The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Nutrition and Anthropometry » EBF Jordan 1990 (Reproduce EBF estimate for Jordan 1990)
EBF Jordan 1990 [message #29959] Thu, 29 August 2024 11:54 Go to next message
staranalyst is currently offline  staranalyst
Messages: 4
Registered: August 2024
Member
Hello DHS colleagues,

I am trying to reproduce the EBF estimate for Jordan 1990. According to the Jordan 2023 DHS Summary Report (attached), Figure 8 Trends in exclusive breastfeeding, the official estimate for is 39%. However, using the KR file and the following STATA code, I got 23.95% with a weighted sample size of 724.7. Please advise.


*load in dataset
	use "$jordhsraw\1990\JOKR21DT\JOKR21FL", clear

*create an age var
	gen age_ch=v008-b3

*drop all vars with bf in it
	capture drop *bf*

*exclusive breastfeeding
	*generate an othfood var to indicate if consumed other foods.
	capture drop othfood
	egen othfood=rowtotal(v409 v409a v410 v410a v411 v411a v412 v413a v413 v414), missing

*create ebf based on age range, last-born, living with mother, current bf status, and other food consumed
	*age_ch<6 smaller than 6 month of age 
	*b9==0 lives with mother 
	*m4==95 currently breastfeeding for the specific child 
	*othfood==0 did not consume any other foods 
	gen ebf=.
	replace ebf=0 if age_ch<6 & b9==0 & midx==1
	replace ebf=1 if age_ch<6 & b9==0 & m4==95 & midx==1 & othfood==0
	bigtab ebf age_ch othfood 
	
	*validate the estimate against DHS report 
	*23.95% Sample size 724.7 Not validated. 
	tab ebf [iweight=v005/1000000]


Re: EBF Jordan 1990 [message #29983 is a reply to message #29959] Thu, 05 September 2024 08:33 Go to previous messageGo to next message
Janet-DHS is currently offline  Janet-DHS
Messages: 880
Registered: April 2022
Senior Member
Following is a response from DHS staff member, Tom Pullum:

I believe you are correct. The estimate of 39% appears to be drawn from STATcompiler. The estimate in the 1990 report is 39% for 0-1 months but an estimate for 0-5 months would be much lower.

I cannot take the time to check your calculation but have asked that the 39% in the KIR and on STATcompiler be re-calculated. The estimate of just 11% in the 1997 survey is also suspicious.

Thanks very much for identifying this likely error.
Re: EBF Jordan 1990 [message #29985 is a reply to message #29983] Thu, 05 September 2024 08:58 Go to previous messageGo to next message
staranalyst is currently offline  staranalyst
Messages: 4
Registered: August 2024
Member
Hi Tom/Janet, thanks for your response. When can I expect to hear back on the corrected EBF estimate for 1990 and 1997?

Re: EBF Jordan 1990 [message #30003 is a reply to message #29985] Mon, 09 September 2024 10:19 Go to previous messageGo to next message
Janet-DHS is currently offline  Janet-DHS
Messages: 880
Registered: April 2022
Senior Member
Following is a response from DHS staff member, Tom Pullum:

The staff who prepare the numbers on STATcompiler have been notified. I have no idea when the numbers posted there will be changed. You will just have to keep checking. We will not re-calculate independently, because of a risk that we will still not match. Thanks again for identifying the apparent error.
Re: EBF Jordan 1990 [message #30006 is a reply to message #30003] Mon, 09 September 2024 13:15 Go to previous message
staranalyst is currently offline  staranalyst
Messages: 4
Registered: August 2024
Member
Thanks Tom and Janet. I will keep checking the STATCompiler. Can you ask the analyst who will update STATCompiler to post their STATA code here when they update STATCompiler for Jordan 1990 EBF?
Previous Topic: Missing women's weight and height measurements in DHS-8
Next Topic: Fish consumption, DHS8
Goto Forum:
  


Current Time: Fri Nov 8 05:36:03 Coordinated Universal Time 2024