Table 3 |
||
|
Parameter estimates γ of the logit model and overall model significance (log-likelihood ratio) |
||
|
Parameter Estimates |
||
|
|
||
|
Variable |
Estimate |
t-score (Pr > |t|) |
|
|
||
|
Geo-indicator2 |
0.006 |
1.98 (0.049) |
|
Rainfall3 |
-0.004 |
-3.91 (<0.001) |
|
Household-indicator4 |
-0.577 |
-2.94 (0.004) |
|
Constant |
2.786 |
2.52 (0.012) |
|
-2 Log Likelihood |
2447.6 |
|
|
-2 Log Likelihood ratio1 |
325.4* |
|
|
Number of observations |
2626 |
|
|
|
||
|
1Under null hypothesis of a constant model, *significant at <0.0001 level. 2Product of aquifer suitability stress, groundwater quality stress, and (2-GW). 3Mean annual rainfall (mm). 4Weighted mean of ED, HG, and DH. |
||
|
Pande et al. International Journal of Health Geographics 2008 7:17 doi:10.1186/1476-072X-7-17 |
||
