| Breastfeeding status [message #30626] |
Fri, 10 January 2025 17:52  |
Varsha
Messages: 51
Registered: November 2020
|
Senior Member |
|
|
Hello,
I have three small doubts.
(1) The questions on breastfeeding in the NFHS questionnaires pertain to the last birth. This last birth doesn't correspond to the one born in the 2 years before the survey, right? I'm confused because the calculations related to breastfeeding in the reports restrict the age of the last birth to the 2 years before the survey?
(2) I want to create a variable on the duration (months) of breastfeeding. For the 'never breastfed' category, how should I assign the value? Can I club it with the '0' category?
(3) Here: https://www.dhsprogram.com/pubs/pdf/DHSG1/Guide_to_DHS_Stati stics_DHS-8.pdf on page 11.28, why don't we exclude 0 as a response from ever breastfed (it excludes 94 and 99 (is it don't know?))?
Thanks
[Updated on: Fri, 10 January 2025 17:52] Report message to a moderator |
|
|
|
| Re: Breastfeeding status [message #30630 is a reply to message #30626] |
Sat, 11 January 2025 16:31  |
Varsha
Messages: 51
Registered: November 2020
|
Senior Member |
|
|
Hello,
I'm trying to create a variable indicating whether a child was ever breastfed as follows:
gen bf_ever= (m4!=94 & m4!=99)
replace bf_ever = . if m4==.
replace bf_ever = . if midx>1
As the information on the duration of breastfeeding in the NFHS was collected for the last birth, why do we have some responses for midx>1 categorized as never breastfed in the KR file? Shouldn't it be missing as the question does not pertain to them (that is why I'm replacing the above created variable with a missing when midx>1)?
Please help me with this and my previously posted doubts.
[Updated on: Sat, 11 January 2025 16:32] Report message to a moderator |
|
|
|
|
|
| Re: Breastfeeding status [message #30651 is a reply to message #30630] |
Tue, 14 January 2025 14:22 |
Bridgette-DHS
Messages: 3229
Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff member, Tom Pullum:
In the final report on the NFHS-5, the first column of table 10.5 gives the percent ever breastfed, among children born in the two years before the survey. All DHS final reports include a table like this. The general approach is as follows, illustrated with the Nigeria 2013 DHS:
use "...NGKR6AFL.DTA" , clear
gen months_ago=v008-b3
gen everbf=1 if m4<.
replace everbf=0 if m4==94 | m4==99
tab v190 everbf if months_ago<24 & m4<. & bidx==1 [iweight=v005/1000000], row
These results can be compared with table 11.2 of the final report on the Nigeria 2013 survey. There is an exact match. In newer surveys that include b19, b19 would be used in place of "months_ago". Not all surveys include a code m4=99, but for those that do, it is grouped with m4=94.
This will match table 10.5 exactly, because this is one of the many tables in the NFHS-5 report that we are unable to replicate. However, these lines describe how the indicator is constructed.
|
|
|
|
|
|
|
|
|
|
|
|
| Re: Breastfeeding status [message #30699 is a reply to message #30686] |
Tue, 21 January 2025 14:15 |
Bridgette-DHS
Messages: 3229
Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff member, Tom Pullum:
Yes, 98 and 99 would be treated the same. Sometimes these codes appear (in many surveys) even though they are not in the label for the codes.
Yes, bidx is 1 for the most recent birth. If it was a twin birth, the second twin has bidx=1 and the first twin has bidx=2. There is also a variable b0 for each birth which tells whether the child was part of a multiple birth, and if so, which number within the multiple birth.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
