weighted sample dhs [message #24689] |
Tue, 21 June 2022 22:13 |
Samishreyaan
Messages: 6 Registered: May 2022
|
Member |
|
|
Hi there,
When i applied svy command to see the sample size for married adolescent woman for six countries of south asia (Afganistan, bangladesh,India,Nepal,Maldives and pakistan ), the sample size for afganistan , maldives, nepal and pakistan is decreased than the unweighted one. I was wondering, isnt it supposed to increase than the unweighted one? Can you please explain me what is it supposed to be in actual ? i have used commands,
tab country
svy: tab country, missing count format (%9.0f)
and the stata outputs are attached as attachments below.
kind regards,
Samikshya
|
|
|
Re: weighted sample dhs [message #24715 is a reply to message #24689] |
Mon, 27 June 2022 15:24 |
Janet-DHS
Messages: 899 Registered: April 2022
|
Senior Member |
|
|
The weights adjust for under-sampling and over-sampling in various strata and for nonresponse. For the sample as a whole, the weighted total and unweighted total are equal (ignoring the factor of 1000000). In subsamples, the weighted total is greater than the unweighted total about half the time, and less about half the time. The two are usually within about 10% of each other but the difference can be wider. I don't think you have a problem here.
|
|
|