Re: st: sign of coefficients in a xtprobit

[Maarten Buis <maartenlbuis@gmail.com>]

One of the first things I do when I see a dataset is use -fre- (you
can install that by typing in Stata -ssc install fre-) on the
variables I want to use. It just gives a frequency distribution but
also shows the numerical values and the value labels. Those three bits
of information can help you detect a lot of problems. Next step would
be to just look at a cross tabulation of your dependent and
independent variable using -tab x y, coll-.

Hope this helps,
Maarten

On Sat, Feb 9, 2013 at 10:38 AM, Luca Marcolin <gonzalo07@hotmail.it> wrote:
> Hello,
> I am currently dealing with a xtprobit regression.
> I
> have experimented a problem with the coefficient of a dummy variable. IT
>  has not the sign that I expect, that is logical and obvious.
> Maybe
> there is a problem in my dataset, when I have put in the data, or is a
> specific - and curious - characteristic of my sample, but maybe is a
> "technical" problem.
> Can you suggest me some commands or procedures to understand if the variable is bad-built or there are some technical problems?
>
> Thanks
>
> Luke
> *
> *   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/



-- 
---------------------------------
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/