Table 1 Household risk of food under-supply in Bosnia (still short of food after receiving food aid) from multivariate analysis, 1995 and 1996
Variable | Bivariate Unadjusted OR Cornfield (95%CI) | Naïve Mantel Haenszel OR-adjusted (95%CI) | Lamothe cluster adjusted ORmh (assuming OR constant)§ (95%CIca) | GEE Exchangeable matrix OR (naïve 95%CI) | Lamothe cluster adjusted ORmh (not assuming OR constant ¤) (95%CIca) | GLMM Laplace approximation OR (95%CI) |
|---|---|---|---|---|---|---|
Household characteristics | ||||||
Disabled in household | 2.24 (1.42-3.25) | 1.52 (1.02-2.27) | ns | ns | ns | ns |
Displaced people | 2.38 (2.18-2.60) | 2.36 (2.16-2.58) | 2.29 (1.53-3.43) | 1.95 (1.56-2.44) | 2.43 (1.78-3.08) | 2.41 (2.2-2.64) |
No remittance | 1.84 (1.63-2.10) | 1.90 (1.67-2.14) | 1.60 (1.28-1.99) | 1.68 (1.33-2.09) | 1.87 (1.40-2.34) | 1.89 (1.67-2.14) |
Female headed household | 1.12 (0.97-1.28) | 1.25 (1.08-1.43) | 1.21 (1.01-1.44) | 1.27 (1.08-1.49) | 1.25 (1.03-1.48) | 1.22 (1.06-1.4) |
No employment | 1.82 (1.65-2.01) | ns | ns | ns | ns | ns |
Muslim vs. non-Muslim | 0.96 (0.88-1.05) | ns | ns | ns | ns | ns |
Cluster characteristics | ||||||
Agency2 (vs others) | 1.71 (1.55-1.89) | 1.65 (1.49-1.82) | ns | ns | 1.8 (1.10-2.50) | 1.59 (1.43-1.76) |
Republic (vs BiH) | 1.21 (1.10-1.32) | 1.37 (1.24-1.50) | ns | ns | ns | ns |
Recent frontline conflict | 0.99 (0.89-1.10) | ns | ns | ns | ns | ns |
Rural (vs urban) | 1.00 (0.88-1.13) | ns | ns | ns | ns | ns |
Number of households | n=17905 | n=17549 | n=17562 | n= | n=17562 | n=17561 |