The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Dataset use in Stata » Use of svy fxn in urban strata analysis
Use of svy fxn in urban strata analysis [message #15883] Wed, 03 October 2018 23:00 Go to next message
deni83 is currently offline  deni83
Messages: 1
Registered: February 2018
Location: Nigeria

Hello all,

I'm attempting to disaggregrate the Nigeria DHS 2013 data at the urban level by wealth index (1st and 2nd quintile-urban poor and 3rd-5th- other urban residents), and all rural residents (as exposure group) and then look at specified outcomes (ANC visits, place of delivery, stunting e.t.c). When I apply the svy fxn to do cross analysis, STATA states that it cannot calculate the SE and therefore cannot give any test statistics or Confidence interval.

Please how do I address this issue?

Thank you


Re: Use of svy fxn in urban strata analysis [message #15998 is a reply to message #15883] Wed, 17 October 2018 17:41 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1626
Registered: February 2013
Senior Member
Please see the response below from Senior DHS Stata Specialist, Tom Pullum:

Are you using "fxn" as an abbreviation for "function"? I just think of svyset, followed by svy, as an adjustment for the survey design. It should not depend upon your choice of outcomes or covariates and is available for most (but not all) estimation commands. I would not think that the disaggregation would cause a problem, but it's recommended that you include the "subpop" option in svy for subpopulations. Can you provide the "svyset" and "svy:" lines?

Previous Topic: Adjusting sample weight for multilevel analysis- SPA dataset
Next Topic: Variable midx
Goto Forum:

Current Time: Wed Jun 19 14:04:33 Eastern Daylight Time 2019