The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Weighting data » Pooled Cross sections
Re: Pooled Cross sections [message #10084 is a reply to message #10080] Thu, 23 June 2016 15:18 Go to previous messageGo to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3040
Registered: February 2013
Senior Member
Following is a response from DHS Stata Specialist, Shireen Assaf:

When you compute the NW the magnitude of the weight is increased by a factor of 1000 because of the CSW. It seems you need to divide the NW by 1000, i.e.:

gen NW2=NW/1000
tab ADM1_CODE [iw=NW2]

If you use svy with pweights the percentage estimates would not be affected by using NW or NW2 just the frequencies. So if you just report the percentages this shouldn't matter and no need to divide the NW by 1000, just use svy and pweights.

I would also suggest not using NW and simply use the weights for the appended data as follows.

gen wt= v005/1000000

egen strata=group(v000 v025 ADM1_CODE) // strata also includes the survey (identified by v000) in the group command

egen v001r = group(v000 v001) // cluster also includes the survey in the group command

svyset v001r [pw=wt], strata(strata) singleunit(centered)
svy: tab ADM1_CODE

You will notice that the percentages are quite close to those produced when using svy and NW or NW2.

svyset v001r [pw=NW], strata(strata) singleunit(centered)
svy: tab ADM1_CODE

 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic: Antenatal visit with skilled providers (Ethiopian DHS 2000, 2005, 2011 AND 2016))
Next Topic: Using R to weight data for domestic violence
Goto Forum:
  


Current Time: Fri Apr 26 01:33:33 Coordinated Universal Time 2024