The DHS Program User Forum      
Discussions regarding The DHS Program data and results
Home » Topics » Nutrition and Anthropometry » How do i get the sample size as Table 11.1 of NDHS 2011 (Page 166-7)
How do i get the sample size as Table 11.1 of NDHS 2011 (Page 166-7) [message #10691] Tue, 30 August 2016 13:15 Go to next message
user is currently offline  user
Messages: 11
Registered: March 2014
Member
Hi,

I tried to calculate the prevalence of Stunting among Nepalese children using PR file (NDHS 2011). However, my calculations resulted in different sample size than the NDHS 2011 final report (2473 Vs 2475) from the Table 11.1 of the final report of NDHS 2011. Can anyone please suggest me what should i do to get the correct number of children (2475)?

The stata codes i had used:

*Declaring Survey data
gen finalwt=hv005/1000000
svyset hv001 [pweight=finalwt], strata(hv022) vce(linearized) singleunit(missing) || hv002

* Calculation of Stunting (Height for Age)
gen HAZ=hc70
replace HAZ=. if HAZ>=9996
gen stunting=.
replace stunting=0 if HAZ ~=.
replace stunting=1 if HAZ<-200
svy: tab stunting


and it produced,


. svy: tab stunting
(running tabulate on estimation sample)

Number of strata = 25 Number of obs = 2,473
Number of PSUs = 287 Population size = 2,536.2319
Design df = 262


stunting proportion

0 .5922
1 .4078

Total 1

Key: proportion = cell proportion

Here, the total number of children included is less (2 less) than the NDHS reported, therefore it impacted on the proportion too.

Any help regarding this would be appreciated.

Sincerely,

[Updated on: Tue, 30 August 2016 13:16]

Report message to a moderator

Re: How do i get the sample size as Table 11.1 of NDHS 2011 (Page 166-7) [message #10692 is a reply to message #10691] Tue, 30 August 2016 14:49 Go to previous messageGo to next message
Trevor-DHS is currently offline  Trevor-DHS
Messages: 520
Registered: January 2013
Senior Member
You are comparing the unweighted number (2473) with the weighted number (2475) from the report. This is the wrong comparison. You should be comparing the weighted number (2536), and you will see that you have more cases. For the anthropometry table there are two other conditions that you need to include:
1) Selection for de facto household member - hv103==1
2) Selection for cases with no imputation of month of birth - hc33==1 (in this case this selection makes no difference as all cases had a complete month and year, but in general this condition should be included)
Thus your tabulation command should be:
svy: tab stunting if hv103==1 & hc33==1

However, after making this change I get 2485 cases. I checked this with the denominator found in the STATcompiler/API and it also had 2485. It appears that the dataset went through additional editing following the production of table 11.1, and that now the denominator is 2485. The proportions stunted don't change however.
Re: How do i get the sample size as Table 11.1 of NDHS 2011 (Page 166-7) [message #10695 is a reply to message #10692] Wed, 31 August 2016 00:14 Go to previous messageGo to next message
user is currently offline  user
Messages: 11
Registered: March 2014
Member
Thank you Trevor for your response. Now i get it. Thanks a lot.
Re: How do i get the sample size as Table 11.1 of NDHS 2011 (Page 166-7) [message #12843 is a reply to message #10692] Sat, 22 July 2017 17:19 Go to previous messageGo to next message
Sanjhama is currently offline  Sanjhama
Messages: 2
Registered: July 2017
Member
Hello,

I am been analysing NDHS data for 2011 and have encountered the same problem. The information posted above is helpful but would like to get the command in SPSS please.

As mentioned above, the weighted number is 2536, but I could not understand why there is need to involve hv103 and hc 33. When I use 2536 as denominator and numerator is 1034, the proportion of stunted is 40.609 which is close to the report (41%) but still is not exact. If you could please post the command for SPSS for generating both the unweighted (2430) and weighted number (2485)((as mentioned in STAT compiler.))

Thank you in advance
Sanjhama
Re: How do i get the sample size as Table 11.1 of NDHS 2011 (Page 166-7) [message #12881 is a reply to message #12843] Thu, 27 July 2017 11:41 Go to previous message
Sanjhama is currently offline  Sanjhama
Messages: 2
Registered: July 2017
Member
Hi,

My issue is solved. Did it.

Thank you
Previous Topic: Food questionnaire vars 412b and v414q
Next Topic: Children dataset sample size does not match with DHS Nigeria report (2008 and 2013)
Goto Forum:
  


Current Time: Tue Aug 22 06:57:27 Eastern Daylight Time 2017