|
Abstract:
|
The factors impacting the use of force by police is a topic of great current interest. We study the impact of community-level socioeconomic features, such as unemployment, as well as event-level features, such as officer tenure, on force used by police. We develop a novel two-stage log Gaussian Cox process model that relates community-level attributes to the location of police use of force events. Our model also relates both community and event-level features to the type of force used. This model addresses the challenge of describing marked point processes in terms of both spatial (community-level) and nonspatial (event-level) characteristics. We find that our two-stage model is flexible and easy to interpret, and we find via simulated examples that it provides reliable parameter estimates. By applying our methods to data from forest fires in Spain, we also learn about the potential impacts of spatial and event-level characteristics on fire location and size.
|
