Predictors | Cameroon | Côte d’Ivoire | Sierra Leone | |||
---|---|---|---|---|---|---|
Empty model | Full model | Empty model | Full model | Empty model | Full model | |
Sociodemographic characteristics and message exposure | ||||||
 Sex | – |  | – |  | – |  |
  Male (RC) |  | 1.000 |  | 1.000 |  | 1.000 |
  Female |  | 3.642d |  | 1.623c |  | 1.258 |
 Age in years | – | 1.045d | – | 1.031d | – | 1.105b |
  Square of age |  | – |  | – |  | 0.998b |
 Education level | – |  | – |  | – |  |
  None (RC) |  | 1.000 |  | 1.000 |  | 1.000 |
  Primary |  | 1.261 |  | 0.621d |  | 0.568c |
  Secondary +  |  | 1.337 |  | 0.434d |  | 0.589c |
 Exposure to malaria-related messages in past six months | – | 0.586b | – | 1.489c | – | 1.598c |
Ideational characteristics | ||||||
 Positive attitudes towards net use | – | 3.171d | – | 7.593d | – | 1.455b |
 Perceived severity of malaria | – | 1.361 | – | 0.874 | – | 1.115 |
 Perceived vulnerability to malaria | – | 2.322d | – | 0.731b | – | 1.658c |
 Discussed malaria with others | – | 0.874 | – | 2.209d | – | 0.539d |
 Perceived response efficacy of ITNs | – | 0.721 | – | 2.005d | – | 2.013d |
 Perceived self-efficacy to use nets | – | 7.026d | – | 3.951d | – | 3.246d |
 Perceived net use as a community norm | – | 2.891d | – | 1.319b | – | 1.168 |
Household and community characteristics | ||||||
 Household has at least one child under age 5 | – | 2.133c | – | 2.035d | – | 1.105 |
 Household wealth quintile | – |  | – |  | – |  |
  Lowest (RC) |  |  |  |  |  | 1.000 |
  Second |  | 1.000 |  | 1.000 |  | 1.188 |
  Middle |  | 2.152b |  | 0.612b |  | 1.184 |
  Fourth |  | 4.050c |  | 0.517b |  | 0.935 |
  Highest |  | 6.519d |  | 0.344d |  | 0.934 |
 Net supply within household | – |  | – |  | – |  |
  Insufficient (RC) |  | 1.000 |  | 1.000 |  |  |
  Sufficient |  | 2.978d |  | 3.044d |  | 2.023c |
 Dwelling unit has completely sealed ceiling | – | 0.972 | – | 0.578c | – | 1.091 |
 Urban residence (RC = rural) | – | 0.532 | – | 0.921 | – | 0.459a |
 Region of residence | –– | 1.000 |  |  |  |  |
  Far North (RC) |  | 0.067d |  |  |  |  |
  North |  |  |  |  |  |  |
 Zone of Residence |  |  | – |  |  |  |
  North (RC) |  |  |  | 1.000 |  |  |
  Central |  |  |  | 1.332 |  |  |
  South |  |  |  | 1.255 |  |  |
  Abidjan |  |  |  | 0.086d |  |  |
 District of residence |  |  |  |  | – |  |
  Bo (RC) |  |  |  |  |  | 1.000 |
  Port Loko |  |  |  |  |  | 0.347c |
Random effects | ||||||
 Household level variance (SE) | 22.792 (6.696) | 14.761 (3.142) | 13.014 (1.8329) | 11.047 (1.568) | 1.833 (0.397) | 2.114 (0.476) |
 Cluster level variance (SE) | 3.419 (0.857) | 1.971 (0.576) | 2.909 (0.576) | 0.344 (0.143) | 3.488 (0.687) | 2.537 (0.560) |
 LR Test (multilevel vs. logistic): Χ2/p | 541.9/0.000 | 393.2/0.000 | 890.5/0.000 | 501.2/0.000 | 645.9/0.000 | 354.0/0.000 |
 Intraclass correlation (Household) | 0.888 | 0.835 | 0.829 | 0.776 | 0.618 | 0.586 |
 Intraclass correlation (Cluster) | 0.116 | 0.098 | 0.151 | 0.023 | 0.405 | 0.319 |
 Proportional change in variance (Household) | – | 35.2% | – | 15.1% | – | − 15.3% |
 Proportional change in variance (Cluster) |  | 42.3% | – | 88.2% |  | 27.3% |
 Number of observations | 2995 | 6164 | 2730 |