Re: [S] Imputation

Alan Zaslavsky (zaslavsk@hcp.med.harvard.edu)
Tue, 31 Mar 1998 10:40:18 -0500 (EST)


> From: "Marc R. Feldesman" <feldesmanm@pdx.edu>
> Question 1 (not an S or S+ question). I've read so many messages on this
> group that mention imputation that I'm wondering whether there is one
> particular or "best" reference on this subject - in particular one that
> deals with the strengths and weaknesses of the various approaches.

I recommend:

Little and Rubin: Statistical Analysis with Missing Data
Rubin: Multiple Imputation for Nonresponse in Surveys
Schafer: Analysis of Incomplete Multivariate Data

The first is a broad overview of analytical methods (of which
imputation is only one). The second deals in detail with the theory
of the multiple imputation approach (the most general imputation
approach). The third dots all the i's and crosses all the t's for
application of that approach to a fairly broad class of specific
models.

> Question 2 (an S or S+ question). Is there an add-on set of functions that
> implements the general class of imputation methods that most find useful?

Schafer has a set of computer programs that are available publicly. The
book says to look at http://www.stat.psu.edu/~jls/

A version of these programs may appear in a future version of S-Plus but I
leave it to Mathsoft to respond to that authoritatively.

Alan Zaslavsky
zaslavsk@hcp.med.harvard.edu
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news