weighting [message #5531] |
Wed, 03 June 2015 11:13 |
asifalokozai
Messages: 1 Registered: June 2015 Location: Afghanistan
|
Member |
|
|
Dear Experts
In surveys when the subject or sample has been selected with replacement, so are the best method to weight such data. e.g EPI coverage survey.
Thanks
Asif
M.Asif Alokozai
|
|
|
Re: weighting [message #5583 is a reply to message #5531] |
Wed, 10 June 2015 12:19 |
Trevor-DHS
Messages: 802 Registered: January 2013
|
Senior Member |
|
|
We can't speak specifically for other surveys, but in general with most surveys you will need to use weights, even if the sample has been selected with replacement. Most surveys like DHS use two stage samples - selecting clusters at the first stage and selecting households within clusters at the second stage. The weights are used to adjust for the probability of selection of the clusters. The second stage of selecting households is where surveys other than DHS use replacement. You thus should use weights (in general - specific surveys may differ)
|
|
|