Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: F-value drops/shows after robust option depending on the magnitude of the independent variable


From   Kaspar Dardas <[email protected]>
To   [email protected]
Subject   Re: st: F-value drops/shows after robust option depending on the magnitude of the independent variable
Date   Tue, 12 Jan 2010 16:17:53 +0100

Hey Maarten,

yes it does. Very simple and plausible explanation.

Thank you.


2010/1/12 Maarten buis <[email protected]>:
> --- On Tue, 12/1/10, Kaspar Dardas wrote:
>> I have a question regarding the ",robust" option after
>> runnig a multivariate regression. One of my independ variables is
>> mcap (in Euros), which is often an extrem high number (some mcap
>> numbers have more than twelve digits) compared to other numbers in
>> my regression. Whenever I use the "robust" option the F-value
>> drops in the result table and becomes blue.
> <snip>
>> However, if I run the exact same multivariate regression
>> with mcap2 == mcap/100,000,000 and then use "robust". The F-value
>> shows again.
> <snip>
>> Is this an appropriate method to "restore" the F-value?
>
> Yes, dividing by a constant is just a linear transformation, so
> they are just equivalent models. From a substantive point of view
> an euro change in mcap is such a minute change that you don't care
> about its effect, you would rather be interested in the change of
> a million euro change in mcap. From a practical point of view, this
> can make your model numerically more stable, as you experienced.
> So it is a good thing whichever way you look at it.
>
> Hope this helps,
> 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/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index