Table 3 Effects of adjusting for environmental, social and personal variables on associations between women's education level and likelihood of walking for transport in multilevel logistic regression models*.
| Variables | Model 1 (education only) | Model 2 (education and environmental mediators) | Model 3 (education, environmental and social mediators) | Model 4 (education, environmental, social and personal mediators) |
|---|---|---|---|---|
| Education | ||||
| Up to 10 years | 1.00 | 1.00 | 1.00 | 1.00 |
| 12 years/certificate/trade etc | 1.59 (1.23 to 1.95) | 1.54 (1.18 to 1.89) | 1.51 (1.14 to 1.87) | 1.49 (1.12 to 1.85) |
| University | 1.30 (0.94 to 1.65) | 1.15 (0.79 to 1.51) | 1.12 (0.75 to 1.48) | 1.10 (0.73 to 1.47) |
| Environmental variables | ||||
| Street connectivity | ||||
| Lowest quartile | 1.00 | 1.00 | 1.00 | |
| Second quartile | 0.82 (0.43 to 1.22) | 0.84 (0.45 to 1.24) | 0.83 (0.44 to 1.23) | |
| Third quartile | 1.57 (1.19 to 1.94) | 1.60 (1.22 to 1.98) | 1.59 (1.21 to 1.97) | |
| Highest quartile | 1.35 (0.96 to 1.74) | 1.33 (0.94 to 1.72) | 1.34 (0.95 to 1.73) | |
| Not coastal neighbourhood | 1.00 | 1.00 | 1.00 | |
| Coastal neighbourhood | 2.93 (2.40 to 3.46) | 2.76 (2.23 to 3.30) | 2.74 (2.20 to 3.28) | |
| Social variables | ||||
| Family social support | ||||
| Low | 1.00 | 1.00 | ||
| Mid | 1.10 (0.78 to 1.42) | 1.08 (0.76 to 1.41) | ||
| High | 1.45 (1.09 to 1.80) | 1.41 (1.05 to 1.77) | ||
| Friends' social support | ||||
| Low | 1.00 | 1.00 | ||
| Mid | 0.93 (0.62 to 1.24) | 0.93 (0.62 to 1.24) | ||
| High | 1.25 (0.86 to 1.64) | 1.22 (0.83 to 1.61) | ||
| Personal variables | ||||
| Self‐efficacy | ||||
| Low | 1.00 | |||
| Mid | 1.32 (0.98 to 1.66) | |||
| High | 1.17 (0.83 to 1.50) | |||
| Random effects | ||||
| Area‐level | ||||
| Median OR | 1.34 | 1.00 | 1.00 | 1.00 |
Values are OR (95% CI).
*Multilevel logistic regression models show the ORs for the likelihood of walking for transport according to education and environmental, social and personal variables. The median OR reflects the level‐2 variation (ie, the variation in the likelihood of walking for transport that is between neighbourhoods rather than between individuals).