Stata has a whole suite of statistical commands designed for survey data.
If you type:
. whelp svy
You will get Stata help window with a list of all the survey commands where you can
click on them and get help for them.
Before using any command you will need to tell Stata what your strata
and your psu variables are with the -svyset- command. Once this is set in your
Stata session any survey command after it will use these strata and psu variables.
. svyset , strata(district_id) psu(school_id)
My example assumes you do not have a sample weight. If you do then the command would
look like this:
. svyset [pweight=sample_wt], strata(district_id) psu(school_id)
Then you can look at your means for example:
. svymean age gpa income
Dan Blanchette
> From rghuang <[email protected]>
> To "[email protected]" <[email protected]>
> Subject st: help for analyse survey data with complex sampling design
> Date Wed, 24 Mar 2004 20:16:28 +0800
>
> hi ,listers
> recently ,i carry out a survey and i hope any of you can give me some advice
> on how i can analyze my data with my own complex sampling design.
> my samling design is as follows:
> the population is all the students in 7 districs(the population id quite big).
> we divide 7 districs in the 3 stratum(each distric have different amount of schools
> and students), then sample one distric each strata,let us name it A,B,C.we can know
> the number of sudents in A,B and C.and A,B,C has different numbers of sub-districs,
> then we sample the 8 sub-distrcis from each randomly.and then we sample randomly
> about 5 schools from each sampled sub-distric.at last we sample about 100 students
> from each sampled school and get the sample that we want.it is so design considering
> the feasibility.we have the information about the amount of sub-districs in each
> distric, the amount of schools in each sub-dsitrics, and the aount of students in
> each school.
> in this case,how should i analyse the survey data ? and exactly,which varibles
> should included in data set,and how to use your program?
> thank you in advance!
> regards!
>
> [email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
