hotspots under 20 was calculated using a trial and error method. This generates ellipses
(i.e. hotspots) that refl ect the underlying crime events and eliminates those ellipses that
could be created to make up the numbers to reach 20. The only guidance for applying a
suitable search radius was experimentation and experience ( Levine, 2004 ). This resulted
in the use of three search radii to determine hotspots – 500 m, 250 m and a search radius
equal to the default bandwidth in Hotspot Detective (the rationale for choosing this
Hotspot Detective value is explained below in the KDE section). This Hotspot Detective
derived parameter determined an alternate search radius size that varied in accordance
with the spatial characteristics of the crime input data. The three approaches were called
STAC-500, STAC-250 and STAC-HD. Once these parameters had been set and run in
CrimeStat, the output was imported into MapInfo for display.