| Stratum variable for Nigeria 2013 [message #18723] |
Wed, 05 February 2020 11:27  |
lbdrew
Messages: 7
Registered: September 2019
|
Member |
|
|
Hello,
I am using couples data from 2013. I know I should use the men's weight variable, but I have questions about the strata variable.
I have
gen wgt=mv005/1000000
I'm confused about which variable I should use for the strata.
Would it be
svyset [pw=wgt], psu(mv021) strata(mv022)
mv022 is listed as the sample stratum number
or
svyset [pw=wgt], psu(mv021) strata(mv023)
mv023 is listed as the sample domain in codebook, but the tutorial video showed v023 being used as strata variable https://www.youtube.com/watch?v=YpXPWMUsb94.
|
|
|
|
|
|
| Re: Stratum variable for Nigeria 2013 [message #18743 is a reply to message #18723] |
Mon, 10 February 2020 09:42  |
lmallick
Messages: 22
Registered: June 2019
|
Member |
|
|
Dear Laura,
Thank you for your question. While mv023 (or v023 or hv023) is typically the variable you should use, this can vary from survey to survey. Often in the dataset mv023 and v022 are the same. In this case, it is OK to use either variable. Regardless, it is always advisable to review Appendix A or B to ensure that the strata variable you are using reflects the sample design. The Nigeria 2013 survey was a 3-stage sample design and both mv022 and mv023 reflect this, so either is fine.
Here is another user forum post with a few examples of different scenarios in different surveys:
https://userforum.dhsprogram.com/index.php?t=msg&goto=11 771&S=Google
-Lindsay
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
