Question regarding -xttobit- and large panel data sets from a fresh Stata user and first-time poster:
I am currently estimating Random Effects tobit models using xttobit on a large data set. The data set has T=11 and n=40325, and is unbalanced so that N= 442318. The equation for estimation has 17 independent variables.
As pointed out in Stata FAQ's concerning xttobit:
"Quadrature is one of the most accepted approaches to fitting these models, but there are three cases where it often breaks down: (1) large panel sizes, (2) high within-panel correlation, and (3) variables that are constant or near constant within panel."
All three points ar valid for my data set. I have not yet been able to finalise any estimations involving the full panel as it generally ends up 'backing up' for iteration after iteration. I have played a bit around with different number of intpoints and briefly tested intmethod(ghermite), but this has not yielded any solution. Since I have little knowledge of technical matters related to algorithms and the maximization, I have not tried any maximize-options (like technique, gradient, ?tolerance etc.).
I know I am vague both in description of the problem and in my question and I am familiar with the FAQ advices with regards to very general questions, but here goes:
Can anyone on the list with experience with large panel data sets and (RE) tobit estimations advice me on estimation strategy? Or maybe point me to some of the general problems with large panels and xttobit?
Best regards,
Even Bergseng
*
* 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/
