10.07.2015 Views

ICCS 2009 Technical Report - IEA

ICCS 2009 Technical Report - IEA

ICCS 2009 Technical Report - IEA

SHOW MORE
SHOW LESS
  • No tags were found...

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Class non-response adjustment (WGTADJ2S)In most countries, one class per school was selected for <strong>ICCS</strong>. Thus, non-response at the classlevel was equivalent to non-response at the school level, and any adjustments for non-responsewere conducted as described above. In a few countries, two classes were selected in some of theschools. If one of the two classes did not participate, the entire school was regarded as nonparticipating.As a consequence, the non-response adjustment was also performed at stratumlevel.However, in situations where a census of schools was taken in a stratum, classes became theprimary sampling units. In situations of class non-participation, a class weight adjustment wascomputed at the school level to correct for class non-response. The class weight adjustment foreach participating class j was calculated as:WGTADJ2S hij= c shiPc hiwhere c hisis the total number of sampled classes and c hiPis the total number of participatingclasses in school i in explicit stratum h.Student non-response adjustment (WGTADJ3S)Two different approaches to calculate student non-response adjustments were taken. Theapproach used depended on differences in non-responses between male and female students.For each country, the percentage of non-responding male students and the percentage of nonrespondingfemale students were compared. If the absolute difference in response rates betweenmale and female teachers did not exceed three percent in a country’s unweighted data, then, forall schools in this country, the adjustment for student non-response inside each class for eachparticipating student k was calculated as follows:WGTADJ3S hijk= s ehijPs hi jeHere, s is the number of eligible students and s Phiis the number of participating students inj hi jclass j in school i in stratum h. In the context of student weight adjustment, students of thetarget population were regarded as eligible if they had not been excluded due to disabilities orlanguage problems and if they had not left the sampled school after class sampling.In Liechtenstein, the overall difference in response rates between male and female students was4.5 percent with respect to the unweighted data. To take this difference in male and femalesurvey participation into account, the sampling team performed, for all schools in this country,student weight adjustment within class-gender cells:WGTADJ3S hijk=se–malehijp–males hije-females hijsp-femalehij, for participating male students, for participating female studentsHere, se–male e-femalehijand s hij are the number of eligible males and females in the class, respectively,p–male p–femaleand s hijand s hijare the number of participating male and female students, respectively, inclass j in school i in stratum h.In order to allow for the calculation of weighted exclusion rates, excluded students withinsampled classes received an adjustment of 1.72 <strong>ICCS</strong> <strong>2009</strong> technical report

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!