The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Nutrition and Anthropometry » Latest birth identification in KR file
Latest birth identification in KR file [message #20415] Fri, 30 October 2020 09:16 Go to next message
id709nvz is currently offline  id709nvz
Messages: 43
Registered: March 2019
Member
Dear DHS experts,
I am using Ethiopian KR file for the whole round to examine the effect of mother's characteristics on child outcome under five years old. Since I am new to DHS data I don't really understand whether KR file has one child per mother or may contain more than 1 child for a mother. What I need is one child for each mother, how can I keep only 1 child with his/her mother in KR file in Stata?
Specifically, I am looking at HAZ, WHZ, and WAZ scores as a child outcome. is it the case that only the latest birth was measured for anthropometry i.e. just one child for each mother?

I like to use anthropometry of the latest birth instead of all births for a mother.

Kind regards
Re: Latest birth identification in KR file [message #20416 is a reply to message #20415] Fri, 30 October 2020 10:11 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3171
Registered: February 2013
Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

All the children under five who were living in the household at the time of the survey are eligible for height and weight measurement. This includes nearly all of the children in the KR file, plus a few others whose mother was not in the household. The KR file includes all children born in the past five years whose mother was in the household survey and eligible for the individual interview. The most recent birth has bidx=1, the next most recent birth has bidx=2, etc. The KR file includes children who died. If the child is alive, b5=1. If died, b5=0. If the child is also in the household file (the PR file) the line number is given by b16.
Re: Latest birth identification in KR file [message #20417 is a reply to message #20416] Fri, 30 October 2020 10:22 Go to previous messageGo to next message
id709nvz is currently offline  id709nvz
Messages: 43
Registered: March 2019
Member
Dear DHS expert,
Thanks for your information.
If I keep bidx==1 & b5==1, will this give me only the latest children and their mothers only?
to put in another way, will this give me one alive child per mother?

Kind regards
Re: Latest birth identification in KR file [message #20420 is a reply to message #20417] Fri, 30 October 2020 15:47 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3171
Registered: February 2013
Senior Member


Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

If you use the KR file, it can happen that the most recent child is not living. If you only want the youngest child who was actually living and measured, you can do the following:

Open the KR file
keep if hw70<.
sort v001 v002 v003 bidx
egen sequence_living_children=seq(), by(v001 v002 v003)
keep if sequence_living_children==1

The trick here is "egen sequence". If you enter "help egen" you will find a lot of helpful functions.

Of course, there may be some women who had a child in the past five years but they do not have any children who are still alive and were measured.

Re: Latest birth identification in KR file [message #20434 is a reply to message #20420] Sun, 01 November 2020 16:16 Go to previous messageGo to next message
id709nvz is currently offline  id709nvz
Messages: 43
Registered: March 2019
Member
Dear DHS expert,
Thank you. It was helpful.

If I may ask one more question. In the Ethiopian KR file, if I consider all children under five, there are cases where weight/height was not measured for some children of a given mother. For instance, the youngest might have been measured while the older isn't. or the reverse or not at all. Was this random or weight/height measurement was selective? I am analyzing a reform for the mother that I believe might affect all the children of a mother. Under such circumstance will it be a problem if I consider all the children record?
This really important in my analysis.

kind regards
Re: Latest birth identification in KR file [message #20439 is a reply to message #20434] Mon, 02 November 2020 08:32 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3171
Registered: February 2013
Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

Height and weight measurements are taken for all the children in the household who have not yet had their 5th birthday. This is done as part of the household survey.

The KR file includes all children born in the past 5 years, whether or not they are alive. If a woman had more than one child in the past 5 years, and one of them was not measured, then it could be because the child has died (b5=0) or because the child was not in the household. Very rarely, the mother will refuse to allow the child to be measured, but when that happens is almost always applies to all children under 5, not just some of them.
Previous Topic: Definition of prelacteal feed in Indonesia DHS
Next Topic: BMI India 2015
Goto Forum:
  


Current Time: Tue Oct 8 14:11:08 Coordinated Universal Time 2024