Log-linear modelling
Model |
c 2
(df) |
p-value |
missing |
95% CI |
1 |
66.7 (3) |
0.00 |
1330 |
1179-1493 |
2a |
54.1 (2) |
0.00 |
1637 |
1387-1919 |
2b |
45.4 (2) |
0.00 |
1152 |
1003-1315 |
2c |
48.2 (2) |
0.00 |
1505 |
1319-1709 |
3a |
43.5 (1) |
0.00 |
1293 |
1041-1595 |
3b* |
1.5 (1) |
>0.05 |
2907 |
2285-3700 |
3c |
34.8 (1) |
0.00 |
1298 |
1111-1507 |
* = selected model
|
Use GLIM software for modelling. The dependent variable is the
observed number of cases in each cell, and the independent variables
are categorical variables representing presence or absence on each
of the lists. The model selected will be used to estimate the missing
value.
|