| Re: using svy function to run cross tabs [message #8403 is a reply to message #8389] |
Fri, 23 October 2015 17:07  |
Reduced-For(u)m
Messages: 292
Registered: March 2013
|
Senior Member |
|
|
One option would be to run OLS ("reg") with dummy variables for the 3 categories of v761 you want (no constant). Then you could use the "test" commands to test the equality of the groups (in whatever pair-wise or all-at-once way you wanted).
That would account for all of the survey design as long as you used the "svy" prefix before the regression. I think that should work.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
