Re: Missing value in Birth file-NFHS4 [message #14360 is a reply to message #14353] |
Tue, 27 March 2018 12:50  |
Trevor-DHS
Messages: 805 Registered: January 2013
|
Senior Member |
|
|
1) It isn't 50% are oversampled. It is that the enumeration areas that people from scheduled tribes lived in were, roughly on average, sampled at twice the level that they represent in the population, and to correct for this in national data, they receive half the weight (roughly).
2) for points 2 &3, you asked me why so many were missing. The reason they are missing is because of the state specific modules selection, and you can confirm this by crosstabbing v701 or v714 with ssmod with the missing data option on. All cases for which ssmod is 1 will be missing for v701 or v714, so yes, obviously the frequency of the non-missing values are going to be exactly the same whether you filter on ssmod == 1 or not.
|
|
|