Re: st: Grouping income variables- RECODE COMMAND

[Maarten Buis <maartenlbuis@gmail.com>]

On Tue, Feb 4, 2014 at 1:48 PM, Antonio Rodriguez Andres wrote:
> Thank you very much for your feedback. What I did is the following
>
> http://www3.nd.edu/~rwilliam/stats2/l12.pdf

Rich quotes Allison in that handout:

This technique of including a missingness indicator variable is
"remarkably simple and intuitively appealing." But unfortunately, "the
method generally produces biased estimates of the coefficients."

The logic of why that leads to biased estimates is discussed in the
link I gave before. So my recommendation is still "do not do that".

> **dummy indicator for missing income values
>
> gen xhincome=hincome
> replace xhincome= 29304.99 if missing(hincome)
> gen md=0
> replace md=1 if xhincome! =hincome
>
> xtmixed dprt age age2 gender married separated divorced widowed eduyrs ichldhm md lhincome ihealth iuemp5yr iuemp12m rgdp06[pw=dweight] if md==0 || cntry: gender , mle
>
>
> But I still got the same message, the md indicator variable is dropped. How can İ estimate the model controlling for missing values in income?

a) I repeat: the method you want to use does _not_
control/adjust/solve/remove for missing values. So you should not use
it.

b) you probably still have the same missing values on lhincome, so in
the estimation sample md is still a constant. (I am delibaretly not
telling you how to solve it, as I don't want you to do something that
is wrong)

-- Maarten

---------------------------------
Maarten L. Buis
WZB
Reichpietschufer 50
10785 Berlin
Germany

http://www.maartenbuis.nl
---------------------------------

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/