Table 5 Comparison of the Gini coefficient and the hierarchical cluster reporting criteria using the simulated cluster configurations in the Northeastern USA
From: Using Gini coefficient to determining optimal cluster reporting sizes for spatial scan statistics
(A) Urban centre with small rural clusters | (B) Rural centre with small urban clusters | |||
|---|---|---|---|---|
One large cluster | Three small clusters | One large cluster | Three small clusters | |
# counties | 57 | 11 | 40 | 6 |
Total population in clusters (%) | 4,344,150 (14.7) | 681,984 (2.3) | 2,736,674 (9.3) | 780,451 (2.6) |
Relative risk | 1.18 | 1.46 | 1.23 | 1.43 |
Gini | ||||
1 | 90 % | 0 | 76 % | 15 % |
2 | 1 % | 12 % | 13 % | 28 % |
3 | 0 | 85 % | 4 % | 56 % |
4+ | 9 % | 2 % | 7 % | 1 % |
Hierarchical | ||||
1 | 100 % | 2 % | 100 % | 64 % |
2 | 0 | 52 % | 0 | 25 % |
3 | 0 | 46 % | 0 | 11 % |
4+ | 0 | 0 | 0 | 0 |