| Question about weights to use [message #29868] |
Thu, 15 August 2024 04:48  |
|
|
Hello,
I have appended men and women DHS for Uganda and also the AIS survey. Which weights can I use?
I plan to use STATA and divide the variable v005 by 100,000 and to set svyset [pweight = weight], psu(v021) strata(v023). For simple descriptive statistics do I also need to use the pweights as I defined them above (using the command svy: ....)?
Hope that someone can help me. Thanks in advance!
|
|
|
|
| Re: Question about weights to use [message #29876 is a reply to message #29868] |
Thu, 15 August 2024 13:44  |
Bridgette-DHS
Messages: 3199
Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff member, Tom Pullum:
Which surveys and files are you using? DHS is not providing any support for the recently-completed survey, which was done without DHS participation. The only AIS I see listed is from 2011. Is that what you are using? DHS stopped doing AIS surveys several years ago.
Are you appending IR and MR files? From one survey or from multiple surveys? You have to rename the mv variables in the MR file to v variables. That includes changing mv005 to v005 and using that as the weight variable in the combined file. You have to add a variable "sex" to specify whether the cases are for women or men.
The weights have a factor of 1000000, not 100000, but the svyset command uses pweights, and for them the factor is irrelevant. Stata automatically renormalizes v005 to have a mean of 1.
Your svyset command is ok but you need to use "egen group" to construct unique ID codes for clusters and strata if you are combining multiple surveys. Please search the forum for posts that describe the steps.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
