Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: dummy variable
From
"Jing Zhou" <[email protected]>
To
<[email protected]>
Subject
Re: st: dummy variable
Date
Fri, 16 Sep 2011 20:06:22 +1000
Thanks a lot, Maarten. your suggestions are very appreciated. Can you
please provide the reference of "The consequence of the fact that there
are so few 1s is that the variance of the variable will be low, and thus
the precision with which the effect of that variable is measured will
also be low, i.e. large standard errors and confidence intervals."?
Regards,
Jing
>>> Maarten Buis <[email protected]> 16/09/11 7:30 下午 >>>
If you want to get results that control for that than you should add
that variable, otherwise you should not do so. If you want to add that
variable I would first try to find out why they are missing. For
example, is there only one shareholder, and is it thus impossible for
anyone to be the second largest shareholder or is the second largest
shareholder unknown. In the former case you can set those missing
values at zero (and possible add a dummy variable for single
shareholder), while in the latter case you can think of multiple
imputation (see: -help mi-).
The consequence of the fact that there are so few 1s is that the
variance of the variable will be low, and thus the precision with
which the effect of that variable is measured will also be low, i.e.
large standard errors and confidence intervals.
Hope this helps,
Maarten
On Fri, Sep 16, 2011 at 10:51 AM, Jing Zhou <[email protected]>
wrote:
> Thanks Maarten, it measues whether the second largest shareholder of a
> listed company is a state shareholder.
>
> Jing
>
>
>>>> Maarten Buis <[email protected]> 16/09/11 6:02 下午 >>>
> On Fri, Sep 16, 2011 at 2:09 AM, Jing Zhou wrote:
>> I have a sample period from 2000-2009. now I am considering to add a
> dummy variable into the regression. for the year of 2009, only 24 out
of
> 550 observations (<5%) valued at 1 of this dummy. for 2008, 31 out of
> 502 observations (<5%) valued at 1. further, for the year of
2000-2002,
> the missing value rate is 61.19%, 31.91%, and 13.33%, respectively.
for
> other variables of the regression, i can get relatively complete data
> value of each observation. therefore, is it still necessary to include
> this dummy in my regression? thanks!
>
> What is this dummy variable supposed to measure?
>
> -- Maarten
>
> --------------------------
> Maarten L. Buis
> Institut fuer Soziologie
> Universitaet Tuebingen
> Wilhelmstrasse 36
> 72074 Tuebingen
> Germany
>
>
> http://www.maartenbuis.nl
> --------------------------
>
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
--
--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany
http://www.maartenbuis.nl
--------------------------
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/