Clustering in space using case locations

Use Cuzick & Edward's method if you have locations of both cases and controls. Controls are selected from the same risk population as the cases. For example, if the cases are places of residence of children with leukemia, the controls might be places of residence of healthy children from the same area as the cases. Cuzick & Edward's method accounts for geographicaly inhomogenous populations.

Geographic clustering using only case locations assumes a homogeneous population density, and is not appropriate for health data where the size of the population at risk varies from one place to another.


Website maintained by Andy Long. Comments appreciated.
longa@nku.edu