Computing the Spatial Correspondence between Geographical Patterns

Bibliographic Data

ID8963550
AuthorsA D Cliff (0000-0002-5636-0959, corresponding author)
Year1970
Issue50
Pages143-143
Publication date1970-07-01
Peer ReviewedYes
Open AccessNo
TypeARTICLE
VenueTransactions of the Institute of British Geographers (JOURNAL)
Journal identifiersISSN: 0020-2754 • E-ISSN: 1475-5661
PublisherWiley (PUBLISHER • GB)
DOI10.2307/621351
OpenAlexW2322796855
LanguageEN
Citations received7
References cited1

A Sign test for evaluating the goodness-of-fit between maps of area-based data is described. The test is based upon measures of contiguity and is used to determine the degree of correspondence between some maps of observed and simulated patterns of adopters of an innovation given by T. Hagerstrand (I953). The observed and simulated patterns are shown to have systematic spatial differences, reasons for which are given. GEOGRAPHICAL research often requires estimates to be made of the amount of agreement between patterns shown on maps. The patterns may consist of points, lines, areas, intensities, flows, or any combination of these elements (W. Tobler, I965). As examples of the kinds of problem which arise, Tobler states (p. 131) that it might be required to compute the spatial correspondence between the railway and road patterns of the United States; or the percentage agreement between the observed distribution of cities within Europe and Christaller's theoretical arrangement of central places. The comparison of an observed pattern with a theoretically derived pattern is of considerable importance (W. W. Bunge, I963), since it is usually undertaken in order to verify how closely the model from which the theoretical pattern was obtained conforms to reality. Several recent studies (H. H. McCarty et al., 1956; A. H. Robinson and R. A. Bryson, I957; R. Bachi, 1962; R. F. Minnick, I964; Tobler, 1965; D. Neft, 1966) have focused upon methods for computing the degree of spatial correspondence between maps of line or point patterns. The problem of comparing maps of area-based or county data has, however, been neglected. The purpose of this paper is to provide a test which measures the spatial correspondence between observed and theoretically derived maps of county data

Economic geography · Geography · Regional Economic and Spatial Analysis · Regional Economics and Spatial Analysis · Spatial and Panel Data Analysis

Tools

Open DOISci-Hub