The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Malaria » household member level data
household member level data [message #2681] Thu, 07 August 2014 07:49 Go to next message
namsgfr is currently offline  namsgfr
Messages: 9
Registered: March 2013
Location: Kampala
Member

Dear DHS,
I'm doing a sub analysis of the Nigeria 2013 data, successfully managed to reproduce household level estimates. I'm embarking on defacto member level estimates e.g. %defacto members that slept under an ITN the previous night.
From the datasets, i'm thinking the HR6 file is what i need to use. however i see it is reshaped wide, up to 35 members. I'm thinking of reshaping it long again and then do the analysis on all indicators whose denominator is defacto hh members
1. Is it the right approach, i'm i using the right dataset?
2. Will i use the same sample weights (hv005)when it is reshaped? i would imagine i need to take into consideration clustering of members in the household?

thanks
Geoffrey Namara
Malaria Consortium
Africa Regional Office
Re: household member level data [message #3143 is a reply to message #2681] Wed, 22 October 2014 17:04 Go to previous messageGo to next message
Liz-DHS
Messages: 1516
Registered: February 2013
Senior Member
Dear Geoffrey,
Here is a response from one of our experts, Cameron Taylor:
Hi Geoffrey,
To answer your question, yes, you need to be using a different file. Since your unit of analysis (case) is at the household member level your should be using our household member file (also referred to as PR file). This dataset has one record for every household member. You will also be using the hv005 weight when working in the PR file. You can read more about sample weights in DHS datasets under Step 7 http://dhsprogram.com/data/Using-Datasets-for-Analysis.cfm. Finally, to restrict your data set to only defacto household members you should use the variable hv103. Hope this answers your question!
Example Stata Code
use " NGPR6AFL.DTA"
g wgt = hv005/1000000
g itn=(hml12==1|hml12==2)
tab itn if hv103==1 [iweight=wgt]
Re: household member level data [message #3146 is a reply to message #3143] Thu, 23 October 2014 02:46 Go to previous message
namsgfr is currently offline  namsgfr
Messages: 9
Registered: March 2013
Location: Kampala
Member

Thanks, yes it does.
Previous Topic: Clustering Respondents in Child Data for Analysis
Next Topic: sample methodology
Goto Forum:
  


Current Time: Thu Mar 28 07:05:17 Coordinated Universal Time 2024