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

Re: st: no log likelihood ratio test after nbreg??


From   Jhilbe@aol.com
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: no log likelihood ratio test after nbreg??
Date   Thu, 10 Mar 2005 19:04:18 EST

I always use the likelihood ratio test when  evaluating which predictors to 
include in a model. During preliminary analysis  I'll accept "highly" 
significant  p-values based Z statitics (or t), but  not for fine-tuning the final 
model. 

Joe Hilbe  


================================================
thanks. I was  indeed using pweight in nbreg regression.  Then it seems the
Wald test  will be appropriate.  But I checked the STATA document and it
says:  "after the manuals were printed, we revisited this subject and
discovered  that we did not, in fact, want to include this test (Wald).  The
likelihood-ratio test is the only test that one should consider in  these
limiting cases".  So do I simply use the wald test reported above  the
coefficient table?

Thanks again.


> At 03:40 PM  3/10/2005 -0500, wying1@umbc.edu wrote:
>>Hi,  all,
>>
>>Usually there is a log likelikhood ratio test  statistics reported after a
>>negative binominal regression (I ran  possion regression first, and then
>>used 'nbreg' regression), and  sometimes the Z score for alpha is reported
>>too.  But in my  case, I didn't see the ratio reported after the
>>regression, and there  is no z score for alpha.  Then I tried to use
>>'lrtest' after the  model, but it says
>>"lrtest not valid after robust
>>specify  force option to perform test anyway".
>>
>>So I put 'force' in  option, but it still didn't give me the result.
>>
>>Can  anybody tell me what's going on?
>
> Sounds like you used the robust  option.  There is some discussion of this
> on p. 28 of the Stata 8  Survey Data Manual. It says that "pseudo-maximum
> likelihood methods"  (which get used with robust standard errors) are not
> "true likelihoods"  and hence "standard LR tests are no longer valid".
>
> For more, see  the FAQ "Why should I not do a likelihood-ratio test after
> an
> ML  estimation (e.g., logit, probit) with clustering or pweights?"  at
>
>  http://www.stata.com/support/faqs/stat/lrtest.html
>
> Other related  FAQs are at
>
>  http://www.stata.com/support/faqs/stat/#robust
>
>
>
>  -------------------------------------------
> Richard Williams, Notre Dame  Dept of Sociology
> OFFICE: (574)631-6668, (574)631-6463
>  FAX:    (574)288-4373
> HOME:    (574)289-5227
> EMAIL:  Richard.A.Williams.5@ND.Edu
> WWW  (personal):    http://www.nd.edu/~rwilliam
> WWW  (department):    http://www.nd.edu/~soc
>
> *
>  *   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/
>


Ying Wang
Ph.D.  Candidate
Department of Public Policy
University of Maryland, Baltimore  County
1000 Hilltop Circle,
Baltimore, MD 21228
Tel(O):  410-455-6531
Tel(H): 410-747-6980
Fax: 410-455-1172
Email:  wying1@umbc.edu


*
*   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/  

*
*   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