The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Weighting data » Cluster level weights for NHFS-5 India (Cluster weights)
Cluster level weights for NHFS-5 India [message #30485] Sun, 08 December 2024 18:05 Go to next message
rghosh is currently offline  rghosh
Messages: 2
Registered: October 2024
Member
India

National Family Health Survey - 5

Hi

I am planning to fit a multilevel level consistent with the data hierarchy of the NFHS-5 survey in India. As far as I can understand, the sampling hierarchy was districts - clusters - households - individual participants. Keeping aside the stratification for now. I have the individual and household weights from various files. Is there a source for the cluster level weights to use in the analysis? I have been reading about cluster weights and the posts range over a decade. Not sure if I ran into the most updated post. Has there been any recent update on cluster weights or do we not have sufficient data to calculate these. Any suggestion will be greatly appreciated.

Best
Rakesh
Re: Cluster level weights for NHFS-5 India [message #30493 is a reply to message #30485] Mon, 09 December 2024 08:25 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3230
Registered: February 2013
Senior Member

Following is a response from Senior DHS staff member, Tom Pullum:

Unfortunately, we still do not have the information to construct cluster-level weights for NFHS-4 and -5. I still hope that some user will come forward and help out. The necessary information for NFHS-5 was briefly posted (as one or two excel spreadsheets) on a website in India and then was unexpectedly removed before DHS had copied it. (There was nothing in the information that would have compromised the confidentiality of the sample.) I believe that a handful of users may have copied it. If anyone reading this can help out, please let me know!

Re: Cluster level weights for NHFS-5 India [message #31100 is a reply to message #30485] Fri, 27 June 2025 09:12 Go to previous message
StephanieBaldwin is currently offline  StephanieBaldwin
Messages: 1
Registered: June 2025
Member
The census or sampling frame they used (usually from Census Solitaired 2011) are not fully public at the PSU level for you to determine cluster selection probabilities → thus it is not possible to calculate exact cluster weights for each PSU yourself.

[Updated on: Fri, 27 June 2025 20:06]

Report message to a moderator

Previous Topic: Question about weights to use
Goto Forum:
  


Current Time: Wed Jul 9 09:56:55 Coordinated Universal Time 2025