Page 20 - nhms2012 Perlis

PERLIS GSHS 2012 Methods

2.0 METHODS

The 2012 Perlis GSHS employed a two-stage cluster sampling design to produce a representative
sample of students in Forms 1 to 5.The _rst-stage sampling frame consisted of all schools containing
any of Forms 1 to 5. Schools were selected with probability proportional to school enrolment size. A
total of 14 schools were selected to participate in the Perlis GSHS. The second stage of sampling
consisted of systematic random sampling of selected classrooms from each participating schools. All
classrooms in each selected school were included in the sampling frame. All students in the sampled
classrooms were eligible to participate in the GSHS.

A weighting factor was applied to each student record to adjust for non-response and for the varying
probabilities of selection. The weight used for estimation is given by:

W = W1 * W2 * f1 * f2 * f3
W1 = the inverse of the probability of selecting the school
W2 = the inverse of the probability of selecting the classroom within the school
f1 = a school-level non-response adjustment factor calculated by school size
category (small, medium, large). The factor was calculated in terms of
school enrollment instead of number of schools.
f2 = a student-level non-response adjustment factor calculated by class
f3 = a post-strati_cation adjustment factor calculated by class

The weighted results can be used to make important inferences about the priority health-risk
behaviours and protective factors of all students in Forms 1 to 5.

For the 2012 Perlis GSHS, 1,560 questionnaires were completed in 14 schools.The school response rate
was 100%, while the students response rate was 86.3%. Overall, the response rate was 86.3%.

The data set was cleaned and edited for inconsistencies. Missing data were not statistically imputed.
Software that takes into consideration the complex sample design was used to compute prevalence
estimates and 95% con_dence intervals. GSHS data is representative of all students attending Forms
1 to 5 in Perlis.

Data collection was conducted from 23 February to 26 April 2012. Approvals from both the Ministry
of Health Research and Ethics Committee and Ministry of Education Ethics Committee were obtained
prior to the survey implementation. Following that, approval from relevant Ministry of Education
o^cials at state, district and selected school levels were obtained. Parental consent forms were
distributed to all students from selected classes and non-consented students were considered as
non-response.

Survey procedures were designed to protect student privacy by allowing for anonymous and
voluntary participation. The students completed the self-administered questionnaire during two
classroom periods and recorded their responses directly on a computer-scannable answer sheet.
A team of data collectors was formed and specially trained to conduct the GSHS. The data collectors
included temporary sta] and o^cials from the Institute for Public Health and Institute for Health
Behavioural Research, Ministry of Health Malaysia.

6