ABSTRACT
A fundamental issue concerning spatial epidemiology is how the disease data have been observed or sampled. The method of data collection can have a great impact on the results. In many occasions, the exact positions of disease incidence at the street address level may not be known and only statistics at the census tract level are available (e.g., changwat or province in Thailand and DC districts in Hong Kong). Even under situations where exact locations are known, grouped or aggregated data are preferred for reasons of data privacy and ease of interpretation (see Section 2.5 of Chapter 2). The use of count models allows point-based data to be aggregated based on some areal units or census tracts.
