Global Statistics

• Published in: Statistical Detection and Surveillance of Geographic Clusters pp. 67 - 108
• Main Authors: Rogerson, Peter, Yamada, Ikuho
• Format: Book Chapter
• Language: English
• Published: United Kingdom Chapman and Hall/CRC 2009; CRC Press LLC
• Subjects: Agriculture & related industries; FAMILY, HOME AND PRACTICAL INTERESTS; Geochemistry; Spatial Randomness; SIDS; Cumulative Distribution Function; Quadrat Methods; Significant Spatial Autocorrelation; Low Values Cluster; Positive Spatial Autocorrelation; Quadrat Size; Click Run; SAR Model; ESRI Shapefile; Nonwhite Population; Moran Coefficients; Binary Connectivity; Null Hypothesis; Guard Area; Simulated 95th Percentile; Case Control Data; Unconditional Approach; Significant Positive Spatial Autocorrelation; SIDS Data; Quadrat Analysis; Random Labeling; NHANES II; Sudden Infant Death Syndrome
• ISBN: 9781584889359; 1584889357
• DOI: 10.1201/9781584889366-7

Global spatial statistics summarize, in a single quantity, the degree to which an observed spatial pattern deviates from a specified null hypothesis. Examples of null hypotheses in this context include those of complete spatial randomness, and those in which observed regional frequencies are consistent with a set of prespecified regional expectations for those frequencies. Observed values of the global statistics are compared with the distribution of statistics that is expected under the null hypothesis, and this comparison leads to acceptance or rejection of the null hypothesis, along with a p-value indicating the likelihood of the observed statistic under the null hypothesis. Global statistics are limited in the sense that they do not give a direct indication of the size and location of regions that are inconsistent with the null hypothesis.