Echantillonnage
Méthode d'échantillonnage
The sample is drawn using a two-stage stratified cluster sampling design. Schools are stratified explicitely according to school type and province, and implicitely (through sorting) according to urbanity, district and school size (number of grade 3 students).
In the first stage, a sample of schools is selected using probability proportional to school size. For each school, two replacements schools are selected in case of non-response.
In the second stage, a grade 3 classroom is randomly selected.
Taux de réponse
Unweighted participation rates:
Schools - 0.92
Students - 0.92
Weighted participation rates:
Schools - 0.99
Students - 0.93
Pondération
There is no weight adjustment for students that were not present at the test. However, students that did not give any valid responses are considered as non-respondents, and the weights are adjusted accordingly.
Télécharger la documentation