Home » Data » Dataset use in Stata » sampling stratification of the Cameroon DHS 2011, 2004, 1998 and 1991 data
| Re: sampling stratification of the Cameroon DHS 2011, 2004, 1998 and 1991 data [message #10226 is a reply to message #10221] |
Mon, 11 July 2016 10:46   |
Bridgette-DHS
Messages: 3062
Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS Stata Specialist, Tom Pullum:
Just some minor modifications. In the syntax for svyset, both the psu and the weight go before the comma, and the psu does not go in parentheses.
It is not necessary to divide v005 by 1,000,000; Stata will automatically normalize the weights to have a mean of one. You can do it if you want but it will have no effect.
I have not looked at the four surveys but for many of the older DHS surveys the stratum variable is incorrect. I would use the combinations of regions and place of residence.
It's a good idea to include the singleunit option. Otherwise you will probably crash.
So this is what I would recommend:
egen strata = group(v024 v025), label
svyset v021 [pweight=weight], strata(strata) singleunit(centered)
If you put all of the surveys into a single file, then the only change I would recommend is that the surveys be numbered (survey=1, 2, 3, 4, for example) and you do the following:
egen clusters=group(survey v021), label
egen strata = group(survey v024 v025), label
svyset clusters [pweight=weight], strata(strata) singleunit(centered)
|
|
|
|
Current Time: Mon May 20 22:41:41 Coordinated Universal Time 2024
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
