Hello,

I have a question regarding how to correctly declare survey design for a dataset I am analyzing.

I am using 2009-2010 Health Behavior in School-Aged Children(HBSC) survey data (United States), with the following description on sampling procedure and weights.

sampling procedure:
The sample was a nationally representative sample, collected through a three-stage stratified design, with census divisions and grades as strata, and school districts (or a group of school districts) as primary sampling units (PSUs). African American and Hispanic students were oversampled to obtain better estimates for these groups.

weights:
For producing population-based estimates of means, totals, proportions, and ratios, each responding student in the sample was assigned a sampling weight. This weight combines a base sampling weight and adjustments for nonresponse at various stages. The base sampling weight assigned to a student (STU_WT) is the inverse of the probability of selection of the student. The probability of selecting a student is the product of the probability of selecting the PSU in which the school to which the student belongs is selected, the probability of selecting the school, and the probability of selecting the class of the student. Weights for each school and district are also present on the file (in fact, the district weight is not available in dataset)

I am interested in fitting some generalized linear models at the student level. How should I specify the survey design and sampling weights in stata?

The dataset does not provide any FPC. So when I am setting up the three-stage sampling, stata says it will ignore further stages in variance estimation. Also, the student weight already accounts for the selection probability at school and district level. In this case, shall I not worry about setting up the 3 stages sampling in stata and just specify the student weight as the sampling weight (pweight)? Please help. Also, could someone provide the stata command for specifying the sampling weight in this case? Thanks much!

Jimmy