Coordinates

Computes the coordinates of feature centroids (works for point and polygon data) and adds them to an ArcView table as X_Coord and Y_Coord.

Clean Shape File

Merges records for multiple polygons that correspond to a single observation into a single record and creates a new ArcView shape file.

Table to SpaceStat Data Set

Exports an ArcView Table or selected variables/records of a Table to the binary format of a SpaceStat data set.

Rook Weights from Shape File

Creates a gal file of spatial weights based on rook-type contiguity (common boundaries) of features in an ArcView shape file.

Queen Weights from Shape File

Creates a gal file of spatial weights based on queen-type contiguity (common boundaries and common vertices) of features in an ArcView shape file.

Table to SpaceStat Input File

Exports an ArcView Table or selected variables/records of a Table to an ascii file in the propper format for input into a SpaceStat data set. (for backward compatibility only)

Boundaries from Shape File

Exports the boundary coordinates of polygons in an ArcView shape file to an ascii file for input into a SpaceStat weights file.
(for backward compatibility only)

Join SpaceStat Report File

Generic join of any SpaceStat Report file to an existing table in ArcView.

SpaceStat Menu

Item

Functionality

Box Map

Creates a new View with a quartile map for a selected variable with the outliers highlighted (a box map).

Percentile Map

Creates a new View with a percentile map for a selected variable.

Spatial Lag Bar Chart

Creates a new View with a bar chart map showing the value of a selected variable and its spatial lag (can also be used for any spatial smoother computed in SpaceStat).

Spatial Lag Pie Chart

Same as spatial lag bar chart, but using a pie chart map.

Spatial Smoother

Creates a new View with a quintile map for the spatially smoothed values of a selected variable. The smoother may be a spatial lag, window average, or any of the rate smoothers computed by SpaceStat.

Moran Scatterplot Map

Creates a new View with a unique value map with four colors corresponding to the four quadrants of the Moran Scatterplot of a selected variable.

LISA Local Moran Map

Creates a new View with a unique value map for those locations with a significant Local Moran statistic.

Moran Significance Map

Creates a new View with a combination of a Moran Scatterplot Map and a Local Moran map, showing the quadrant of the Moran Scatterplot only for those locations with a significant Local Moran statistic.

G-Stat Map

Same as LISA Local Moran Map but for the Gi or Gi* statistic.

Residual Map

Creates a new View with a standard deviational map for the residuals of any spatial regression in SpaceStat.

Predicted Map

Creates a new View with a bar chart map showing the observed and predicted values for any spatial regression in SpaceStat.