Crime Hot Spot Analysis
Also known as: Hot Spot Mapping, Crime Hotspot Detection, Getis-Ord Gi* Crime Analysis, Spatial Cluster Analysis of Crime
Crime hot spot analysis identifies the places where crime concentrates far more than chance — the small number of street segments, blocks, or addresses that account for a large share of incidents. Building on Sherman and Weisburd's landmark demonstration that crime clusters tightly in space and that patrolling those clusters deters offending, the method uses spatial statistics such as the Getis-Ord Gi* local statistic to separate genuine, statistically significant clusters from random noise and to classify each place as a hot spot, a cold spot, or neither.
Key highlights
- Distinguishes statistically significant clusters from incidental density, giving a defensible basis for targeting resources.
- Provides per-location significance values, so analysts can map both hot and cold spots with explicit confidence levels.
- Directly supports hot-spots policing, an intervention with strong randomized-trial evidence for crime reduction.
- Works with either point events aggregated to units or pre-aggregated area counts, fitting many data situations.
- Implemented in standard GIS and statistical software, making it accessible and reproducible for agencies.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use crime hot spot analysis when you have geocoded incidents or area crime counts and need to identify, with statistical confidence, where crime concentrates so resources can be targeted. It is central to hot-spots policing, resource allocation, and evaluating place-based interventions, and the Getis-Ord Gi* approach is appropriate when crime is aggregated to comparable spatial units and you want to distinguish real clusters from noise. It is less suitable when events are too sparse to populate units, when the spatial unit is so large that within-unit variation is hidden, or when the question concerns continuous risk surfaces (kernel density) or short-term space-time contagion (near-repeat analysis). It describes where, not why, crime concentrates.
Strengths & limitations
- Distinguishes statistically significant clusters from incidental density, giving a defensible basis for targeting resources.
- Provides per-location significance values, so analysts can map both hot and cold spots with explicit confidence levels.
- Directly supports hot-spots policing, an intervention with strong randomized-trial evidence for crime reduction.
- Works with either point events aggregated to units or pre-aggregated area counts, fitting many data situations.
- Implemented in standard GIS and statistical software, making it accessible and reproducible for agencies.
- Results depend strongly on the spatial unit and the definition of neighbors (spatial weights), which can change which areas appear hot.
- Aggregating points to units invites the modifiable areal unit problem, where conclusions shift with the zoning of the map.
- It identifies where crime concentrates but offers no explanation of the underlying causes or risk factors.
- Sparse crime or very small units yield unstable counts and unreliable significance values.
- Testing many units simultaneously inflates false positives unless multiple-comparison corrections are used.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
What is the difference between a kernel density hot spot and a Getis-Ord hot spot?
Kernel density estimation produces a smooth surface showing where incidents are dense, but density alone does not test whether that concentration is more than chance. The Getis-Ord Gi* statistic explicitly tests each location against a null of spatial randomness and returns a significance value, so its 'hot spots' are statistically defensible clusters. Many analysts use kernel density for visualization and Gi* for inference; they answer related but distinct questions.
Why does the choice of spatial unit matter so much?
Crime counts and clustering depend on how space is divided. Large units can hide concentration within them, while units drawn differently can merge or split the same cluster — this is the modifiable areal unit problem. Because Gi* also depends on how neighbors are defined through spatial weights, the same data can yield different hot spots under different choices. Good practice fixes the unit and weights on substantive grounds and reports sensitivity to alternatives.
Does finding a hot spot tell us why crime concentrates there?
No. Hot spot analysis is descriptive: it locates statistically significant concentrations but says nothing about the environmental features, opportunities, or offenders that cause them. To understand and act on the drivers, analysts pair hot-spot maps with diagnostic methods such as risk terrain modeling or problem-oriented analysis of the places involved.
Sources
- 1.Sherman, L. W., & Weisburd, D. (1995). General deterrent effects of police patrol in crime "hot spots": A randomized, controlled trial. Justice Quarterly, 12(4), 625–648.
- 2.Getis, A., & Ord, J. K. (1992). The analysis of spatial association by use of distance statistics. Geographical Analysis, 24(3), 189–206.
You have read it. What now?
Cite this page
ScholarGate. (2026, June 22). Crime Hot Spot Analysis. ScholarGate. https://scholargate.app/criminology/hot-spot-analysis-crime