The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » India » India DHS 2015-2016 Births Recode Data ( Missing standard errors because of stratum with single sampling unit)
India DHS 2015-2016 Births Recode Data [message #21898] Wed, 06 January 2021 18:29 Go to next message
zkayani is currently offline  zkayani
Messages: 1
Registered: January 2021
Member
Hi,

I am having trouble creating a logistic regression model while using the 2015-2016 India DHS Births Recode dataset. I am trying to adjust for a weighted analysis and have used the following code:

svyset [pw=wgt], psu(v021) strata(v023)
svy:ologit b1 quartageatbirth2 quartageatbirth3 quartageatbirth4, or

Survey: Ordered logistic regression

Number of strata = 2,514 Number of obs = 1,315,617
Number of PSUs = 28,518 Population size = 1,289,652
Design df = 26,004
F( 0, 26004) = .
Prob > F = .

------------------------------------------------------------ ----------------------
| Linearized
b1 | Odds Ratio Std. Err. t P>|t| [95% Conf. Interval]
-----------------+------------------------------------------ ----------------------
quartageatbirth2 | .9567935 . . . . .
quartageatbirth3 | .9242713 . . . . .
quartageatbirth4 | .8839336 . . . . .
-----------------+------------------------------------------ ----------------------
/cut1 | -2.388488 . . .
/cut2 | -1.651026 . . .
/cut3 | -1.125956 . . .
/cut4 | -.7196191 . . .
/cut5 | -.3491667 . . .
/cut6 | .0052559 . . .
/cut7 | .3622409 . . .
/cut8 | .775032 . . .
/cut9 | 1.171656 . . .
/cut10 | 1.707994 . . .
/cut11 | 2.521563 . . .
------------------------------------------------------------ ----------------------
Note: Estimates are transformed only in the first equation.
Note: Missing standard errors because of stratum with single sampling unit.

I am particularly confused as to why I am getting the error note specifying there are strata with only a single sampling unit and unsure as to how proceed. When I conduct these analyses (using the same svyset and svy:ologit command on data from other nations in the DHS data, I do not get the same errors and thus have the appropriate standard error estimates. Is this something that is unique to the India data or is my command code incorrect. I have tried reading up on this issue online and rehashing my command code, but have had no succeed.

I would greatly appreciate any assistance on this matter!

Regards,
Zeb
Re: India DHS 2015-2016 Births Recode Data [message #21907 is a reply to message #21898] Thu, 07 January 2021 09:51 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3016
Registered: February 2013
Senior Member

Following is a response from Senior DHS Stata Specialist, Tom Pullum:

You just need to add the "singleunit" option to svyset, as follows: "svyset [pw=wgt], psu(v021) strata(v023) singleunit(centered)".

Previous Topic: Stillbirth risk factor
Next Topic: NFHS-4 India sample design in R
Goto Forum:
  


Current Time: Thu Mar 28 05:11:59 Coordinated Universal Time 2024