|
Re: Weighted result's total gives different total value with unweighted one. [message #19435 is a reply to message #19433] |
Thu, 18 June 2020 07:55 ![Go to previous message Go to previous message](/theme/default/images/up.png) |
Bridgette-DHS
Messages: 3152 Registered: February 2013
|
Senior Member |
|
|
Following is a response from DHS Research & Data Analysis Director, Tom Pullum:
The child weights in the KR file are the same as child weights in the BR file and the same as the mother's weights in the IR file. They are normalized to have a mean of 1 (multiplied by 1 million to eliminate the need for a decimal point) in the IR file. However, that does not mean that the weights will have a mean of 1 in the BR and KR files. Women with different weights have different numbers of children.
When you use v005 as a pweight in Stata, it always re-normalizes the weights to have a mean of 1, meaning that the total weight in any estimation command will be the same as the number of cases. So what you are seeing is ok--not a problem.
|
|
|