Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: how to handle missing observations in a regression model


From   "Rodrigo A. Alfaro" <ralfaro76@hotmail.com>
To   <statalist@hsphsun2.harvard.edu>
Subject   Re: st: how to handle missing observations in a regression model
Date   Tue, 5 Sep 2006 09:55:08 -0400

In Stata you have -ice- and -hotdeck- procedures for missing data. The first 
is based on Chained-Equations technique (CE), which is explained at 
http://www.multiple-imputation.com/, the second is regression technique... 
an interesting analysis is available in Efron, B. (1994). Missing Data, 
Imputation, and the Bootstrap. Journal of the American Statistical 
Association, 89, 463-478. A third technique based on EM algorithm (basically 
Maximum Likelihood) and it uses Bayesian-principle is available in the 
free-software NORM by J. Schafer [His book (1997) Analysis of Incomplete 
Multivariate Data (1997) is a classic in the Multiple Imputation literature 
as Rubin (1987). Multiple Imputation for Nonresponse in Surveys]. EM has 
well-known asymptotic properties in compare with CE, but CE has good results 
in Monte Carlo experiments. I suggest you NORM, because it is pretty easy to 
use and it is very fast in compare with -ice-. If you can use SAS as well P. 
Allison wrote a code that mimics NORM. See his webpage 
http://www.ssc.upenn.edu/~allison/ and you can have an introductory lesson 
with this paper: http://www.ssc.upenn.edu/~allison/MultInt99.pdf. Rodrigo.


----- Original Message ----- 
From: "Richard Williams" <Richard.A.Williams.5@ND.edu>
To: <statalist@hsphsun2.harvard.edu>
Sent: Tuesday, September 05, 2006 9:15 AM
Subject: Re: st: how to handle missing observations in a regression model


At 05:56 AM 9/5/2006, Joseph Coveney wrote:
>You can explore the behavior of this approach using -simulate- with a
>data-generating process that mimics what you expect prevails in your study.
>(This includes the mechanism of missingness.)  A rudimentary example of 
>this
>is shown below.  It has 5% randomly missing in both predictors.  The 
>results
>indicate that for this approach, compared to just listwise deletion, there

One of the interesting points in Allison's Missing Data book is that,
of all the more or less traditional approaches to handling missing
data, listwise deletion tends to work as well or better as
anything.  You have to go to the more recent and advanced techniques
if you want to do better.


-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
FAX:    (574)288-4373
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu
WWW (personal):    http://www.nd.edu/~rwilliam
WWW (department):    http://www.nd.edu/~soc

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index