| Re: Weighting Data for Pooled Indonesian DHS Dataset [message #10051 is a reply to message #10050] |
Mon, 20 June 2016 09:12   |
adelia
Messages: 4
Registered: May 2016
|
Member |
|
|
Thank you very much for your reply.
Regarding the weighting, I did the re-normalization as follow for IDHS 2007 BR, is this correct?
. gen wgt=v005/1000000
. gen wgt_denorm=v005*64140000/32895
. gen wgt_denorm_des=wgt_denorm/1000000
. sum v005 wgt wgt_denorm wgt_denorm_des
Variable | Obs Mean Std. Dev. Min Max
-------------+---------------------------------------------- -----------
v005 | 84,726 955475.8 1167341 21872 7968061
wgt | 84,726 .9554758 1.167341 .021872 7.968061
wgt_denorm | 84,726 1.86e+09 2.28e+09 4.26e+07 1.55e+10
wgt_denorm~s | 84,726 1863.025 2276.129 42.64691 15536.45
My intention is to use this weight to generate a pooled wealth index using IDHS 2007 and 2012. I think for this, in the pooled data the weighted sample sizes need to be proportional to the number of women in the population, no? Or the same weighted number of cases in each survey is enough?
Many thanks beforehand
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
