Table 2 |
|||||
| Results of the weighted ANOVAs of qualitative moderator variables on the lifetime prevalence estimates | |||||
| Moderator variable | k | p+ | 95% | C. I. | ANOVA results |
| pl | pu | ||||
| Continent: | |||||
| Europe | 21 | 0.390 | 0.314 | 0.472 | QB(4) = 1.95, p = .745; R2 = 0.0 |
| North America | 3 | 0.455 | 0.255 | 0.670 | QW(25) = 4227.25, p < .001 |
| Oceania | 1 | 0.403 | 0.128 | 0.757 | |
| Africa | 2 | 0.277 | 0.114 | 0.532 | |
| Asia | 3 | 0.490 | 0.285 | 0.698 | |
| Origin of the sample: | |||||
| School | 25 | 0.398 | 0.333 | 0.467 | QB(3) = 0.87, p = .833; R2 = 0.0 |
| Community | 2 | 0.396 | 0.197 | 0.638 | QW(26) = 3833.03, p < .001 |
| University | 2 | 0.470 | 0.249 | 0.703 | |
| School + sport | 1 | 0.280 | 0.085 | 0.619 | |
| Data extraction method: | |||||
| Questionnaire | 23 | 0.411 | 0.354 | 0.471 | QB(1) = 0.36, p = .546; R2 = 0.143 |
| Interview | 4 | 0.365 | 0.243 | 0.507 | QW(25) = 2345.19, p < .001 |
k: number of studies. p+: mean prevalence. pl and pu: lower and upper confidence limits of the 95% confidence interval around the mean prevalence. QB: between-categories statistic to test the influence of the moderator variable on the prevalence rates. QW: within-categories statistic to test the model misspecification. R2: proportion of variance accounted for by the moderator variable.
Calvo-Muñoz et al. BMC Pediatrics 2013 13:14 doi:10.1186/1471-2431-13-14
