[R] Missing data analysis in R

[Marc Schwartz]

On Tue, 2006-10-31 at 11:17 -0600, Inman, Brant A. M.D. wrote:
> I am looking for a book that discusses the theory of multiple imputation
> (and other methods of dealing with missing data) and, just as
> importantly, how to implement these methods in R or S-Plus.  Ideally,
> the book would have a structure similar to Faraway (Regression),
> Pinheiro&Bates (Mixed Effects) and Wood (GAMs) and would be very modern
> (i.e. published within the last couple of years).  
> 
> Any ideas?  If such a book does not exist, one of the experts on this
> help list should write it! (I will gladly buy the first copy.)
> 
> Brant Inman
> Mayo Clinic

One of the better references is:

Statistical Analysis with Missing Data, Second Edition
by Roderick J. A. Little, Donald B. Rubin
Wiley-Interscience; 2nd edition (September 9, 2002)
ISBN: 0471183865
http://www.amazon.com/Statistical-Analysis-Missing-Data-Second/dp/0471183865


In addition, see Frank Harrell's book "Regression Modeling Strategies",
Chapter 3 on Missing Data.  More information here:

  http://biostat.mc.vanderbilt.edu/twiki/bin/view/Main/RmS

and see Frank's function 'aregImpute' in the Hmisc package. More
information is here:

  http://biostat.mc.vanderbilt.edu/twiki/bin/view/Main/Hmisc

with function specific help here:

  http://biostat.mc.vanderbilt.edu/s/Hmisc/html/aregImpute.html



Within R, if you use:

  RSiteSearch("Missing Data")

you will also get many hits.



Finally, the Multivariate task view on CRAN has a Missing Data section
about half way down the page:

  http://cran.r-project.org/src/contrib/Views/Multivariate.html

HTH,

Marc Schwartz