Multiple imputation

Article View: bit.listserv.sos-data

Article #337

Back to bit.listserv.sos-data Previous Next

#337

From: SHERMAN DORN
Date: Sun, 16 Jul 1995 00:00

18 lines
1649 bytes

I'm using several of the decennial census PUMS files to examine young adults
and high school graduation.  I am trying to use multiple imputation (a la Rubin)
to gauge how important the <15% of the sample living away from relatives is
to the parameters relating household economic circumstances to high school
graduation.  (In other words, when I have an odds ratio associating household
head's graduation status to the individual's graduation, how much am I losing
because some of the sample are *not* living as dependents of relatives?)

I am looking for others familiar enough with multiple imputation to look at
some preliminary results so I know I'm not entirely off the mark.  Or
just interested and willing to go for a bit of adventure in social-science
computing.  (I'm thinking that "just interested" means having read the Rubin
[or is willing to] enough to get the basic principles.)

Please reply by e-mail.

Sherman Dorn
dornsj@ctrvax.vanderbilt.edu

Technical Details

Message-ID: <Pine.CVX.3.91.950716102009.19124A-100000@gibbs.oit.unc.edu>

Path: arc1-ca.pugleaf.net!unknown!not-for-mail

Show Full Headers

Complete Article Headers

X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: fc2f8,2e17f9da4c329c53,start
X-Google-Attributes: gidfc2f8,public
From: SHERMAN DORN <DORNSJ@ctrvax.Vanderbilt.Edu>
Subject: Multiple imputation
Date: 1995/07/16
Message-ID: <Pine.CVX.3.91.950716102009.19124A-100000@gibbs.oit.unc.edu>
X-Deja-AN: 106486671
sender: sos-data@gibbs.oit.unc.edu
comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
errors-to: irsslmgr@gibbs.oit.unc.edu
precedence: bulk
x-listprocessor-version: 6.0 -- ListProcessor by Anastasios Kotsikonas
reply-to: DORNSJ@ctrvax.Vanderbilt.Edu
newsgroups: bit.listserv.sos-data,bit.listserv.sos-data
originator: sos-data@gibbs.oit.unc.edu