|
Parameter
|
Estimate
|
95% CI
|
SE
|
VIF
|
P value
|
|---|
|
Model A.
| | | | | |
|
Intercept
|
0.75
|
0.46 – 1.05
|
0.15
| |
< 0.001
|
|
Head of household is a woman
|
7.77
|
0.73 – 15.17
|
3.67
|
1.14
|
0.034
|
|
Age of head of household
|
0.10
|
0.00 – 0.21
|
0.05
|
1.06
|
0.051
|
|
Residual of HCI regressed on households with no access to piped water inside the home
|
9.04
|
3.98 – 14.37
|
2.64
|
1.05
|
< 0.001
|
|
Distance to central hospital
|
−0.0005
|
−0.0007 – 0.0
|
0.0001
|
1.20
|
< 0.001
|
|
Model B.
| | | | | |
|
Intercept
|
−7.59
|
−14.24 – −1.32
|
3.28
| |
0.021
|
|
Head of household is a woman
|
7.30
|
0.11 – 14.83
|
3.74
|
1.18
|
0.051
|
|
Age of head of household
|
0.14
|
0.002 – 0.26
|
0.07
|
1.60
|
0.052
|
|
No piped water inside the home
|
−3.18
|
−6.08 – −0.4
|
1.44
|
3.64
|
0.027
|
|
HCI
|
9.16
|
4.06 – 14.56
|
2.66
|
3.11
|
0.001
|
|
Distance to central hospital
|
−0.001
|
−0.001 – 0.0
|
0.0
|
1.31
|
<0.001
|
- (Model A.) The best fit model, which included the residual of the HCI regressed on the variable for no access to piped water inside the home. (Model B.) The same model is shown with separate parameters for the HCI and no access to piped water inside the home to indicate the direction of the effects of the parameters in the model. VIF values indicate a high degree of multicollinearity in model B compared to model A. High values of HCI indicate poor housing condition.
