Declaring my Data Survey Data [message #13889] |
Thu, 18 January 2018 16:25 ![Go to next message Go to next message](/theme/default/images/down.png) |
Dangboye
Messages: 1 Registered: September 2017
|
Member |
|
|
Good afternoon everyone,
I am trying to do a difference in difference using DHS Data on Nigeria in 1990, 1999, 2003, 2008, and 2013. I am concerned about how to weight each observation as I am not sure using weight=hv105/1000000 from my pooled data works fine. Also, in declaring the data as a survey data, I wrote my command as:
svyset [pweight=weight], psu(hv021) strata(strata) having defining my hv023 as my strata. When I run this command, it states single unit missing and running a regression based on that yields to estimates with no t statistics, confidence intervals etc.
I would appreciate any response that could help me work through this.
Regards.
|
|
|
Re: Declaring my Data Survey Data [message #13891 is a reply to message #13889] |
Fri, 19 January 2018 07:03 ![Go to previous message Go to previous message](/theme/default/images/up.png) |
Bridgette-DHS
Messages: 3151 Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS Stata Specialist, Tom Pullum:
Yes, you are ok with that weight. I recommend that you add "singleunit(centered)" to the end of the svyset statement. There are other options besides "centered" but they all give about the same results.
|
|
|