Table 2 Multivariate logistic regression analysis of characteristics associated with changed dwelling type between baseline and follow-up survey
From: On the impact of nonresponse in logistic regression: application to the 45 and Up study
Baseline characteristics | Complete case | Complete case with IPW | Selection model with λ = 0 |
|---|---|---|---|
Gender | |||
 Male | Ref | Ref | Ref |
 Female | 1.156 (0.92–1.46) | 1.102 (0.86–1.41) | 1.158 (0.93–1.45) |
Age (yrs) | |||
 45–54 | Ref | Ref | Ref |
 55–64 | 5.369 (2.29–12.60) | 6.395 (2.62–15.60) | 5.371 (2.47–11.55) |
 65–74 | 16.301 (6.83–38.93) | 21.294 (8.53–53.17) | 16.445 (7.30–36.31) |
 75+ | 43.021 (17.80–104.00) | 58.047 (23.11–145.82) | 43.598 (19.01–97.32) |
Marital status | |||
 Single | Ref | Ref | Ref |
 Married/de facto | 0.806 (0.49–1.33) | 0.773 (0.45–1.34) | 0.816 (0.49–1.41) |
 Widowed/divorced/separated | 0.608 (0.36–1.04) | 0.613 (0.35–1.09) | 0.613 (0.36–1.09) |
Work status | |||
 FT/self-employed | Ref | Ref | Ref |
 PT | 2.235 (1.20–4.16) | 2.455 (1.29–4.68) | 2.189 (1.15–4.15) |
 Fully retired | 3.710 (2.25–6.13) | 3.816 (2.24–6.49) | 3.717 (2.20–6.16) |
 Partially retired | 2.690 (1.32–5.49) | 2.409 (1.15–5.06) | 2.591 (1.22–5.28) |
 Disabled/sick | 1.754 (0.51–6.05) | 2.400 (0.64–9.03) | 1.484 (0.32–4.90) |
 Unemployed/look after home/study/unpaid | 2.386 (1.18–4.83) | 2.604 (1.25–5.43) | 2.308 (1.11–4.60) |
Dwelling type | |||
 House/house on farm | Ref | Ref | Ref |
 Flat/unit/apart. | 1.522 (1.13–2.04) | 1.485 (1.09–2.03) | 1.511 (1.12–2.02) |
 Mobile home/other | 1.929 (1.02–3.66) | 1.797 (0.94–3.44) | 1.866 (0.93–3.51) |
Functional limitation (fl) | |||
 No fl | Ref | Ref | Ref |
 Slight fl | 1.702 (1.30–2.23) | 1.704 (1.29–2.26) | 1.703 (1.30–2.24) |
 Moderate fl | 1.301 (0.88–1.91) | 1.318 (0.87–1.99) | 1.287 (0.86–1.90) |
 Significant fl | 2.060 (1.39–3.06) | 2.120 (1.40–3.22) | 2.040 (1.37–3.03) |
 Severe fl | 1.917 (1.15–3.20) | 1.934 (1.12–3.35) | 1.885 (1.10–3.10) |