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 #22281 is a reply to message #22277] Tue, 23 February 2021 12:12 Go to previous messageGo to previous message
analyst_till is currently offline  analyst_till
Messages: 10
Registered: January 2021
Member
Yes, the above described situation fits my case and I understand why we adjust for the sampling design.

My confusion was more specifically about at what level I should cluster standard errors. Abadie et al. (2017) argue that clustering SEs is justified either by the sampling design (the case for DHS), or by the experimental design, when clusters of units are assigned to the treatment.

For my analysis the latter case also applies, since districts were assigned to the programme I evaluate. Therefore, I was wondering whether sampling process or experimental design have to be addressed in the first place.
So, I guess it is the sampling design that has to be addressed in the first place, as you mentioned, and therefore standard errors should be clustered at PSU-level here.

Thanks a lot for your help again!
 
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: Fri Mar 29 10:16:55 Coordinated Universal Time 2024