Table 2.
Ordinal logistic regression analyses (n = 1,020; Grahamstown 2007).
| Model | B | SE | Wald Z | P | OR |
|---|---|---|---|---|---|
| Intercept 1 ≤ 0 | -6.061 | 0.589 | 105.177 | ≤ 0.001 | 0.002 |
| Intercept 2 ≤ 1 | -3.253 | 0.532 | 37.413 | ≤ 0.001 | 0.039 |
| Intercept 3 ≤ 2 Intercept 4 ≤ 3 |
-1.700 0.987 |
0.522 0.522 |
10.610 3.502 |
≤ 0.001 0.061 |
0.183 2.683 |
| Socio-economic conditions/resources | |||||
| Married | 0.029 | 0.135 | 0.045 | 0.833 | 1.029 |
| Male | -0.156 | 0.142 | 1.204 | 0.273 | 0.856 |
| Age | -0.059 | 0.005 | 139.069 | ≤ 0.001 | 0.942 |
| Employment | 0.722 | 0.154 | 22.094 | ≤ 0.001 | 2.058 |
| Education (none or some primary) | -0.767 | 0.194 | 15.617 | ≤ 0.001 | 0.464 |
| Education (some secondary) | -0.085 | 0.165 | 0.269 | 0.604 | 0.919 |
| Income | 0.001 | 0.035 | 0.000 | 0.985 | 1.001 |
| Social capital | 0.501 | 0.119 | 17.845 | ≤ 0.001 | 1.650 |
| Housing quality | |||||
| Leaky roof | -0.050 | 0.134 | 0.138 | 0.710 | |
| Formally build house | -0.029 | 0.150 | 0.037 | 0.847 | 0.971 |
| Neighbourhood quality | |||||
| Crime in the area | 0.080 | 0.072 | 1.255 | 0.263 | 1.083 |
| Residents get value for their rental rates | -0.156 | 0.088 | 3.186 | 0.074 | 0.856 |
| Test | |||||
| Overall model evaluation Wald test |
Chi2 374.550 |
df 12 |
P ≤ 0.001 |
||
| R2 Nagelkerke | 35.3% |
Notes: reference group of the ordinal logistic model is excellent health (4). B = unstandardized regression coefficient.