The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Bangladesh » Merging Community file with individual recode file
Merging Community file with individual recode file [message #9038] Sat, 30 January 2016 17:15 Go to next message
pgroy is currently offline  pgroy
Messages: 1
Registered: January 2016
Location: Illinois
Member
Hello,
I am trying to merge the community file (BDSQ) with the individual recode file (BDIR). I have merged it using the cluster number (COCLUST and v001) in the respective files. Is this the right way to do it? Can I analyze individual information with respect to the cluster? For example, if i were to examine association between birth weight and a specific community characteristic, could i do it using the merged file?

Any suggestion would be very helpful.

Thank you.

P G Roy


P G Roy
Re: Merging Community file with individual recode file [message #9054 is a reply to message #9038] Tue, 02 February 2016 14:49 Go to previous message
Reduced-For(u)m
Messages: 291
Registered: March 2013
Senior Member

Can I analyze individual information with respect to the cluster? For example, if i were to examine association between birth weight and a specific community characteristic, could i do it using the merged file?

- This is something people definitely do, and what you describe would be the right dataset merge (note: I didn't check the merge variables, but the idea is right). The biggest issue is probably going to be how much variation you actually get in the cluster level measure: how many clusters and how much your X of interest varies. But there is no particular reason why you can't do this, it is just likely to be a little bit noisy.

Also, remember that the cluster level information you are getting comes from the time of the survey, which is not necessarily the time of birth, so if things are changing over time in the region you will have some measurement error problems that will be correlated with date of birth (the further back you go the less reliably well the information will actually match what the mother experienced during pregnancy/birth. Some variables this is more of a problem than others, but it is up to your judgment.
Previous Topic: Repitition of the same variable
Next Topic: Only 2 cases in a PSU in DHS 2014 data
Goto Forum:
  


Current Time: Sun Mar 29 14:02:02 Eastern Daylight Time 2020