Skip to main content
. 2020 Apr 6;3:1722. Originally published 2019 Dec 4. [Version 2] doi: 10.12688/gatesopenres.13089.2

Figure 3. Relationship between homogeneity index and residual for three values of epsilon.

Figure 3.

The homogeneity index, defined as the number of cells with precise count of zero in the detailed histogram, is positively associated with the bias (markers show the mean difference between the DP count estimated by TopDown and the precise count, and shaded area shows the distribution of individual differences). This plot shows the association for enumeration districts, and a similar relationship holds at the county level. As ϵ increases, the scale of the bias decreases. (Enumeration districts attained only a subset of the homogeneity index values between 0 and 23, which is why there are different width gaps between markers. We pooled the residuals for the four runs of TopDown with different random seed.)