Thank you so much for the response.
I want to analyze how various characteristics (household's, mother's and father's) affect the children's schooling. I wanted to do a panel analysis, but from your response, it seems this is impossible.
"I see that this survey only had a subsample of men. That may be an issue for the kind of analysis you want to do." Do you think this will be an issue in my analysis?
I would appreciate clarification on when to renormalize weights.
I assumed that since I will be merging various files, then I would need to renormalize.
What I have understood is that it is better to use the weight from the file likely to have the highest nonresponse.
I will be using variables from all the 3 files (PR, MR and IR)and below is my svyset code after merging the files:
gen weight = mv005/1000000
svyset hv021 [pweight=weight], strata(hv023) singleunit(scaled)
Ps. I will later on merge the gps file to the already merged file. I hope I do not have to worry about weights when merging the gps file to the merged file.
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
