| Issue in child age variables in DHS08? [message #3381] |
Tue, 02 December 2014 11:23  |
Elizabeth
Messages: 9
Registered: October 2013
|
Member |
|
|
Dear DHS,
I'm analyzing child feces disposal by child age in years.
I get the correct percentages for stool disposal, and the total number of children, of all under fives.
But the number of children in each year bracket (0-11 months, 12-23 months, etc) differs from your report. Is there an issue with variable B3 or v008?
In the same processing for other DHS surveys, everything matches fine.
My syntax:
use " DHS.STATA.datasets/ChildRecodeKR/EgyptDHS08.EGKR5AFL/EGKR5AD T/EGKR5AFL.DTA "
keep v001 v002 v003 bidx b9
tab bidx b9
*keep if child lives with respondent
keep if b9==0
collapse (min) bidx, by(v001 v002 v003)
sort v001 v002 v003 bidx
save "/Desktop/temp2.dta", replace
use " DHS.STATA.datasets/ChildRecodeKR/EgyptDHS08.EGKR5AFL/EGKR5AD T/EGKR5AFL.DTA "
keep bidx b3 h11 hw1 hw13 hw70 v001 v002 v003 v005 v008 v021 v022 v024 v025 v116 v160 v190 v440 v465 v026
sort v001 v002 v003 bidx
merge v001 v002 v003 bidx using "/Desktop/temp2.dta"
keep if _merge==3
drop _merge
tab bidx
gen stratid = v022
gen psu = v021
gen weight = v005/1000000
svyset psu [pw=weight], strata(stratid)
gen age_interval=0
*modify age_interval to age in years.
replace age_interval=1 if v008-b3>11
replace age_interval=2 if v008-b3>23
replace age_interval=3 if v008-b3>35
replace age_interval=4 if v008-b3>47
*to check against report
tab age_interval v465 [iweight=weight]
The output and relevant comparison section in your report, are attached in one excel file.
Thanks in advance for your response.
Elizabeth
|
|
|
|
| Re: Issue in child age variables in DHS08? [message #3383 is a reply to message #3381] |
Tue, 02 December 2014 16:56  |
Trevor-DHS
Messages: 805
Registered: January 2013
|
Senior Member |
|
|
Hi Elizabeth,
It appears that there is an error in the age in months panel of table 12.13 in the Egypt DHS 2008 report. This variable was incorrectly constructed for this table. An errata will be issued for this table. Your logic and results are correct.
A small note - you can shorten your code for creating age_interval to just:
gen age_interval=int((v008-b3)/12)
[Updated on: Tue, 02 December 2014 16:58] Report message to a moderator |
|
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
