| Senegal 2015 - Birth registration [message #11610] |
Mon, 16 January 2017 15:30  |
fewerling
Messages: 19
Registered: November 2016
Location: Brazil
|
Member |
|
|
Dear all,
We are trying to reproduce the % of under-5 children with a birth registration for Senegal 2015 (Table 15.1 in the country report).
We are using the following codes, which we already used for other surveys, to calculate the results:
** PR dataset **
gen breg = .
replace breg = 0 if inrange(hv105,0,4) & hv102==1
recode breg (0=1) if inrange(hv140,1,2)
svyset hv021 [pweight=hv005]
svy: prop breg
. svy: prop breg
(running proportion on estimation sample)
Survey: Proportion estimation
Number of strata = 1 Number of obs = 7087
Number of PSUs = 214 Population size = 6279670112
Design df = 213
------------------------------------------------------------ --
| Linearized
| Proportion Std. Err. [95% Conf. Interval]
-------------+---------------------------------------------- --
breg |
0 | .3173947 .0209918 .2775279 .3601337
1 | .6826053 .0209918 .6398663 .7224721
------------------------------------------------------------ --
We found an estimate of 68.3% and a weighted sample size of 6,280, however the results in the report are 80.4% and sample size the same (6,280).
Was the estimated % from the report calculated from another dataset or we are missing something in the code?
Thank you!
Best regards,
Fernanda
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
