The study was designed to select a representative sample
of high school students at private schools in São
Paulo, Brazil. The city has 578 private high schools and
28 were included in this study. There was some degree
of non-participation at the school level, and the standard
protocol is to replace these schools with similar
schools; 68% of our sampled schools consented; the
others were replaced with similar schools in the same
stratum. The sample size was set for a maximum relative
error of 10% and a confidence interval of 95%. The
schools were stratified according to the average incomes
of the neighborhoods in which they are located (schools
in low- and high-income neighborhoods, in a rate of
2:3). In a second phase, the sample was selected by conglomerates
(schools and classes). All students in each
selected class were invited to answer the questionnaires.
The process generated a final sample of 2691 high
school students. The response rate among the students
invited to participate was 99.4% (only 16 adolescents
refused to participate in the study). The complex survey
design took into account the stratum (neighborhoods in
which schools are located), the conglomerate (school as
primary sampling unit), the expansion weight and the
probability of drawing the student who answered the
questionnaire.
The study was designed to select a representative sampleof high school students at private schools in SãoPaulo, Brazil. The city has 578 private high schools and28 were included in this study. There was some degreeof non-participation at the school level, and the standardprotocol is to replace these schools with similarschools; 68% of our sampled schools consented; theothers were replaced with similar schools in the samestratum. The sample size was set for a maximum relativeerror of 10% and a confidence interval of 95%. Theschools were stratified according to the average incomesof the neighborhoods in which they are located (schoolsin low- and high-income neighborhoods, in a rate of2:3). In a second phase, the sample was selected by conglomerates(schools and classes). All students in eachselected class were invited to answer the questionnaires.The process generated a final sample of 2691 highschool students. The response rate among the studentsinvited to participate was 99.4% (only 16 adolescentsrefused to participate in the study). The complex surveydesign took into account the stratum (neighborhoods inwhich schools are located), the conglomerate (school asprimary sampling unit), the expansion weight and theprobability of drawing the student who answered thequestionnaire.
การแปล กรุณารอสักครู่..
