| NFHS - 5 [message #24434] |
Sun, 15 May 2022 02:12  |
neera.vohra@competitivene
Messages: 1
Registered: May 2022
|
Member |
|
|
Hello,
I wanted to extract district-wise data on the percentage of women who have a bank account from the Household members (or Persons) Recode state file. I am not sure whether I have set the syntax right; I need clarity on the same :
gen pwt=.
replace pwt= hv005/1000000
svyset hv021 [pw=pwt], strata(hv023)
Since the NFHS-5 sample is a stratified two-stage sample. therefore shouldn't the syntax look like :
svyset hv021 [pw=pwt], strata(hv023) || hhid , strata(hv001)
My understanding is that in the second stage: about 25-30 households are looked after, identifying the villages ( in the first stage)
However, I have no experience working on NFHS data sets, so I am not sure what's the right thing to do. Kindly help me and respond to it as soon as anyone can.
Thank you for your help.
|
|
|
|
| Re: NFHS - 5 [message #24471 is a reply to message #24434] |
Thu, 19 May 2022 10:30  |
Janet-DHS
Messages: 687
Registered: April 2022
|
Senior Member |
|
|
Following is response from DHS Research & Data Analysis Director, Tom Pullum:
First, for pweights you do not need to divide hv005 by 1000000. Many people do this, but it is not necessary because Stata automatically re-scales pweights to have a mean of 1.
Second, you do not need a multi-level version of svyset. Your first version of svyset is sufficient.
Third, the clusters and strata were numbered separately in each state. You need commands such as "egen cluster_id=group(v021 v024)" and "egen stratum_id=group(v023 v024)".
Then "svyset cluster_id [pw=hv005], strata(stratum_id)" should be correct for svyset.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
