The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Weighting data » Ghana 2008 Birth Recode Weights
Ghana 2008 Birth Recode Weights [message #2814] Wed, 27 August 2014 17:42 Go to next message
lkohler is currently offline  lkohler
Messages: 2
Registered: August 2014
Member
Hello,

I'm using the birth recode data set for Ghana 2008. I've generated the appropriate sampling weights in STATA and am using svy: for descriptive statistics. I'd like to get denominators for how many children/mothers/households are in my reduced data set. When I sort by household then respondent line number, I notice that some children in the same household (with even the same mother) have different sampling weights (v005). Can anyone explain this?

Thanks!!
Lindsay
Re: Ghana 2008 Birth Recode Weights [message #2836 is a reply to message #2814] Fri, 29 August 2014 12:01 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3016
Registered: February 2013
Senior Member
Following is a response from Senior Data Processing Specialist, Noureddine Abderrahim:

The weights in DHS are generally created for the enumeration areas. Some enumeration areas have the same sampling weights and other do not. The weight variables are: HV005 for the household, V005 for women, MV005 for men, and D005 for women interviewed in the domestic violence module.

All these weights should be divided by 1000000 ( 1 million). All individuals belonging to the same enumeration area would have the same weight. My assumption is that you must have made a mistake in computing the weights.
Re: Ghana 2008 Birth Recode Weights [message #2838 is a reply to message #2836] Fri, 29 August 2014 16:25 Go to previous messageGo to next message
lkohler is currently offline  lkohler
Messages: 2
Registered: August 2014
Member
The variable v005 from the original Birth Recode dataset has different sampling weights for observations (children) within a single household or even mother. This is not something I computed. I would think children of the same mother would have the same sampling weight or even children of different mothers in the same household? Thanks for any further clarification.

[Updated on: Fri, 29 August 2014 16:42]

Report message to a moderator

Re: Ghana 2008 Birth Recode Weights [message #2919 is a reply to message #2838] Wed, 10 September 2014 09:34 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3016
Registered: February 2013
Senior Member
Another response from Noureddine Abderrahim:

Try to use the distributed dataset rather than the one you constructed. There might be a problem when you constructed the reduced dataset.

DHS weights are created for the whole Enumeration Area not for the individuals within the EA.
Previous Topic: Confusion re: "Numbers in parentheses are based on 25-49 unweighted cases"
Next Topic: Weighting data
Goto Forum:
  


Current Time: Thu Mar 28 09:16:09 Coordinated Universal Time 2024