Bookmark and Share

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: Outlier diagnostics for tobit (postestimation)


From   "JVerkuilen (Gmail)" <[email protected]>
To   [email protected]
Subject   Re: st: Outlier diagnostics for tobit (postestimation)
Date   Fri, 19 Oct 2012 09:34:33 -0400

On Fri, Oct 19, 2012 at 4:36 AM, Maarten Buis <[email protected]> wrote:
>
> Here I have to disagree. Graphical methods can also work for bounded
> variables if you know what kind of pattern to expect if the model is
> true.

The latter has been, in my experience a pretty big if which is what my
caveat "the usual graphical methods."

I do think that tools like -hangroot- help a lot, but most of the
usual graphical tools in statistics are built with the same
assumptions as linear regression, though they are not explicitly
stated. So for instance, the boxplot has marked trouble if you have a
bimodal distribution or a markedly non-normal one with too little
kurtosis.

Finding better tools for bounded response spaces would be a very
worthy research agenda.
*
*   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/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index