The DHS Program User Forum: Weighting data » Regarding de-normalizing and weighting procedures in Stata

Home » Data » Weighting data » Regarding de-normalizing and weighting procedures in Stata

Show: Today's Messages :: Show Polls :: Message Navigator

Re: Regarding de-normalizing and weighting procedures in Stata [message #9655 is a reply to message #9643]

Fri, 29 April 2016 11:03

Bridgette-DHS
Messages: 3230
Registered: February 2013

Senior Member

Following is a response from Senior DHS Stata Specialist, Tom Pullum:

The first procedure, without svy but including pweight, is sufficient.

For this purpose, you did not need to denormalize the weights. If you try it again, using [pweight=v005] rather than [pweight=wgt], you will again get the same result. With wgt, the numerator of the mean will be weighted up to the population level and the denominator will also be weighted up to the population level, but the numerator divided by the denominator will be exactly the same as if you had not weighted up.

The pweight command in Stata will always RE-normalize the weights. That's as it should be.

Report message to a moderator

[Message index]

		Regarding de-normalizing and weighting procedures in Stata By: AmsP on Mon, 11 April 2016 06:47
		Re: Regarding de-normalizing and weighting procedures in Stata By: Bridgette-DHS on Mon, 18 April 2016 11:21
		Re: Regarding de-normalizing and weighting procedures in Stata By: AmsP on Thu, 28 April 2016 07:03
		Re: Regarding de-normalizing and weighting procedures in Stata By: Bridgette-DHS on Fri, 29 April 2016 11:03
		Re: Regarding de-normalizing and weighting procedures in Stata By: AmsP on Fri, 29 April 2016 15:26
		Re: Regarding de-normalizing and weighting procedures in Stata By: Bridgette-DHS on Mon, 09 May 2016 06:29

Previous Topic:	Sampling weights, cluster, strata in AIDS recode (AR)
Next Topic:	Weighting in pooled data

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Fri Oct 24 03:39:19 Coordinated Universal Time 2025