[R] Multiple imputation for nominal data

[John Sorkin]

Thank you!
John




John David Sorkin M.D., Ph.D.
Chief, Biostatistics and Informatics
University of Maryland School of Medicine Division of Gerontology
Baltimore VA Medical Center
10 North Greene Street
GRECC (BT/18/GR)
Baltimore, MD 21201-1524
(Phone) 410-605-7119
(Fax) 410-605-7913 (Please call phone number above prior to faxing)>>> Andrew Miles <rstuff.miles at gmail.com> 11/2/2010 3:59 PM >>>
There are a couple of packages that do MI, including MI for nominal  
data.  The most recent of these is "mi", but I believe "mice" might do  
it as well.  Both are available on the CRAN, and both have useful  
articles that teach you how to use them.  The citations for these  
articles can be found at the bottom of the help page that appears by  
typing

?mi
OR for mice
?mice

mi is the newer package and has some useful control features, but as  
it is newer it still is under development.

Andrew Miles


On Nov 2, 2010, at 3:38 PM, John Sorkin wrote:

> I am looking for an R function that will run multiple imputation  
> (perhaps fully conditional imputation, MICE, or sequential  
> generalized regression) for non-MVN data, specifically nominal data.  
> My dependent variable is dichotomous, all my predictors are nominal.  
> I have a total of 4,500 subjects, 1/2 of whom are missing the main  
> independent variables. I would appreciate any suggestions that the  
> users of the listserver might have.
> John
>
>
> John David Sorkin M.D., Ph.D.
> Chief, Biostatistics and Informatics
> University of Maryland School of Medicine Division of Gerontology
> Baltimore VA Medical Center
> 10 North Greene Street
> GRECC (BT/18/GR)
> Baltimore, MD 21201-1524
> (Phone) 410-605-7119
> (Fax) 410-605-7913 (Please call phone number above prior to faxing)
>
> Confidentiality Statement:
> This email message, including any attachments, is for ...{{dropped:18}}