ABSTRACT
It is generally assumed that the process by which data describing a spatial process are organized into mutually exclusive and collectively exhaustive areal units affects analytical results. It is further assumed that these analytical results will vary in response to a change in either the number of units, or the organization of smaller units into larger units. This apparent indeterminacy has been called variously the aggregation problem, the ecological fallacy, or the modifiable areal unit problem. While the literature dealing with this issue is growing, it is surprisingly small given the rather fundamental nature of the problem as it pertains to spatial analysis (Openshaw 1984).
