| Weighting for the Pakistan Special 2019 DHS on Stata [message #24986] |
Sat, 13 August 2022 12:04  |
kgeorg7
Messages: 8
Registered: August 2022
|
Member |
|
|
Hi, I am working with the Pakistan Special 2019 DHS in Stata, I've combined the individual, household and services datasets with household and services data being applied to the individual level.
I am currently trying to do the weighting for the dataset. From what I understand from the final report, the Primary Sampling Unit (PSU) is the household (qhnumber) with the weight variable being qweight. For the strata variable, the final report states that 16 sampling strata were created with urban/rural for each of the eight regions. However, there is no specific strata variable, as such I was wondering if I should be creating a new strata variable to account for the 16 strata? Furthermore, would also appreciate feedback regarding my understanding of the PSU and weighting variables as well as whether my approach is appropriate considering I've combined the individual, household, and services dataset.
The code I'll be using: svyset qhnumber [pweight = qweight], strata(newstratavar) vce(linear) singleunit(center)
Any help will be appreciated, thank you!
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
