The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Nepal » How to create a single weight when combining surveys (Trying to combine weights for women's, children's, and births surveys)
How to create a single weight when combining surveys [message #22540] Mon, 29 March 2021 14:32 Go to next message
PerrinKrisko is currently offline  PerrinKrisko
Messages: 1
Registered: March 2021
Member
Hi all,

I am new to the DHS and trying to combine the women's, births, and children's surveys because these are the files that provide details pertaining to respiratory conditions. My unit of analysis is individuals. I couldn't figure out how to combine weights and strata since each of these surveys contain a unique weight and strata variables. See below what I have so far:

* Set survey parameters based on Nepal DHS Appendix A *
* DHS Guidelines state to use variable v005 divided by 1,000,000 as the weight variable *
* DHS Guidelines state to use variable v021 as the primary sampling unit variable *
* DHS Guidelines state to use variable v023 as the stratification variable *
rename v005 female_weight
rename v021 psu
rename v023 strat_unit

gen svy_wt = female_weight/1000000
svyset psu [pweight=svy_wt], strata(strat_unit) singleunit(centered)

I could offer more information if needed about how I have cleaned the data so far or about the question I am aiming to analyze. Would someone be able to offer me advice on how to do this? Thank you!

Perrin
Re: How to create a single weight when combining surveys [message #22555 is a reply to message #22540] Wed, 31 March 2021 16:26 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3203
Registered: February 2013
Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

The weight for a child in the KR or BR file is exactly the same as the weight of the mother in the IR file. The clusters and the strata are also the same. These are not separate surveys. The lines you have for constructing svyset look good to me.

Previous Topic: Ethnicity variable in NDHS 2016
Next Topic: Staff trained for ANC
Goto Forum:
  


Current Time: Sun Dec 8 11:39:11 Coordinated Universal Time 2024