ABSTRACT
Environmental criminology begins its analysis of the criminal event by asking where and when a crime has occurred (Brantingham and Brantingham, 1991). Through this explicitly spatial-temporal approach to studying crime (broadly encompassing geometric, routine activity, rational choice, opportunity, and pattern theories), significant insight is garnered: human activity patterns and, hence, crime patterns follow both spatial and temporal regularities. The development of geographic information science (GIScience) and the availability of spatially referenced crime data allow the spatial dimension of these theories to flourish. Most often, this occurs through aggregation of criminal event data to census boundary units and analyses in relation to census variables. These types of analyses, however, may be problematic. In order to calculate crime rates for census boundary units, the residential population is most often used. As discussed in the Chapter 2, a potential problem arises because people leave their census boundary units during the day, which is a spatial and temporal regularity. Consequently, crime rate calculations based on the residential population may not provide accurate representations of risk.
