Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Dummy variables for missing values


From   "Alan Acock" <acock@comcast.net>
To   <statalist@hsphsun2.harvard.edu>, <statalist-digest@hsphsun2.harvard.edu>
Subject   st: Dummy variables for missing values
Date   Sat, 12 Feb 2005 08:07:17 -0800

On Sat, 12 Feb 2005 00:48:28 -0500, Daniel Egan <dp.egan@gmail.com> wrote:
> Hi Quang,
> 
> 
> > My friend has a panel data in the following format:
> >
> > Year        Investment
> > 1900        . (Missing value)
> > 1930        .
> > 1960        12
> > 1980         .
> > 1990        28
> > 1991        56
> > 1992        35
> > 1993        45
> > etc.,
> >
> > Could you please tell me how she can use the dummy variables for this
> > exercise (regression)? Also, how she can interprete the coeficients
> > for Investment and the dummy variables?
> >
A better solution to missing data is available with Stata. Do a findit on
mvis. This set of commands imputes multiple datasets, does your regressions
on each of these, and combines them to obtain unbiased estimates and
unbiased standard errors. Imputing the missing values make sense if the
missing values are "missing at random." If this is not the case, then the
assumption can be approximated by including "mechanism" variables, i.e.,
variables that predict patterns of missingness.
Alan Acock


*
*   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