Examples

Two examples of crime audit hot spots are included here to demonstrate how this type of analysis could be carried out. The first is one that required little in the way of complex mapping applications. The second is more advanced in terms of quantitative content and method.

High tech hot spot example

The example from the London Borough of Brent http://www.brent.gov.uk displays sophisticated geographical analysis as central to informing stakeholders within the Local Authority. This borough was selected as an example here on the grounds that they maintain a good property database (BS7666 compliant) and a well developed GIS environment. Both of these have been credited with the Council’s success in obtaining funding for Home Office funding.

Crime data is cleaned by a customised tool that matches the police database addresses from the Ordinance Survey Address-Point addresses. This process sanitises the police address data thereby providing a higher successful geo-coding rate and a more representative sample of incidents.

Once the data are accurately geo-coded, an application (Arcview’s Spatial Analyst) is used to analyse the spatial distribution of crime. Spatial Analyst calculates "surfaces" by interpolating values between crime locations, producing a contour map of crime. Hot spots are represented by the peaks in the contour maps.

Crime surfaces can then be compared to underlying socio-economic or demographic data. For instance, the crime surface for burglary would be compared to the baseline distribution of residential households.

Low tech hot spot example

The example of a technologically simple hot spot system described here comes from the Knowsley area within Merseyside Police. Used explicitly for tactical police purposes at present, it use is currently being extended to a more strategic focus.

Every week, crime counts for each beat are calculated. The count for each beat is compared to the average and standard deviation for that beat, using the previous 52 weeks counts. Those beats displaying statistically significant crime counts, where the count exceeds two standard deviations above the mean are deemed hot spots.

The next phase involves a detailed analysis of within beat patterns for the hot spots. The locations, times and circumstances of the criminal incidents are analysed to elicit patterns that generate the relative high crime count.

The analysis is performed entirely within the Excel application using two standard functions (average and standard deviation). No mapping application is required, although utilising some form of geographical representation could further aid in highlighting patterns of hot spot areas.

Back to Hotspots