Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: xtpcse vs. xtpcse, noconstant


From   "carambas" <mcaramba@uni-bonn.de>
To   "Statalist" <statalist@hsphsun2.harvard.edu>
Subject   st: xtpcse vs. xtpcse, noconstant
Date   Sun, 8 Aug 2004 23:16:57 +0200

Dear All,

  I estimated an xtpcse model, and I observed that any of my runs without
noconstant option has very low R-squared (i.e. 0.08--and this is the same
result if I run it in xtreg which has no nocons option) but with noconstant
option, all R-squared become high (i.e. from 0.80).Does anyone have
explanation for that?


  In addition, I did an lrtest to test for heteroskedasticity using xtgls
(since xtpcse does not allow these tests) but there appeared a comment ---
log-likelihoods of null models cannot be compared, followed by

likelihood-ratio test                                  LR chi2(72)
  -115.82
(Assumption: hetero nested in .)                       Prob > chi2 =
1.0000


  Could anyone pls give me hint what's going on there? Thanks

Cris

*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index