| Merging and weighting PR and KR files for under-five children [message #31427] |
Wed, 24 September 2025 09:44  |
Yiqun Luan
Messages: 2
Registered: February 2021
|
Member |
|
|
Hello,
My research team and I are conducting an analysis to estimate the proportion of children under five who are simultaneously stunted, have mothers with low education, and whose mothers report low frequency of internet use. For this purpose, we are working with the KR files. We would greatly appreciate your guidance on the following methodological questions:
1. Can under-five children who appear in the PR file but not in the KR file (because their mothers were not interviewed) be merged into the KR dataset for analysis?
2. If so, what weight should be assigned to these children? Should household weights (hv005) be used, or is there a recommended approach to deriving appropriate weights for them?
3. If household weights are used for these added children, how should they be integrated with the children in the KR files who already have women's weights (v005), in order to produce a single set of estimates?
We are raising these questions because we want to avoid excluding children whose mothers were not interviewed - since they may represent a particular underprivileged group - while at the same time applying the appropriate weights to ensure nationally representative estimates when working on the KR files after the merge. Any guidance, documentation, or best practices you could share would be extremely helpful for our analysis.
Thank you very much for your time.
|
|
|
|
|
|
|
|
|
|
Members
Search
Help
Register
Login
Home
