The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Weighting data » Apply Weights to household members (PR) files
Apply Weights to household members (PR) files [message #25967] Wed, 18 January 2023 05:04 Go to next message
joangar is currently offline  joangar
Messages: 1
Registered: January 2023
Member
I am computing indicators using sample weights (HV005) and the number of cases doesn't match the number of weighted cases. For example in Brazil 1996 there are 56755 cases in the samples and 55317 cases when I use weights. Afghanistan 2015 has 203708 cases and the frequency is 197710 if I use weights.

Thanks for your help in advance
Re: Apply Weights to household members (PR) files [message #25971 is a reply to message #25967] Wed, 18 January 2023 08:54 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 2670
Registered: February 2013
Senior Member
Following is a response from Senior DHS staff member, Tom Pullum:


The household weights are normalized so that the weighted and unweighted number of households are the same--NOT the weighted and unweighted number of individuals. The latter type of normalization is used for the IR and MR files.

The following Stata lines will confirm this for the PR file from the Brazil survey. I use hv101=1 to get one record per household. Alternatively you could use hvidx=1.

/index.php?t=getfile&id=2001&private=0
  • Attachment: hv000.jpg
    (Size: 56.72KB, Downloaded 33 times)
Previous Topic: Resampling data to obviate the need for weighting
Next Topic: Sex ratio at birth using IR file , India, NFHS-5
Goto Forum:
  


Current Time: Mon Feb 6 06:57:17 Coordinated Universal Time 2023