| Revisiting the topic of weighting data [message #30695] |
Tue, 21 January 2025 04:09  |
acseng
Messages: 8
Registered: November 2024
|
Member |
|
|
Dear colleague,
I would like your advice and opinion. I accept my question has been discussed many times in this forum, and I have already read most of the replies, including the file pool_and_reweight_surveys_do_22Ot2024.txt written by Tom Pullum.
My point to revisit is on how to append different DHS databases and weights: let's say I want to use data from children under 5 years using KR databases from the same country and from 4 different survey years. My intention is to create a database that after adjustment for potential confounders would represent a hypothetical average population. My questions:
1) in the Stata routine in the txt text, what is the target population to generate the var factor? a)the total population, b) women 15-49 yo, or c) children <5yo?
2)would it make sense to use weightr as wtr=hv005r/1000000/4 to create a kind of "average" population of the 4 years?
Thank you in advance for your help!
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
