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]
st: Cragg's Tobit
From
Andrea Rispoli <[email protected]>
To
[email protected]
Subject
st: Cragg's Tobit
Date
Wed, 4 Jul 2012 13:17:27 +0100
Dear Statalisters,
I am performing an analysis on panel data and my dependent variable is
continuous non negative, therefore I started by running a xttobit
model.
However I am now considering the possibility that the data generating
process is different when y=0 and when y>0. Therefore I was
considering using a craggit model where in the first stage I predict
the probability that y>0 and then, in the second stage, the actual
value of y.
In case you think that this is sensible, my question is how to reflect
the panel nature of the data in this case? I don't think craggit
allows it?
(The alternative I was thinking about is to use xtprobit to predict
y=0 vs y=1 ( if y>0), calculate the inv Mills Ratio and then use it as
a regressor in a xttobit. where I use all the values of y).
I am looking forward to receiving your valuable comments.
Thank you very much in advance!
Best Regards
AR
*
* 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/