Adjusting previously weighted variables to account for missing

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

Posts: 2
Joined: Sat Jan 14, 2017 11:48 pm

Adjusting previously weighted variables to account for missing

Postby medresearch » Sun Jan 15, 2017 4:56 pm


So I am working with an insurance database and the issue I am wrestling with is this:
The database specifically covers assorted discharged patient cases. In order to make the cases representative of the nation as a whole (since they're only about 20% of the hospitals nationwide) each hospital has a weight provided by the database owners (DISCWT).
I know how to wt cases in SPSS, but my issue is that for certain variables there are missing values. Keeping in mind that <5% of cases are missing in any given variable, do I need to re-weight the cases?
If so what is the method? My understanding is you can take the variable and multiply it by (total value of original weights/total value of weights without cases with the missing variable).
That is if there were 10,000 cases, but only 9,000 had an age, I would: DISCWT*(10,000/9,000) if I was using age for a calculation.
Any help in clarifying this would be appreciated. Thank you.
Posts: 2760
Joined: Tue Jun 12, 2007 12:08 pm
Location: Florida, USA

Re: Adjusting previously weighted variables to account for missing

Postby statman » Mon Jan 16, 2017 2:04 pm

If I understand then the weight should be applied as is and the MV will be accounted for based on the model being used, but .................
See the note below

NOTE: Please read the Posting Guidelines and always tell us your OS, the SPSS version and information about your study and data!

Statistical Services

Who is online

Users browsing this forum: No registered users and 1 guest