The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Uganda » URGENT QUESTION: Missing child nutrition values in Uganda DHS2016 (Missing child nutrition values for Domestic Violence respondents in Uganda DHS 2016)
URGENT QUESTION: Missing child nutrition values in Uganda DHS2016 [message #19429] Tue, 16 June 2020 18:16 Go to next message
smd80 is currently offline  smd80
Messages: 6
Registered: June 2020
Member
Dear DHS Program

I would like to examine the intimate partner violence variables (RECDV WD1) and child height and weight under age 5 (REC44 W44) contained in the Individual Questionnaire (IR File UGIR7BSV). Unfortunately, there seems to be an issue in the dataset: there are no child nutrition values for any of the women who responded to the domestic violence module, even though the dataset indicates that there are 5615 respondents to the domestic violence module with at least one child. Do you know why/how the nutrition-related variables related to all of these children can be blank?

I wondered whether mothers needed to be matched to children via the household number or birth index, but this is not the case in other DHS datasets. Moreover, there are child nutrition values for other respondents, just not the respondents to the domestic violence module.

I would very grateful for a response, as this seems to be a serious issue and may affect other important ongoing Uganda DHS analyses.

Thank you very much in advance and kind regards
JSD
Re: URGENT QUESTION: Missing child nutrition values in Uganda DHS2016 [message #19482 is a reply to message #19429] Tue, 30 June 2020 10:19 Go to previous messageGo to next message
jfleuret is currently offline  jfleuret
Messages: 3
Registered: May 2019
Member
Hi, I hope this finds you well. I was the ICF Survey Manager for the 2016 UDHS. In that survey, we did not conduct biomarker collection in the households where we had the domestic violence module for women, so there will not be anthropometry data for the children of those women. There is an explanation of the sample on page 3 of the Final Report (https://dhsprogram.com/pubs/pdf/FR333/FR333.pdf), please don't hesitate to let us know of other questions.

Best,
Julia
Re: URGENT QUESTION: Missing child nutrition values in Uganda DHS2016 [message #22417 is a reply to message #19482] Wed, 10 March 2021 11:00 Go to previous messageGo to next message
smd80 is currently offline  smd80
Messages: 6
Registered: June 2020
Member
Dear Julia,

Thank you so much for your timely response and apologies for only responding now. Due to COVID-19 I took a break of my PhD and am only resuming now.

Based on your response, does that mean that it is not possible to link any of the domestic violence variables to child nutrition outcomes in any of the Uganda DHS datasets? Was there a particular reason why under 5 anthropometric measures (weight, height and age) were not collected for mothers responding to the domestic violence module? Will this change in future DHS? Do you think there is any way to handle this missing data by linking other datasets?

Thank you very much for your help in advance!

Best wishes,
Sarika
Re: URGENT QUESTION: Missing child nutrition values in Uganda DHS2016 [message #22427 is a reply to message #22417] Thu, 11 March 2021 12:05 Go to previous messageGo to next message
jfleuret is currently offline  jfleuret
Messages: 3
Registered: May 2019
Member
Dear Sarika,

Yes, because in the 2016 UDHS anthropometry data were not collected in households where the DV module was implemented, there is no link between those variables.

Surveys often have subsamples (quite often for biomarkers and, to the best of my knowledge, always for DV). While the DV subsampling also has to do with wanting to increase confidentiality, subsampling is generally recognized as a tool to reduce survey burden and increase the feasibility of survey implementation - it's not a specific decision to avoid collecting anthro in DV households.

I encourage you to check Chapter 1 of the Final Report for any survey in which you are interested to understand whether that survey collected all data in all households or had any subsampling (different biomarkers in different households, different questionnaires in different households).

Subsampling decisions are survey-specific and are taken during survey design by the survey implementing organization & steering committee. Precedents are not set in stone.

I can't think of any way to extrapolate from the children with anthro data (but no mother DV data) to children without anthro data (but yes mother DV data), although I am not a statistician/expert in this domain.

Please don't hesitate to let us know of any other questions & good luck with your PhD.

Best,
Julia
Re: URGENT QUESTION: Missing child nutrition values in Uganda DHS2016 [message #22429 is a reply to message #22427] Thu, 11 March 2021 12:25 Go to previous messageGo to next message
smd80 is currently offline  smd80
Messages: 6
Registered: June 2020
Member
Dear Julia,

Thank you so much for your response! That makes a lot of sense although this news is very disappointing for my research study (focusing on the association of IPV and child malnutrition in Uganda). I suppose this is the same for the 2006 and 2011 datasets? I wonder if you might know of any other sources I could examine to look at this association? I will also try and ask a statistician if that extrapolation is in any way possible, although I believe it would undermine establishing a link between maternal care-giving behaviour to her child.

Thank you again - your responses have been extremely helpful.

With best wishes from Kampala,
Sarika
Re: URGENT QUESTION: Missing child nutrition values in Uganda DHS2016 [message #22430 is a reply to message #22429] Thu, 11 March 2021 14:18 Go to previous message
jfleuret is currently offline  jfleuret
Messages: 3
Registered: May 2019
Member
Dear Sarika,

For the 2006 and 2011 UDHSes, I was not involved with them and don't know the answer offhand - as I mentioned, checking the first chapter of the final report should be enlightening.

For other sources - I don't think I know of anyplace you wouldn't have already thought of (MOH, Makerere, USAID, CDC, UNICEF, etc.), although of course not all research/datasets from those sources would be publicly available.

In the 2016 UDHS, you do of course have IYCF, supplementation, birth weight, etc. info for children (in the right age group) of women who responded to DV questions; perhaps those may be fruitful variables to explore.

Best,
Julia
Previous Topic: Causes of maternal mortality
Goto Forum:
  


Current Time: Thu Jun 17 21:45:25 Coordinated Universal Time 2021