Using crime data and Kernel Density

[RickKepler]

We have been using Kernel Density to identify hotspots of various types of crimes. 

I was wondering if there were others out there that have found a good representation of the data using this method.  Normally we use the Natural Breaks and up to 7 Classes with the first being void.

[LarryGaudieri]

I'm working on a similar project using a whole year's worth of fire rescue response data. Kernel Density definitely produces an effect and shows incident density for a given geo-unit. However, as I research the concept, I'm finding that there is too much subjectivity using density mapping.

Given that your data is rich with possible trends and patterns, you should really be using Hot Spot Analysis so that you can discover not just how many crimes are happening in a given geo-unit, but a pattern of statistical relevance or interest. Just finished a pretty good tutorial about it and there's some very comprehensive reading that influences the reason why it's important to use it.

Here's a link to the tutorial: http://www.arcgis.com/home/item.html?id=6626d5cc81a745f1b737028f7a519521

Good Luck,