The DHS Program User Forum: India » Taking slum into account for sampling (NFHS4)

Home » Countries » India » Taking slum into account for sampling (NFHS4)

Show: Today's Messages :: Show Polls :: Message Navigator

Taking slum into account for sampling (NFHS4) [message #14036]

Mon, 05 February 2018 16:05

omogeonago
Messages: 5
Registered: May 2015

Member

I am working with the the new NFHS 4 dataset. I am trying to figure out the correct svyset.
I used this syntax
egen strata = group(v024 v025)
svyset v021 [pw=v005/1000000], strata(strata) singleunit(centered)

However, the recode documentation indicates that the survey was "weighted by state and urban/rural, and within major cities by slum/non-slum"

In that case, should I be using this syntax instead?

gen slum=sslumc==1|sslumo==1
egen strata = group(v024 v025 slum)
svyset [pw=v005/1000000], psu(v021) strata(strata) singleunit(centered)

Thank you for your help

Report message to a moderator

Re: Taking slum into account for sampling (NFHS4) [message #14096 is a reply to message #14036]

Thu, 15 February 2018 09:11

Bridgette-DHS
Messages: 3230
Registered: February 2013

Senior Member

Following is a response from DHS Senior Sampling Specialist, Mahmoud Elkasabi:

The stratification in the NHFS4 is so complicated; within each rural districts we had 3 sub-strata and within urban districts we have slum/non-slum sub-stratification. Anyway, any analysis should consider HV022 or HV023 as the stratification variable. The user does not need to recreate the stratification variable.

Report message to a moderator

Previous Topic:	Merging child and Household recode (2015-2016 India survey)
Next Topic:	Women's Questionnaire

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sun Jan 25 14:40:29 Coordinated Universal Time 2026