Vicinity
approach to spatial analysis
Index
> research > vicinity
If
you have crime locations in a GIS they are generally points. If you want to
aggregate these by enumeration districts (usually polygons) then this relies
on the accurate geocoding of the crime location. Unfortunately this is rarely
the case, especially if you are using the 100 metre resolution Postcode Address
File. As the polygons of enumeration districts in urban areas can be quite
small this means that a number of crimes can end up in the wrong polygon.
In an attempt to get round this it is possible with a GIS to use an areally
weighted average of the area in the immediate vicinity of the point to correct
for any locational errors in the point location.
In the diagram the crime location (black diamond) is located near two other polygons. To reduce the possibility of location error a radius (r) can be drawn from the point (i) and the variable from the three areas (j=1 to 3) can be weighted according to the area (a) within the circle.
For further details of this research, a suggested radius,
and an application see:
Ratcliffe, J.H. and McCullagh, M.J. (1999) Burglary, victimisation and
social deprivation
Crime Prevention and Community Safety: An International Journal, 1999, Volume
1 Issue 2: 37-46. Download
the paper...
