Data analysis was carried out on diverse levels, though mainly concentrating on the analytical methods, statistical measures and spatial statistics. Different types of methods and statistics as employed in crime studies are available for both tabular and spatial analysis. GIS has also enabled a variety of analyses to be conducted as listed in the examples below (CMAP, 2002):
- Offender/Victim Movement analysis – monitoring the travel patterns of offenders and victims
- Temporal analysis – analysing crime over time (yearly, etc) and at different time periods
- Environmental analysis – analysing the location of offences and offenders
- Property analysis – analysing the risk of crime by property type
- Link analysis (criminal relationships) – analysing the relationships between offenders and groups prone to offend
- Victim/Target Characteristic analysis – monitoring the interactions of offenders and victims in their routine activities
- Causal analysis – analysing the cause and effect of offence and offenders on the physical and social fabric
These methods require statistical tests to enable analysis and such software as CrimeStat aid in the analytical process prior to the crime-mapping analysis through a dedicated GIS application.
As an example of types of spatial statistics used in Crime, listed below are the CrimeStat categories clustered in four-groups : Spatial distribution, Distance statistics, ‘Hot spot’ analysis routines and, Interpolation statistics:
1. Spatial distribution - the mean center, center of minimum distance, standard deviational ellipse, Moran’s I spatial autocorrelation index, or angular mean;
This study employs a variety of these methods, particularly Moran’s I, hot spot and interpolation. It also bases its procedures on the following methods.
2. Distance statistics - nearest neighbour analysis, linear nearest neighbour analysis, and Ripley’s K statistic;
3. ‘Hot spot’ analysis routines - hierarchical nearest neighbour clustering, K-means clustering and local Moran statistics;
4. Interpolation statistics - a single-variable kernel density estimation routine for producing a surface or contour estimate of the density of incidents (e.g. auto thefts) and a dual-variable kernel density estimation routine for comparing the density of incidents to the density of an underlying baseline (e.g., auto thefts relative to the number of households).
The following section shows some map outputs of crime in the Maltese islands
Maps - Various Methodologies
The following maps depict crime offences based on a spatial statistical
measures known as NNA (Nearest Neighbour Analysis), NNH (Nearest Neighbour Hierarchical Cluster Analysis), and 3D analysis. These processes
create hotspots for crimes where red represents the highest
rates of offences and green/blue the lowest rates in the first two images. Each map
contains the landuse activities related to that area. Whilst
the harbour map shows a high relationship to retail and recreation,
the Pieta map (town on the suburbs of Valletta) depicts a
relationship that has a more detailed social construct: hospital,
drug services, housing estate, etc. In the case of the last two maps, migration of crime is investigated over time whilst the 3D model gives a visual perspective of crime in an easily dicernable mode.
The Harbour Area -
The Pieta Local Council -
The Malta Migration -
One can also investigate migration of crime using cluster analysis.
The 3D Analytical Component -
One can also investigate crime through 3D visualisation-component analysis.
Malta's FIRST CrimeMaps
Model Map: Local Councils - 2006
Model Map: Local Councils - 2006