The DHS Program User Forum: Merging data files » Child nutrition and spousal violence

Home » Data » Merging data files » Child nutrition and spousal violence (Merging files)

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

Child nutrition and spousal violence [message #28214]

Thu, 30 November 2023 07:26

darji.shah_ug24@ashoka.ed is currently offline

darji.shah_ug24@ashoka.ed
Messages: 10
Registered: November 2023

Member

Hi I'm working on a paper that wishes to assess the impact of spousal violence on children's nutrition. I have compiled the list of variables relevant to my study however am unsure of how to merge household and IR files. What unique identifiers do i need to incorporate to facilitate this?

Report message to a moderator

Re: Child nutrition and spousal violence [message #28216 is a reply to message #28214]

Thu, 30 November 2023 11:34

Bridgette-DHS
Messages: 3230
Registered: February 2013

Senior Member

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

The child nutrition variables are in the KR file. That file includes most of the variables for the mother that are in the IR file, but it does not include the DV variables, so you need to merge the IR and KR files, using v001 v002 v003 in both files. For this merge, note that the mother appears only once in the IR file but can appear more than once in the KR file (if she had more than one child in the past 5 years). If you have more questions about the merge, please say what survey you are using. Please also look at the Guide to DHS Statistics ( https://www.dhsprogram.com/Data/Guide-to-DHS-Statistics/inde x.cfm) for more information about the variables in the standard recode files.

Report message to a moderator

Re: Child nutrition and spousal violence [message #28218 is a reply to message #28216]

Sat, 02 December 2023 15:14

darji.shah_ug24@ashoka.ed
Messages: 10
Registered: November 2023

Member

I am using NHFS-5 data for India. I understand that i have to merge IR and KR. However, i still have a doubt. I was hoping to add controls of region (urban/rural), wealth, etc. these characteristics of households are unavailable in the KR and IR files, right? Moreover, if these variables are only there in the Household file will it be possible for me to merge all 3?

Report message to a moderator

Re: Child nutrition and spousal violence [message #28233 is a reply to message #28218]

Mon, 04 December 2023 12:05

Bridgette-DHS
Messages: 3230
Registered: February 2013

Senior Member

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

It is possible to do a merge of the PR, IR, and KR files, but it may not be necessary. It depends on what you mean by "etc.". The KR file includes region (v024), place of residence (v025), and wealth quintile (v190). What other variables do you need? You can check whether they are in the IR file, for example with Stata commands such as "lookfor residence" or "lookfor wealth".

Report message to a moderator

Re: Child nutrition and spousal violence [message #28234 is a reply to message #28233]

Mon, 04 December 2023 12:21

darji.shah_ug24@ashoka.ed
Messages: 10
Registered: November 2023

Member

Yes I have located those variables it makes sense now. Yet another follow up is that: I am attempting make the standard deviations of weight-for-height z-scores, weight-for-age z-scores and weight-for-age z scores my explanatory variable, however, I am unsure of whether the HW-5, HW-8 and HW-11 are the relevant variables here or HW70, HW71 and HW72? Can DHS guide me through that? How do I obtain the z-scores

Report message to a moderator

Re: Child nutrition and spousal violence [message #28241 is a reply to message #28234]

Tue, 05 December 2023 07:01

Bridgette-DHS
Messages: 3230
Registered: February 2013

Senior Member

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

There have been other postings on this. The relevant variables are hw70-hw72 in the PR file (the same as hc70-hc72 in the KR file). These are based on the WHO Child Growth Standards. hw5, hw8, and hw11 use an earlier CDC standard and in my opinion should no longer be included in the standard recode file. They are only useful if you want to construct continuous longterm trajectories that go back before the WHO standards were developed, almost 20 years ago ( https://www.who.int/tools/child-growth-standards/who-multice ntre-growth-reference-study).

Report message to a moderator

Re: Child nutrition and spousal violence [message #28253 is a reply to message #28241]

Wed, 06 December 2023 04:49

darji.shah_ug24@ashoka.ed
Messages: 10
Registered: November 2023

Member

Is the hw70 standard deviations from z-score? if yes, when we are creating stunting as a variable of hw70<-2 (to capture stunting) it says that there no observations when we try and regress it the new variable with our independent variables. I'm not sure why that is happening, is there a misunderstanding in our interpretation of the variable hw70?

Report message to a moderator

Re: Child nutrition and spousal violence [message #28256 is a reply to message #28241]

Wed, 06 December 2023 05:40

darji.shah_ug24@ashoka.ed
Messages: 10
Registered: November 2023

Member

this is the code we have used, for context:
log using "c:\users\dell\documents\project\log_file.log", replace
clear all
set maxvar 7000
use "c:\users\dell\documents\project\working_dataset.dta"
keep v012 v013 d101* d102 d103* d104 d105* d110* d118a d121 hw1 hw2 hw3 hw51 hw70 hw71 hw72 hw73 m18 v025 m10 v130 v131 v190 v136 v137 v044 v106 v191 hw13
keep if v012 >= 18 & v012 <= 49 | hw1 < 60
keep if hw71 >= -600 & hw71 <= -201
tab hw71 reg hw71 d105*
regress hw71 d105*
log close

Report message to a moderator

Re: Child nutrition and spousal violence [message #28257 is a reply to message #28256]

Wed, 06 December 2023 06:14

darji.shah_ug24@ashoka.ed
Messages: 10
Registered: November 2023

Member

actually i have identified that the regression works on most independent variables except the ones that have V502 = 1 as their base, how do i change my code to accomodate for that?

Report message to a moderator

Re: Child nutrition and spousal violence [message #28261 is a reply to message #28257]

Wed, 06 December 2023 10:27

Bridgette-DHS
Messages: 3230
Registered: February 2013

Senior Member

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

Yes, hw70 is the number of standard deviations from the WHO standard, but it is multiplied by 100. Below I will paste the Stata code for the basic anthropometric indicators.

gen HAZ=hw70/100 if hw70>-600 & hw70<600
gen WAZ=hw71/100 if hw71>-600 & hw71<500
gen WHZ=hw72/100 if hw72>-500 & hw72<500
gen stunted=0
gen underweight=0
gen overweight=0
gen wasted=0
replace stunted=1 if HAZ<-2
replace underweight=1 if WAZ<-2
replace overweight=1 if WHZ>2
replace wasted=1 if WHZ<-2

Report message to a moderator

Re: Child nutrition and spousal violence [message #28262 is a reply to message #28261]

Wed, 06 December 2023 10:35

Bridgette-DHS
Messages: 3230
Registered: February 2013

Senior Member

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

From your very brief statement, I cannot tell what the problem is. v502=1 identifies women who are currently in a union. Most (not all) of the DV questions relate to the current partner. If you want to include more women, just drop the condition. You will then add in women who were previously in a union (v502=2). Women who were never in a union (v502=0) are NA (or mostly NA) for the module.

Report message to a moderator

Re: Child nutrition and spousal violence [message #28275 is a reply to message #28262]

Thu, 07 December 2023 02:02

darji.shah_ug24@ashoka.ed
Messages: 10
Registered: November 2023

Member

Hi thank you for all your help. I had another query about how i can add state level fixed effects or district level fixed effects in STATA using nhfs. is there a variable for it

Report message to a moderator

Re: Child nutrition and spousal violence [message #28276 is a reply to message #28275]

Thu, 07 December 2023 11:14

Bridgette-DHS
Messages: 3230
Registered: February 2013

Senior Member

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

In the list of covariates for the model, you would include categorical forms of v024 or sdist, that is, "i.v024" or "i.sdist" .

Report message to a moderator

Re: Child nutrition and spousal violence [message #28277 is a reply to message #28276]

Thu, 07 December 2023 11:21

Bridgette-DHS
Messages: 3230
Registered: February 2013

Senior Member

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

In the list of covariates for the model, you would include categorical forms of v024 or sdist, that is, "i.v024" or "i.sdist" .

Report message to a moderator

Previous Topic:	Merging two rounds
Next Topic:	Merging and appending data files

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sat Nov 8 01:08:34 Coordinated Universal Time 2025