Figures & data
Figure 1. Results from the six KRE estimators. These results are based on the same data of disease cases and population at risk. Both fixed bandwidths use the same geographic distance; both background-adaptive bandwidths use the same threshold about population; and both case-adaptive bandwidths use the same threshold about number of cases.
Figure 2. An RCMC process for disaggregating polygon-level location data. In this example, the 14 disease cases are likely to be actually located within a small portion of this rural town where the population is concentrated; without the disaggregation based on the high-resolution population data, the researcher has to assume that the 14 cases are either evenly distributed across the entire town or are concentrated at a point (typically the centroid of the polygon, however defined), both are much less realistic than the scenario created by the disaggregation.
