The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » General » Weithted analysis in R
Re: Weithted analysis in R [message #29863 is a reply to message #29825] Wed, 14 August 2024 10:33 Go to previous messageGo to previous message
Mahir is currently offline  Mahir
Messages: 12
Registered: September 2023
Member
Dear DHS team,

Thank you for your response.

I promise might be the last time I come back with question regarding sampling weight.

the reason why I am asking for denormalising weight because I want to do poolsed analysis of eight countires I have mentioned. I have gone through almost all the questions on sampling weight on this forum now. I also found a manual by DHS (attaching here) which recommends denormalizing weights for doing pooled analysis using the formula given. I using the exact same formula that the manual is sharing. I will share the example of DHS Benin KR recode (2017-18).

I apply fthe fooolwing code in

KRdata$denorm<-(KRdata$v005 * (2717666/13589)))/1000000
#2717666 are the total number of woemn aged 15-59 at the time of the survey. I got this data from this site ( https://platform.who.int/data/maternal-newborn-child-adolesc ent-ageing/indicator-explorer-new/MCA/women-of-reproductive- age-(15-49-years)-population-(thousands)) and 13589 is the total number of respondents in the KR file

Now ideally with the sum of all the values of the new weight (denrom) should be 2717666 but that is not the case, the total comes out to be 2809377, this is a huge discrepency.

Would you be able to explain why this mgiht be happening? Is there a way to resolve this? I am afraid this same problem will occur with rest of the seven countires (Cameroon 2018, Cote d'Ivoire 2021, Ghana 2022, Kenya 2018-19, Liberia 2022, Nigeria 2021 and Uganda 2016)

Best
Mahir
 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic: Child Marriage
Next Topic: WEIGHTS AFTER DROP OUTS
Goto Forum:
  


Current Time: Thu Nov 28 09:15:19 Coordinated Universal Time 2024