Home » Data » Weighting data » Pooled Cross sections
| Re: Pooled Cross sections [message #20217 is a reply to message #20192] |
Fri, 09 October 2020 11:57   |
Bridgette-DHS
Messages: 3039
Registered: February 2013
|
Senior Member |
|
|
Following is another response from Senior DHS Stata Specialist, Tom Pullum:
This looks fine to me. The use of egen group is appropriate. Your adjustment to the weights is appropriate, but I believe that for some analyses it will have no effect (that is, the results with adjusted weights will match the results with unadjusted weights).
Yes, using iweight with tab is fine.
I am not familiar with ivreg2 (I see it is in an econometrics package you can add to Stata). Does it allow pweights and the cluster option? If so, I would recommend using them (pweights and the cluster option). Then not being able to use svy will only mean you are dropping the stratification adjustment.
In general, when a complex estimation procedure does not allow some option (such as svy) I would recommend pairing it with a simpler procedure that DOES allow svy. Check them against each other to see what is gained or lost by using the more complex procedure. Then it's a judgment call about which one to use. I would usually go with the one that is more conservative, that is, has fewer significant coefficients, and NOT with the one that is more favorable for your research hypothesis.... Good luck!
|
|
|
|
Current Time: Thu Apr 25 05:34:03 Coordinated Universal Time 2024
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
