Why not? This is unusual, but possible. The underlying key here is
that a probability density for a continuous variable is _not_ itself
a probability. It is thus not bounded by 0 and 1, and will have
units that are the reciprocal of those of your (unstated) response
variable. Follow that through, and likelihood can exceed 1;
and so log likelihood can exceed 0. So, on this evidence, there
is no problem to fix.
Nick
[email protected]
Richard Boylan
> I am using ivtobit for the first time and I am obtaining positive log
> likelihood. I did not think that this was possible. Any explanations
> and possible solutions? Thanks.
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
