Table 3 |
|||
|
Bayesian hierarchical models with different priors for the clustering component. |
|||
|
coefficient |
(95% CI) |
DIC; pD |
|
|
|
|||
|
All cancers in women |
|||
|
Gaussian CAR prior |
|||
|
adjacency |
0.502 |
(0.083, 0.957) |
12776; 298 |
|
distance ≤ 5 km |
0.560 |
(0.185, 0.924) |
12770; 277 |
|
distance ≤ 10 km |
0.581 |
(0.235, 0.931) |
12739; 289 |
|
Median based CAR prior |
|||
|
adjacency |
0.526 |
(0.071, 0.981) |
12773; 292 |
|
|
|||
|
Breast cancer in women |
|||
|
Gaussian CAR prior |
|||
|
adjacency |
0.680 |
(0.040, 1.315) |
9707; 121 |
|
distance ≤ 5 km |
0.707 |
(0.076, 1.239) |
9725; 94 |
|
distance ≤ 10 km |
0.717 |
(0.176, 1.258) |
9717; 128 |
|
Median based CAR prior |
|||
|
adjacency |
0.660 |
(0.052, 1.284) |
9708; 116 |
|
|
|||
|
The estimates of regression coefficients, their 95% credible intervals for the index of exposure to MSWIs, DIC and pD terms |
|||
|
Goria et al. International Journal of Health Geographics 2009 8:31 doi:10.1186/1476-072X-8-31 |
|||
