Table 5.
Chi square automatic interaction detector gain index
| Cluster node | Node description | Node | Gain | % prompte | Indexf | ||
|---|---|---|---|---|---|---|---|
| Na | %b | Nc | %d | ||||
| A | B | C | D | E | F | G | H |
| Cluster 1 | 2059 | 32.5 | 1504 | 37.5 | 73.0 | 115.3 | |
| 14 | Mozambique and Zanzibar: children from 20 to 29 years old mothers | 199 | 3.1 | 171 | 4.3 | 85.9 | 135.6 |
| 15 | Mozambique and Zanzibar: children from 15 to 19 years old mothers and 30–49 years old mothers | 209 | 3.3 | 157 | 3.9 | 75.1 | 118.5 |
| 16 | Uganda: children from urban areas | 334 | 5.3 | 240 | 6.0 | 71.9 | 113.4 |
| 4 | Malawi, Rwanda, Senegal, and Tanzania: all children | 1317 | 20.8 | 936 | 23.3 | 71.1 | 112.1 |
| Cluster 2 | 3366 | 53.1 | 2102 | 52.3 | 62.4 | 98.5 | |
| 11 | Liberia: children from lowest and highest wealth quintiles | 303 | 4.8 | 205 | 5.1 | 67.7 | 106.8 |
| 17 | Uganda: children from rural areas | 1515 | 23.9 | 1000 | 24.9 | 66.0 | 104.2 |
| 7 | Angola, Ghana, and Zambia: children from mothers with secondary education and above | 358 | 5.6 | 220 | 5.5 | 61.5 | 97.0 |
| 13 | Liberia: children from second and third quintiles | 445 | 7.0 | 263 | 6.5 | 59.1 | 93.3 |
| 9 | Kenya and Madagascar: children from mothers with secondary education and above | 111 | 1.8 | 63 | 1.6 | 56.8 | 89.6 |
| 18 | Angola, Ghana, and Zambia: children from mothers with primary or no education; children from lowest, fourth, and highest wealth quintiles | 634 | 10.0 | 351 | 8.7 | 55.4 | 87.4 |
| Cluster 3 | 912 | 14.4 | 410 | 10.2 | 45.0 | 70.9 | |
| 12 | Liberia: children from the fourth wealth quintile | 147 | 2.3 | 71 | 1.8 | 48.3 | 76.2 |
| 19 | Angola, Ghana, and Zambia: children from mothers with primary or no education; children from second and third wealth quintiles | 388 | 6.1 | 176 | 4.4 | 45.4 | 71.6 |
| 10 | Kenya and Madagascar: children from mothers with primary or no education | 377 | 5.9 | 163 | 4.1 | 43.2 | 68.2 |
| Total | 6337 | 100 | 4016 | 100 | 63.4 | n/a | |
N number of children
aNumber of children who received any anti-malarial treatment per node (demographic size in the sample)
bDemographic size in percentage = (0.1/Σ0.1) × 100
cNumber of children who received prompt treatment with ACT
dDemographic size among children who received prompt treatment with ACT in percentage = (0.3/Σ0.3) × 100
eProportion of children who received prompt treatment with ACT out of those who received any anti-malarial medicine = (0.3/Σ0.1) × 100
fNode index = [(0.3/Σ3)/(0.1/Σ0.1)] × 100