GIS5100: Module 4 - Crime Analysis

Hotspot Analysis

This week's lab involved the use of ArcGIS Pro to examine the occurrence of 2018 Washington DC burglaries contained in census tracts and 2017 Chicago homicides in half-mile grids to uncover crime patterns via three different types of hotspot analysis methods: grid overlay, kernel density, and Local Moran's I.  These methods inspect a dataset and scale of analysis for the presence of clustering, which ultimately becomes the focal points to visualizing hotspots and coldspots of crime activity.

Crime activity (CA) is a measure of crime events per some unit of area.  When examining burglaries per census tract, the crime rate was calculated by dividing the total burglary count per total household units multiplied by 1000, ([Join_Count] / [Total Household Units] * 1000).  In the case of homicide points overlaid by half-mile grids, CA was calculated by dividing the total homicides per half-mile square (2640 * 2640 = 6,969,600 feet).  In the above calculations, notice the regional difference between census boundaries and the grid (the scale is important).  In both scenarios, when no crimes were observed within a census tract or grid, those empty record sets were excluded to avoid unreliable results when determining crimes per area, density.

Below is a side-by-side comparison brief analysis steps and results.

Grid-based thematic mapping

Kernel density thematic mapping

Local Moran's I thematic mapping

In Closing

In this module, I learned that Hotspot Analysis (HA) is a spatial analysis and mapping technique used in identifying clusters of spatial phenomena.  Regardless of the method, HA used vectors to identify the locations of statistically significant hot spots and cold spots in the data analyzed.  Crime points were aggregated to polygons for the analysis.  I discovered that density may indicate where clusters exist in a dataset, but NOT if those clusters are statistically significant

The benefit of using any of the HA methods is that they are pretty easy to use and the results are statistically significant that reveal patterns that may or may not have been able to detect with non-statistical methods like the grid-based thematic mapping technique.

Citations

Ned Levine (2010). CrimeStat: A Spatial Statistics Program for the Analysis of Crime Incident Locations (v 3.3). Ned Levine & Associates, Houston, TX, and the National Institute of Justice, Washington, DC. July. Chapter-6

Eck, John E, et al. “Mapping Crime: Understandin Hot Spots.” Mapping Crime: Understandin Hot Spots, National Institute of Justice, Aug. 2005, www.ncjrs.gov/pdffiles1/nij/209393.pdf.