| Weighting of pooled country and year data [message #30671] |
Fri, 17 January 2025 10:15  |
Christiaan
Messages: 1
Registered: January 2025
|
Member |
|
|
Dear DHS community
I have been performing analyses using DHS data pooled over various countries and years. I would like to weight this data accordingly. I have been reading up on how this should be done and I came across the following thread:
https://userforum.dhsprogram.com/index.php?t=msg&goto=99 82&S=Google
What interests me in particular is the first response by DHS Stata Specialist Shireen Assaf:
gen wt= v005/1000000
egen strata=group(v000 v025 ADM1_CODE) // strata also includes the survey (identified by v000) in the group command
egen v001r = group(v000 v001) // cluster also includes the survey in the group command
svyset v001r [pw=wt], strata(strata) singleunit(centered)
svy: tab ADM1_CODE
Now, I would like to display mean values of certain outcome variables by birth cohort. How would I weight this appropriately, drawing on the code provided above?
gen wt= v005/1000000
egen strata=group(v000 country v007 birth_cohort) // Where v007 is the year in which the survey was conducted and birth_cohort is the cohort in which the observation was born.
egen v001r = group(v000 country)
svyset v001r [pw=wt], strata(strata) singleunit(centered)
svy: sum outcome_var, by(birth_cohort)
Would this approach to weighting make sense? If not, how can I improve upon it? This is the first time that I am using weights and I would just like to make sure that I do it correctly.
Thank you so much for your kind assistance!
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
