Re: Using weights for merged PR&IR file NFHS-4(2015-16), India [message #16670 is a reply to message #16651] |
Thu, 14 February 2019 13:26 ![Go to previous message Go to previous message](/theme/default/images/up.png) |
preshit
Messages: 13 Registered: March 2018 Location: Tucson, AZ, USA
|
Member |
|
|
Tom,
Thank you so much for your reply. My primary question was on which weights to use in my case and you have clarified my doubts. Regarding my logit models, I have created caste dummies on sh35 and running separate models for each of them. You are right, Fairlie decomposition is for binary dependent variable and that's why I have created caste dummies. When I rechecked my codes, I found that my problem was due to estimation command ado file and sorting I did while creating new variables. I was able to resolve it now. Thank you so much once again for your prompt reply. I really appreciate it.
Regards
Preshit
|
|
|