The DHS Program User Forum: General » Children's nutritional outcomes and father's characteristics

Home » Topics » General » Children's nutritional outcomes and father's characteristics (Rwanda DHS VII)

Show: Today's Messages :: Show Polls :: Message Navigator

Children's nutritional outcomes and father's characteristics [message #27970]

Fri, 27 October 2023 12:25

fraeus
Messages: 14
Registered: February 2023

Member

Hi All. I am wowrking on the attached code. I am analsying nutritional outcomes for children, so stunting, wasting, and underweight. However, when I restrict the sample to only those children eligible for these measurement and I try to check some characteristics of the father (for example land ownership), I only see missing values. The same does not happen for other rounds of the DHS. Do you know why? Is there any non-random choice behind this?

Any help would be much appreciated.
Thank you

Attachment: merge_children_mothers_fathers_do_14Feb2023+(2).txt
(Size: 8.56KB, Downloaded 108 times)

Report message to a moderator

Re: Children's nutritional outcomes and father's characteristics [message #28005 is a reply to message #27970]

Wed, 01 November 2023 12:05

Janet-DHS
Messages: 888
Registered: April 2022

Senior Member

Following is a response from DHS staff member, Tom Pullum:

Linking children with their biological fathers is complicated. I see that you are using a program I posted earlier. I also see that you are using the Rwanda 2010-11 survey. If, in RWPR61, you enter "tab1 hv117 hv118" you will see that there was subsampling of men. You can check the report, but there are about half as many eligible men as eligible women. You will lose about half of the fathers that way. However, you should definitely not lose ALL the fathers.

I will try to check what has happened and will add another post on this.

Report message to a moderator

Re: Children's nutritional outcomes and father's characteristics [message #28007 is a reply to message #28005]

Wed, 01 November 2023 12:14

fraeus
Messages: 14
Registered: February 2023

Member

Hi Janet, thank you for your reply. Actually I was using 2010, 2014, and 2019. I lose the fathers on the 2014 and not on the other years with the same code, but reading from the report I noticed that differently from 2010 and 2019, in 2014 they check those children's nutritional indicators in the half sample with women only. Can you confirm that? So, by default, I cannot have information on fathers land ownership, because their information is in the other half of the sample, where children have not been measured in terms of Height-for-weight etc.

Report message to a moderator

Re: Children's nutritional outcomes and father's characteristics [message #28010 is a reply to message #28005]

Wed, 01 November 2023 14:19

Janet-DHS
Messages: 888
Registered: April 2022

Senior Member

Following is a response from DHS staff member, Tom Pullum:

I just ran the program you are using and it works fine. (However, I don't know if your interest is in the Rwanda 2010 survey. I used that survey as a test.) In the program you will find the following lines:

* Copy of the PR file for children
use PR.dta, clear
keep if hv105<=17
keep hv*

If your interest is in wasting, etc., then you need to include hc70, hc71, and hc72 in the "keep hv*" line. You would construct the anthropometry indicators from hc70-hc72. The code for those indicators is elsewhere on the forum. They apply only to children age 0-4, but you don't need to change the selection on hv105.

Please actually try to run the program, on whatever survey you are working with. You will have to change the path to the data (see "scalar spath= ..."). Please let us know if you have other questions.

Report message to a moderator

Re: Children's nutritional outcomes and father's characteristics [message #28028 is a reply to message #28007]

Fri, 03 November 2023 09:39

Janet-DHS
Messages: 888
Registered: April 2022

Senior Member

Following is a response from DHS staff member, Tom Pullum:

What you are describing with the 2014 survey has happened several times with various surveys. The sample may be split in 2 or more ways, and it is then impossible to relate the variables in one part with the variables in another part. Subsampling reduces costs and reduces the burden on the respondents by shortening the length of the interview. Sorry that it has affected your analysis.

Report message to a moderator

Re: Children's nutritional outcomes and father's characteristics [message #28036 is a reply to message #28028]

Fri, 03 November 2023 12:10

fraeus
Messages: 14
Registered: February 2023

Member

I understand, thank you so much. Let's consider 2010. My merged file has as unit of analysis children. If I want to run some descriptive statistics for their mothers and fathers, without counting twice a mother or father of two children, how can I do?

For example, I want to do tab v745b and tab mv745b, so I want to check the percentage of parents owing land alone, jointly, or both. But, if a mother has two children, I will have two lines in the dataset for the same mother (child 1 and child2), that will be counted as twice in my tab v745b. Any suggestions on that?

thank you so much

Best,
Francesca

Report message to a moderator

Re: Children's nutritional outcomes and father's characteristics [message #28078 is a reply to message #27970]

Wed, 08 November 2023 10:13

fraeus
Messages: 14
Registered: February 2023

Member

I also have another question on this code. How do I now if the two parents of the child are a couple? If they are married or just living together ?

Thank you

Report message to a moderator

Previous Topic:	Household Health Expenditure
Next Topic:	writePff_ExternalFile

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sun Nov 24 09:49:33 Coordinated Universal Time 2024