ABSTRACT
An interesting problem in spatial epidemiological study is largely attributable to the fact that most available data are in the form of complete enumeration of events based on administrative boundaries, such as census tracts within a fixed study area. Such data already form a realization and represent a completely mapped occurrence.1 The problems of such a representation of aggregated data arising from zonation, as presented in Section 5.3.2 of Chapter 5, have sparked increased interest in spatial sampling that involves a limited number of locations in a geographic space to measure a phenomenon.
