The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » General » Aggregating up to district level in NFHS-4
Aggregating up to district level in NFHS-4 [message #20319] Thu, 22 October 2020 09:27 Go to next message
zoeg is currently offline  zoeg
Messages: 6
Registered: October 2020

I am trying to merge the NFHS-4 dataset with another dataset that contains variables at the district-level. Are there district weights somewhere in the NFHS-4 dataset that will allow me to aggregate up the individual-level responses to the district level? Has someone else done this and knows how to do it in Stata?

Thank you!
Re: Aggregating up to district level in NFHS-4 [message #20386 is a reply to message #20319] Mon, 26 October 2020 17:58 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 2550
Registered: February 2013
Senior Member

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

I don't know which file you want to use, but let's say it's the PR file (IAPR74FL.dta). In that file, district is given by shdistri. The numerical codes for the names will be given if you enter "label list SHDISTRI". In your district-level file, be sure that the district ID has the same name and the same codes. Sort that file by shdistri and save it. Then open the PR file and sort it by shdistri. Then merge with the saved district-level file. There are many examples of merges on the forum.

This will attach the district-level variables to the individual cases in the PR file. You can then treat those variables as covariates. You can collapse the file (try "help collapse") but that's not necessary. Ideally, you would do the analysis with multi-level models.
Previous Topic: Question of STATA
Next Topic: Merge Service Availability data with individual data for zimbabwe
Goto Forum:

Current Time: Mon Aug 15 14:31:31 Coordinated Universal Time 2022