The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Weighting data » Clustered Standard Errors
Re: Clustered Standard Errors [message #22115 is a reply to message #9278] Wed, 03 February 2021 11:09 Go to previous messageGo to previous message
analyst_till is currently offline  analyst_till
Messages: 10
Registered: January 2021
Member
Dear all,

I am operating a diff-in-diff analysis at the district-level with the Ethiopian DHS surveys of 2000 and 2011 and have a similar issue. Mr. Pullums answer was very helpful, however, I am still wondering about how I can adjust the clustering in the svyset specification.

More precisely, I want to cluster standard errors at the district-level (since this is my pseudo-panel cohort), which is not possible in a "svy:" regression.

If I now regress: "reg y i.post##i.Treatment i.district [pweight=v005], cluster(district)", I can account for weights and cluster standard errors at district level. However, in this case I do not take the stratification into account.
Would it in general also be okay to ignore the stratification in my analysis to be able to cluster SE at district-level?

Or, could you tell if there is a code to directly incorporate the fact that I cluster SE at district level e.g. through?:

svyset cluster_ID [pweight=v005], cluster(district) strata(stratum_ID) singleunit(centered)

Thanks in advance!

 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic: Confirming correct survey design
Next Topic: What is the difference between v005 and sweight DHS India?
Goto Forum:
  


Current Time: Sun Oct 2 00:01:46 Coordinated Universal Time 2022