re:Re: st: heteroskedasticity test in panel data

["Jing Zhou" <jing.zhou@rmit.edu.au>]

Thanks Kit. I referred the command http://www.stata.com/support/faqs/stat/panel.html on FAQ to conduct the heteroskedasticity test in panel-data model. but the problem is just as you expected. I got the negative Chi-sq and p-value is 1. there must be something wrong. 

in my case, my panel dataset is large N, small T, so I am relatively sure there should exist heteroskedasticity in my models. but I want to do the heteroskedasticity test before chosing the proper models, could you please advise me how can I deal with this excepting for using the recommended commands on FAQs?

Many thanks!

Jing 

>>> Christopher Baum <kit.baum@bc.edu> 29/07/2010 4:43 am >>>
<>
Michael asked

   In that case, is the Stata FAQ (at http://www.stata.com/support/faqs/stat/panel.html)
 on which this code example is based also backward? 


I have not taken a careful look at that FAQ. It may be quite right and the user has not applied it correctly. In any case, if you compare two LLF values and come up with a negative Chi-square, something is wrong. Don't bother looking at the reported p-value in that case!

Kit

Kit Baum   |   Boston College Economics & DIW Berlin   |   http://ideas.repec.org/e/pba1.html 
                              An Introduction to Stata Programming  |   http://www.stata-press.com/books/isp.html 
   An Introduction to Modern Econometrics Using Stata  |   http://www.stata-press.com/books/imeus.html 


*
*   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/


*
*   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/