Best way to define search radius for Kernel density analysis

Discussion created by catrionalbrogan on Jan 19, 2012
Latest reply on Feb 12, 2013 by peter_m
I'm new to GIS so this is probably pretty simple stuff.  I have a set of co-ordinates and I am doing kernel density on it.  Rather than randomly just picking an arbitrary search radius is there any way that I can find out what the most appropriate search radius is for my points?  I have 20 different categories and they are all quite different in their distribution patterns so I am wary of just applying the same search radius to them all.