The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Nigeria » Table D.4 Births by calendar years- 2013 NDHS
Table D.4 Births by calendar years- 2013 NDHS [message #5610] Tue, 16 June 2015 10:08 Go to next message
omogeonago is currently offline  omogeonago
Messages: 5
Registered: May 2015
Member
Reposting because I got no response to my earlier post dated 10 June. Thanks
-----------------------------------------------------

Hi,
I am trying to replicate "Table D.4 Births by calendar years" on page 399 of the Nigeria 2013 final report. Using the birth recode dataset with weights applied, I tabulate variable "b2" but my numbers don't align at all with the report. However, when I use the same code with the Burkina 2010 birth recode in order to replicate the same table i.e. "Tableau C.4 Naissances par année de naissance" on page 363, my results match the numbers in the Burkina report. Similarly, when I do the same for Congo DR 2013/14, my numbers match those in the report.

Can you please let me know if the calculation of this table is different for Nigeria. Thanks

gen wt=v005/1000000
egen strata=group(v024 v025)
svyset [pw=wt], psu(v021) strata(strata)
svy:tab b2, missing count stubwidth(20) format(%11.10g)
svy, subpop(if b2==2010):tab b5, missing count stubwidth(20) format(%11.10g)
Re: Table D.4 Births by calendar years- 2013 NDHS [message #6717 is a reply to message #5610] Wed, 01 July 2015 17:28 Go to previous message
Liz-DHS
Messages: 1381
Registered: February 2013
Senior Member
Dear User,
Here is a response from one of our technical experts, Dr. Tom Pullum:
Quote:
When I open NGBR6AFL.dta and crosstab b2 and b5 with weights, I can match table D.4 in the report, but with an interesting error! The row for 2013 in the table should be labelled 2011. The row for 2012 in the table should be labelled 2010, etc.

The correct numbers for 2012 and 2013 are not stated. You have probably calculated them correctly. This is what I get:

     Year        Dead      Alive      All
      2012 |449.923453  6,845.892 | 7,295.816 
      2013 | 96.092803  1,911.094 | 2,007.186

I have not rounded to the nearest integer. When I looked at table D.4 I could immediately see there was a problem, because the Nigeria 2013 DHS was conducted early in the year, and there had to be many fewer births in 2013 than in 2012. As it is, those two numbers are about the same. It appears that the omission of 2012 and 2013, and the incorrect labeling of the years 2011 and earlier, carried over to the groupings of years given in the table. For example, I see that the total number of births for 2009-2013 is the sum of the stated numbers for 2009-2013, not the correct numbers.

By the way, for a table like this there is no point in making the stratum and cluster adjustments. Those adjustments will not make any difference. I just used this command: "tab b2 b5 [iweight=v005/1000000]".

I will inform our data processing people of this error, although I hope someone had already noticed it. Thanks for calling attention to it.

Previous Topic: Strata Computation for Survey Sampling Weighting
Next Topic: Value labels for LGAs not available
Goto Forum:
  


Current Time: Tue Nov 13 00:42:45 Eastern Standard Time 2018