The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Weighting data » Re-weighting combined (female+male) dataset
Re-weighting combined (female+male) dataset [message #677] Mon, 12 August 2013 10:35 Go to next message
Jasmine is currently offline  Jasmine
Messages: 3
Registered: August 2013
Location: Bethesda
Member
FAQs indicate there are no direct estimates from jointed population (female and male). I was considering reweighting the combined the dataset to run the regression on the usage of malaria treatment. Could you please advise me how to calculate and apply the new weights and how to get the svy set up? Thank you!
Re: Re-weighting combined (female+male) dataset [message #719 is a reply to message #677] Fri, 30 August 2013 08:25 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3016
Registered: February 2013
Senior Member
Dear Jasmine,

Your August 12 posting on the DHS forum has not been answered, so I have asked the DHS staff to respond. They would like to have more information before trying to answer.

What survey or surveys are you working with? What malaria treatments are you analyzing? When you refer to "females+males", are you referring to children or adults?

Normally, information about malaria treatment is only obtained for children, and the information is attached to either the mother's data or the household's data. There is no pooling of datasets, and no issue of re-weighting the data. Or perhaps you are pooling multiple surveys?

If you can provide this information, we will have a better chance of giving you a useful reply.

Best regards,
Bridgette-DHS
Re: Re-weighting combined (female+male) dataset [message #755 is a reply to message #719] Wed, 11 September 2013 10:17 Go to previous messageGo to next message
Jasmine is currently offline  Jasmine
Messages: 3
Registered: August 2013
Location: Bethesda
Member
Dear Bridgette,

I am sorry that I gave the wrong example on malaria treatment. I am working on the adult datasets (individual and male) for their HIV test. I am going to pool the individual(female) and male datasets together to see what the general HIV test ultilization rate is in one country. Thank you!

Best,
Jasmine
Re: Re-weighting combined (female+male) dataset [message #785 is a reply to message #755] Fri, 20 September 2013 10:48 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3016
Registered: February 2013
Senior Member
Following are comments from one of our STATA experts Tom Pullum:

We never (so far as I know) give a combined (male+female) estimate of HIV prevalence. Obtaining a pooled estimate is not trivial, partly because in some surveys the men or women are subsampled, and partly because men and women in the same household may have different weights.

We recently did a report on HIV testing (Comparative Report 30, available on the website). If there are still questions after looking at that report, please contact the first author, Sarah Staveteig.
Previous Topic: Question re-weighting combined survey data
Next Topic: Sample weight/Survey design
Goto Forum:
  


Current Time: Thu Mar 28 07:13:52 Coordinated Universal Time 2024