The DHS Program User Forum - RDF feed
https://userforum.dhsprogram.com/index.php
Re: Maternal Education and Stunting
https://userforum.dhsprogram.com/index.phpindex.php?t=rview&goto=28928&th=13564#msg_28928
To proceed with your analysis in Stata, you will need to use regression commands. For your first question, you can use a logistic regression model where the dependent variable is stunting (coded as 1 for stunted, 0 for not stunted) and the independent variable is mother's education (measured in years). You can also include an interaction term between mother's education and a binary variable for urban/rural location to see how the impact of mother's education on stunting varies across urban and rural areas. The Stata command for this would be:
logit stunting c.mother_education##c.urban_rural
For your second question, you will need to recode mother's education into categories (e.g., no education, incomplete primary, complete primary, etc.) and then run a logistic regression model with these categories as independent variables. You can then compare the coefficients to determine which category yields the greatest relative return. The Stata command for this would be:
xi: logit stunting i.education_category
Please note that these are basic commands and you may need to adjust them based on the specifics of your dataset and research question.]]>Janet-DHS2024-03-28T19:25:17-00:00Maternal Education and Stunting
https://userforum.dhsprogram.com/index.phpindex.php?t=rview&goto=28900&th=13564#msg_28900
I am very unsure about how to progress in my analysis. I have gotten all of the classifications of stunting, underweight and wasting done (eg, -2SD), but now Im unsure about how to progress.
I want to look at the impact that each additional year of education a mother has on stunting and then look at how this varies across urban and rural areas. I then want to look at which classification of education (eg. no education, incomplete primary, complete primary etc) has the greatest relative return.
Any ideas on the commands I would need to put into stata?
Thanks]]>Morrilu2024-03-26T14:03:39-00:00Re: Calculating Undernutrition Indicators for Under-5 Children in Older surveys
https://userforum.dhsprogram.com/index.phpindex.php?t=rview&goto=28659&th=13470#msg_28659
The formulas for the WHO Child Growth Standards (https://www.who.int/tools/child-growth-standards) have not changed since 2006. The only part of the standards that DHS uses to construct hw70-hw72 are weight, height/length, and whether the child was standing (height) or lying (length) at the time of measurement. The assumptions, cutoffs, etc., are all consistent with WHO standards. We have had several reports related to the quality of the anthropometric data (see our publications page) and there is good evidence of improved measurement, especially of height/length, in the past 6 years or so. With the HW files you can get the values that would have been calculated in earlier years if the formulas had been available in earlier years. If you go back to the earliest years of DHS, however, there was no anthropometry at all.]]>Janet-DHS2024-02-15T21:31:42-00:00Re: Calculating Undernutrition Indicators for Under-5 Children in Older surveys
https://userforum.dhsprogram.com/index.phpindex.php?t=rview&goto=28652&th=13470#msg_28652
I got the HW file and I should be able to merge with the KR file accordingly. My concern is that is there any adjustment to be made to this data to make them comparable to the current WHO standard in the most recent surveys?
I am interested in constructing the indices of malnutrition over the range of the cross-sections. I have successfully done for the recent surveys but I am stuck with the older surveys and I want to get it right at this stage.]]>Adetunji2024-02-14T10:42:30-00:00Re: Calculating Undernutrition Indicators for Under-5 Children in Older surveys
https://userforum.dhsprogram.com/index.phpindex.php?t=rview&goto=28647&th=13470#msg_28647
The older surveys that had anthropometry have hw5, hw8, and hw11 in the KR file instead of hw70, hw71, hw72. Those variables were based on an earlier CDC standard rather than the current WHO standard. (Those variables are still included in the data files but are hardly ever used.)
You should be able to find an HW file (the file name has "HW" in place of "KR") for such surveys. You just have to merge it with the KR file from the same survey. Let us know if you have difficulty doing this.]]>Janet-DHS2024-02-13T21:30:56-00:00Calculating Undernutrition Indicators for Under-5 Children in Older surveys
https://userforum.dhsprogram.com/index.phpindex.php?t=rview&goto=28630&th=13470#msg_28630
I am working on the coexisting forms of malnutrition in under-5 children in SSA. Using the KR file in the newer surveys DHS-V upward, I have successfully computed the indicators of undernutrition and their coexistence respectively. I used the variables hw71, hw71 and hw72 for my computation. However, these variables are not found in the same KR file in DHS-IV downward.
Kindly advise on what I should do. My unit of analysis remains the under-5 children.
Thanks in anticipation of your help.]]>Adetunji2024-02-09T13:40:54-00:00Re: Which variable corresponds to residence, mothers education
https://userforum.dhsprogram.com/index.phpindex.php?t=rview&goto=28615&th=13456#msg_28615
This helps me a lot.]]>Mahir2024-02-06T15:23:52-00:00Re: Which variable corresponds to residence, mothers education
https://userforum.dhsprogram.com/index.phpindex.php?t=rview&goto=28609&th=13456#msg_28609
The variables you listed for urban/rural are all the same. Type of place of residence is written (excuse the obsolete terminology!) on the cover page of the household questionnaire and is the same for the household and for everyone in it. It is also the same for all households in the same cluster and is a characteristic of the cluster or enumeration district in the sampling frame, using the definitions of the country's statistical office. It is hv025 in the household files; v025 or v102 in the women's file; and mv025 or mv102 in the men's file. 1 is urban and 2 is rural.
Hv106 is education and it also appears with hv in the household files, v in the women's file, and mv in the men's file, but it varies from one person in the household to another. In the HR file, with one record per household, the members of the household are assigned subscripts _01, _02, etc. which are the line number in the household listing. The PR file has household members as units and the subscripts are not needed (hvidx is the line number). For children under 5, in the PR file the mother's education is put on the child's record as hc61, just for convenience. (The mother's line number is hv112, and her value of hv106 is copied onto the child's record.) Hope this helps!]]>Janet-DHS2024-02-05T20:40:56-00:00Which variable corresponds to residence, mothers education
https://userforum.dhsprogram.com/index.phpindex.php?t=rview&goto=28593&th=13456#msg_28593
I hope you are doing well.
I have 2 questions which I wanted to clarify
1) Which variable exactly defines residence of the respondent (urban/rural)? I see that there are multiple variables that collect this information(hv025, v025, mv025, v102, mv102), which of these should I use to find proportion of respondents in urban vs rural? Additionally, how is this variable coded? responses are (1/2) what is rural and what is urban? I could not find this information in Guide to DHS statistics.
2) Which variable should I use for mothers education? I can see that hv106, hc62 and hc61 all record mothers education. Which of these should I use? And even within these variables there are multiple variables hv106_01, hv106_02 and so on. Which One should I use?
It would be great if you can help me with these queries.
Warm wishes
Mahir ]]>Mahir2024-02-02T11:54:18-00:00Re: child health relation with his mother's health in India
https://userforum.dhsprogram.com/index.phpindex.php?t=rview&goto=28552&th=13436#msg_28552
Mukesh2024-01-29T20:02:42-00:00