The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Child Health » Confine my Dataset
Confine my Dataset [message #11454] Thu, 22 December 2016 11:19 Go to next message
Messages: 10
Registered: November 2016
Location: UK

How can I handled cases of children who have died? ,say in the first year of their life!

So I am using STATA and trying to merge the files, I just wanted to confine my data set to the following cases:

1- Household with children under five age.
2- Ever married women with children under five age.
3- Child is a live
Meaning to omit the records for died children, unmarried women, household does not have children under five.

Not having them included in the analysis Does it mean that I obtain results may be subject to "sample selection biases" ?!!

Can anyone guide me Please!

Re: Confine my Dataset [message #11960 is a reply to message #11454] Mon, 13 March 2017 12:42 Go to previous messageGo to next message
Messages: 1516
Registered: February 2013
Senior Member
Dear User,
We now have a resource of YouTube videos to assist our users. Please look these short videos over. Once you have viewed, if you still have questions, please feel free to post again. The videos can be found here: https:// y4CT
Thank you!
Re: Confine my Dataset [message #14559 is a reply to message #11454] Sat, 21 April 2018 18:42 Go to previous message
kingx025 is currently offline  kingx025
Messages: 95
Registered: August 2016
Location: Minneapolis. Minnesota
Senior Member
I am not sure of your motivation for limiting your dataset, but an easy solution is to create a customized dataset with just the samples and variables you need from the IPUMS-DHS database (at This will limit the number of variables and size of the file to what fits your research project. Downloading data from IPUMS-DHS also saves you the work of merging across file types, because the IPUMS-DHS staff have attached household characteristics to the files for women and attached the household and mother's characteristics to files for young children and births. Once you have made your datafile, check the universe (who is included with answers in variable) in the documentation for variables of interest. For example, you will find that for many variables relating to children, dead children are coded as "Not in universe" and don't have meaningful responses for issues like recent illness or nutrition. If you want to further restrict your population for analysis, I think it would be better to do it with programming statements after you have created a dataset for analysis.

IPUMS-DHS currently includes DHS data for 28 countries; you don't mention which countries interest you.

Miriam King

Dr. Miriam King
IPUMS-DHS Project Manager (
Previous Topic: Cooking place
Next Topic: Check ups after delivery - interviewed population
Goto Forum:

Current Time: Fri Jul 12 20:49:44 Coordinated Universal Time 2024