| Pooled weights for men and women sample. [message #31653] |
Thu, 11 December 2025 00:03 |
ask_ans_sri
Messages: 3
Registered: June 2018
Location: DELHI
|
Member |
|
|
I am analysing Indian dataset. I want to estimate a combined prevalence for the women and men, for example overall prevalence of hypertension. Since women are oversampled as compared to men, I am restricting my analysis to state level.
All the tables in the NFHS reports the weighted estimates, I have matched my estimates with the report for both men and women. However, for overall prevalence, should I just append the data and generate weight variable as follows:
gen wt=.
replace wt=v005 if sex==1
replace wt=mv005 if sex==2
gen hypertension [aw=wt/1000000]
Is this an appropriate method? or is there any other way?
Thanks for your time.
|
|
|
|
Members
Search
Help
Register
Login
Home
