Hot Spot

Hot Spots can help you visualize geographic data easily. For instance you can use a 'Hot Spot' Grid to highlight areas which report more burglary incidents.

For both types of spatial analysis, a colour gradient is used to indicate areas of increasingly higher density.

Select Analysis > Grids > Scatter Grid.

This opens the Choose Hot Spot Type dialog:

KDE - Creates a Hot Spot Grid using a Kernel Density Estimation algorithm. Click here for a description of creating a KDE Hot Spot Grid.
Getis-Ord (Gi*) - Creates a Hot Spot Grid using the Getis-Ord (Gi*) spatial statistics. Click here for a description of creating a Getis-Ord (Gi*) Hot Spot Grid.

Points to consider when selecting the type of Hot Spot Grid to create:

Getis-Ord (Gi*) is a more statistical method than KDE, it will ignore any hotspots it deems to be outlying. This means it is better at recommending the most important areas to target resources.
Getis-Ord (Gi*) highlights both hot and cold spots as opposed to just hotspots with KDE. This will identify areas where incidence is low as well as where it is high.
Getis-Ord (Gi*) tends to give a better Predictive Accuracy Index (PAI) score when predicting the future location of crime than all other spatial analysis methods, including KDE.
Getis-Ord (Gi*) allows a theme to be set to the standard confidence intervals regardless of the data used, the output map will not be significantly affected by the choice of colour scheme intervals etc.
Getis-Ord (Gi*) should ideally be used to complement KDE rather than replace it, with Getis-Ord (Gi*) being used to highlight important areas and KDE being used to investigate these areas further.
Getis-Ord (Gi*) is particularly useful when dealing with fields which have a strong spatial connection (crime data, voting analysis, retail analysis, etc.)