Table 1 Ipop global clustering. Case count correlations for the geographic distribution of invasive and late-stage prostate or breast cancer incidence within or among selected areal units of analysis, Connecticut, 1988–92.
From: Lumping or splitting: seeking the preferred areal unit for health geography studies
Areal Unit | Within % | Among % | p-value |
|---|---|---|---|
Prostate Cancer Incidence | Â | Â | Â |
   Block Group | 60.0 | 40.0 | 0.0002 |
   Tract | 60.1 | 39.9 | 0.0002 |
   Town | 57.4 | 42.6 | 0.0002 |
Breast Cancer Incidence | Â | Â | Â |
   Block Group | 72.1 | 27.9 | 0.0002 |
   Tract | 63.7 | 36.3 | 0.0002 |
   Town | 75.8 | 24.2 | 0.0002 |
Late Stage Prostate Cancer Incidence | Â | Â | Â |
   Block Group | 99.7 | 0.3 | 0.0008 |
   Tract | 87.9 | 12.1 | 0.2190 |
   Town | 77.0 | 23.0 | 0.2740 |
Late Stage Breast Cancer Incidence | Â | Â | Â |
   Block Group | 100.7 | -0.7 | <0.0001 |
   Tract | 100.2 | -0.2 | 0.0002 |
   Town | 89.1 | 10.9 | 0.0846 |