Rwanda 2000 only 13 unique women's weights (v005) [message #12204] |
Sun, 09 April 2017 20:02 ![Go to next message Go to next message](/theme/default/images/down.png) |
jkweku
Messages: 2 Registered: April 2017
|
Member |
|
|
Dear forum,
I am using the children's recode of the Rwanda 2000 survey for some analysis. The v005 variable for the women's sample weight has only 13 unique values. This matches the number of regions in the survey. I am baffled as to how this can be the individual weight. Should I multiply this by the household weight (after dividing both by 1,000,000) to get the correct individual woman's sample weight?
Many thanks!
|
|
|
Re: Rwanda 2000 only 13 unique women's weights (v005) [message #12215 is a reply to message #12204] |
Tue, 11 April 2017 10:59 ![Go to previous message Go to previous message](/theme/default/images/up.png) ![Go to next message Go to next message](/theme/default/images/down.png) |
Bridgette-DHS
Messages: 3143 Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS Stata Specialist, Tom Pullum:
You should use hv005 when analyzing data in the HR or PR files, v005 when analyzing data in the IR, KR, or BR files, mv005 for the MR and CR files, and hiv05 for the AR file. You would never multiply weights together as you suggest.
I just looked at v005 in RWIR41 and see the same thing, which I had not noticed before, even though I have used that file many times. I agree that it is very strange to have only 13 values. Usually v005 is cluster-specific. However, I recommend that you go ahead and do what you would do with any other survey. I will ask others at DHS for an explanation and let you know what they say.
|
|
|
|
Re: Rwanda 2000 only 13 unique women's weights (v005) [message #12223 is a reply to message #12217] |
Wed, 12 April 2017 16:00 ![Go to previous message Go to previous message](/theme/default/images/up.png) |
Bridgette-DHS
Messages: 3143 Registered: February 2013
|
Senior Member |
|
|
Tom Pullum investigated further, and got the following response from Senior DHS Sampling Specialist, Ruilin Ren:
The Rwanda 2000 DHS survey used a self-weighting sample at survey domain level, with two urban domains and 11 rural domains, so the weight variables (both HV005 and V005) have only 13 different values. This can be achieved by a variant number of HH selected per cluster, in relation with the number of HH listed and the number of HH in the frame. Appendix A of the final report did not give the details.
|
|
|