TABLE 4—
Average Marginal Effects of Nonsmoking Adolescents’ Exposure to Secondhand Smoke Outside the Home: Global Youth Tobacco Survey, 9 West African Countries, 2006–2009
Characteristic | Cape Verde (n = 970), % | Cote d’Ivoire (n = 1382), % | Ghana (n = 3154), % | Guinea (n = 938), % | Mali (n = 1593), % | Mauritania (n = 1217), % | Niger (n = 904), % | Senegal (n = 1124), % | Togo (n = 1610), % |
Gender | |||||||||
Male | 0.3 | −1.5 | 0.9 | 7.3* | −2.4 | 1.3 | 1.0 | −2.0 | 0.9 |
Female (Ref) | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
Age, y | |||||||||
13 (Ref) | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
14 | −1.1 | 2.6 | −0.2 | 2.0 | 0.2 | −2.9 | 1.0 | −1.1 | 4.3 |
15 | 3.5 | 4.9 | −0.3 | −4.6 | 3.8 | −5.9 | 2.2 | 2.4 | 1.3 |
Parent smokes | |||||||||
Yes | 16.9*** | 9.7** | 19.7*** | 12.8** | 3.2 | 6.8 | 12.5 | 13.6*** | 23.0*** |
No (Ref) | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
Peer smokes | |||||||||
Yes | 18.9*** | 8.6** | 8.8*** | 10.2* | 11.6*** | 18.9*** | 21.1*** | 6.1 | 18.5*** |
No (Ref) | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
Knowledge of smoking harm | |||||||||
Yes | 5.4 | 5.9 | 8.7*** | 13.8*** | 11.8*** | 15.5*** | 10.4* | 17.8*** | 5.5* |
No (Ref) | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
Support smoking ban | |||||||||
Yes | 4.7 | 5.7 | 4.8** | 15.4*** | 8.0** | 11.3*** | 3.6 | −4.3 | 6.4 |
No (Ref) | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
Exposed to antismoking media message | |||||||||
Yes | 6.3* | 11.3*** | 13.4*** | 5.1 | 4.0* | 13.8*** | 10.9** | 10.6** | 7.3* |
No (Ref) | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
Exposed to antismoking education in school | |||||||||
Yes | −0.6 | 2.0 | 6.3*** | 1.9 | 10.9*** | 6.0* | 10.4** | 7.4* | 4.6 |
No (Ref) | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
H-L χ2(8)a | 1.8 (0.9) | 10.3 (0.2) | 14.5 (0.1) | 13.4 (0.1) | 27.2 (0.0) | 8.7 (0.4) | 13.2 (0.1) | 23.8 (0.0) | 13.4 (0.1) |
Note. Average marginal effects were estimated from the multivariable logistic regression and converted to percentages.
H-L χ2(8) shows the results of the Hosmer–Lemeshow test for the goodness of fit of the model with 10 groups. The values in parentheses indicate Prob > χ2.
*P < .05; **P < .01; ***P < .001.