The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Weighting data » Using region as a predictor in a regression analysis different from sample strata
Using region as a predictor in a regression analysis different from sample strata [message #11822] Tue, 14 February 2017 11:31 Go to next message
karoun11 is currently offline  karoun11
Messages: 4
Registered: December 2015
Location: Gainesville, FL
Member
Hello,
I am running linear regression analyses using STATA, in several East African countries, but each regression analysis is run separately for each country. I am using the child file, with the appropriate weights (svyset v021 [pweight = sampwt], strata(v022) vce(linearized) singleunit(scaled)) to run my regression analyses in STATA.

I am trying to determine that if I use a country's region as a predictor variable, whether it is incorrect to use an administrative unit (such as province or DHS designated region; v024) different than one one used in the stratification of the sample design(v022 or v023). I still use the svyset command to run my regression.

Thank you,
Karoun


Karoun H. Bagamian
Postdoctoral Associate, HEDGE lab
Emerging Pathogens Institute/Department of Environmental and Global Health
University of Florida
Re: Using region as a predictor in a regression analysis different from sample strata [message #11829 is a reply to message #11822] Wed, 15 February 2017 10:53 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:

It is ok to use a regional categorization that differs from the stratification in the survey design.
Previous Topic: Interrupted time series analysis using DHS data and applying sample weights
Next Topic: Weighing SPA data Tanzania
Goto Forum:
  


Current Time: Thu Mar 28 11:55:24 Coordinated Universal Time 2024