Salient Properties of Geographic Connectivity Underlying Spatial Autocorrelation

Abstract

Features of spatial autocorrelation can be established with analytical, computational, and conceptual techniques. Analytical techniques rely upon algebra and geometry to obtain mathematical generalizations about spatial autocorrelation. Computational techniques rely upon computing power, algorithms, and statistical theory to obtain numerical outcomes, experimentally exploring the behavior of spatial autocorrelation in the absence of analytical equations. Tools commonly employed in experimental exercises include resampling and simulation. Conceptual techniques often involve logical arguments that sometimes derive expectations through analogies; many facets of spatial autocorrelation are paralleled with those of serial correlation (e.g., time series).