The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Bangladesh » Estimates on birth registration do not match for wealth index
Estimates on birth registration do not match for wealth index [message #12306] Fri, 28 April 2017 06:25 Go to next message
dgodha
Messages: 44
Registered: November 2016
Location: India
Member
Hello,

I have been trying to replicate the birth registration estimates dis-aggregated by area, sex, division/region, and wealth index. My commands are pasted below:

**********************************************************
/*BDPR70FL.DTA used*/
gen bc = .
replace bc = 0 if inrange(hv105,0,4)& hv102==1
recode bc (0=1) if inrange(hv140,1,2)

/*Set data for survey weighting*/
	gen wt=hv005/1000000

/*Generating weights*/
	svyset [pw=wt], psu(hv021) strata(hv025)

tabout hv025 hv104 hv270 hv024  bc using t4.txt if hv105<5,cells (row ) ci2col format(2) clab (Row_% ) percent npos(row) lay(col) svy ptotal(single) replace

**********************************************************

All my estimates match those in Table 2.12 on Page 20 of the DHS 2014 Report except for those for wealth index. The percents as well as the frequency match exactly for all other dis-aggregations. My output for wealth index is pasted below along with estimates from the DHS report.

My calculations DHS 2014 Report
WI % Count % Count
Poorest 15.26 1737 15.4 1652
Poorer 18.88 1493 18.7 1442
Middle 17.53 1490 17.5 1404
Richer 21.15 1568 20.4 1583
Richest 28.61 1511 28.0 1717
Total 20.16 7798 20.2 7798

I will appreciate any insight into my error and how to fix it.

Thanks
D Godha


Deepali

[Updated on: Fri, 28 April 2017 06:31]

Report message to a moderator

Re: Estimates on birth registration do not match for wealth index [message #12323 is a reply to message #12306] Mon, 01 May 2017 10:56 Go to previous messageGo to next message
Trevor-DHS is currently offline  Trevor-DHS
Messages: 787
Registered: January 2013
Senior Member
Thanks for pointing this out. Due to late changes related to the wealth index, there are some tables in this report that used a prior version of the wealth index, and not the final one. We will be correcting these estimates for the wealth index. Your calculations are correct.
Re: Estimates on birth registration do not match for wealth index [message #12328 is a reply to message #12323] Mon, 01 May 2017 15:10 Go to previous messageGo to next message
dgodha
Messages: 44
Registered: November 2016
Location: India
Member
Thanks Trevor.

Deepali
Re: Estimates on birth registration do not match for wealth index [message #19366 is a reply to message #12328] Fri, 05 June 2020 11:17 Go to previous messageGo to next message
ronaldmusi is currently offline  ronaldmusi
Messages: 3
Registered: June 2020
Member
Im also trying to reproduce Table 2.1 0 Birth registration of children under age 5 for Zimbabwe DHS 2015. My figures are not the same with those in the report. Below is my code


use "ZWPR72FL.DTA"

/*Set data for survey weighting*/
gen wt=hv005/1000000

/*Generating weights*/
svyset [pw=wt], psu(hv021) strata(hv025)


recode hv140 (0=0) (1/2=1) (else = .) if hv102==1, gen(birthreg)
label var birthreg " birth registered "
ta birthreg
Re: Estimates on birth registration do not match for wealth index [message #19367 is a reply to message #19366] Fri, 05 June 2020 11:48 Go to previous messageGo to next message
Trevor-DHS is currently offline  Trevor-DHS
Messages: 787
Registered: January 2013
Senior Member
You have two problems with your code:
1) In your recoding of hv140 you are dropping a number of cases. To match DHS results use the following. This recodes "don't know" and missing responses to 0, but restricts the recoding just to children age 0-4
recode hv140 (0=0) (1/2=1) (else = 0) if hv102==1 & hv105<5, gen(birthreg)
2) You are tabulating the data unweighted. To match the results you must tabulate the data using the weights.
svy: tab birthreg
Re: Estimates on birth registration do not match for wealth index [message #19374 is a reply to message #19367] Fri, 05 June 2020 17:09 Go to previous message
ronaldmusi is currently offline  ronaldmusi
Messages: 3
Registered: June 2020
Member
Thank you very much. I got it right now.
Previous Topic: Mother's BMI
Next Topic: How to calculate mother's age at most recent birth (last birth) in BDHS 2011
Goto Forum:
  


Current Time: Fri Mar 29 08:11:20 Coordinated Universal Time 2024