Premium
Toward a Theory of Spatial Statistics: Another Step Forward
Author(s) -
Griffith Daniel A.
Publication year - 1987
Publication title -
geographical analysis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.773
H-Index - 65
eISSN - 1538-4632
pISSN - 0016-7363
DOI - 10.1111/j.1538-4632.1987.tb00114.x
Subject(s) - multivariate statistics , autoregressive model , spatial analysis , statistics , econometrics , autocorrelation , computer science , sampling (signal processing) , sample (material) , mathematics , chemistry , filter (signal processing) , chromatography , computer vision
An update is presented to an earlier piece outlining some steps that needed to be taken toward the establishment of a theory of spatial statistics. Findings that have appeared since this first paper are summarized and interpreted, and extensions and suggestions are offered for the further establishment of a basis for a theory of spatial statistics. Topics include boundary considerations, the role of latent spatial dependencies, and small‐sample‐size issues. These topics embrace the problems of data transformations, edge effect bias, reference sampling distributions, multivariate autocorrelation models, conditional expectations, and higher‐order autoregressive structures. In part, a course is charted for the next step to be taken.