Abstract by Spencer Galbraith
Realignment of Areal Data using Spatial Point Processes
Census data contain a rich store of information useful to many fields of study. Unfortunately, the use of census data is often a challenge because the census regions don't align with those used by other sources of data. In statistics, this is the so called ``change of support" problem. Many researchers attempt to realign multiple data sources to a common scale; however, this process implies several inherent complexities. This paper proposes a method to realign areal data using discrete kernel convolutions and a spatial point process. The proposed methodology is applied to census data from the city of Houston, Texas. We show that areal data can be successfully realigned using spatially relevant information.