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

RE: st: How to obtain intercepts using svyologit


From   "YP Choi" <choiyp@cuhk.edu.hk>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: How to obtain intercepts using svyologit
Date   Tue, 22 Feb 2005 14:16:39 +0800

Dear Richard, 

Thanks for your clarification. It is very helpful. 
Regards
Susanne 

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Richard Williams
Sent: Tuesday, February 22, 2005 8:39 AM
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: How to obtain intercepts using svyologit


At 03:05 PM 2/21/2005 +0800, YP Choi wrote:
>Dear all,
>
>I am using the survey commands in stata to analyze stratified data. I 
>am running some ordinal logistic regressing models and want to know how 
>I can obtain the odds ratio of the intercepts? Many thanks for your 
>help.

Somebody can correct me if I am wrong, but I don't believe the term "odds 
ratio" is appropriate when referring to the intercept (which is why Stata 
doesn't report it in the column labeled odds ratio; although other programs 
like SPSS will have a column labeled EXP(B) which includes the 
intercept).  The idea of the odds ratio is that you are contrasting the 
odds of 2 otherwise identical cases where one case scores 1 unit higher on 
the X variable in question.  For a constant, this doesn't make any sense, 
since all cases have a score of one on  the constant.

Nonetheless, in some contexts, you might be interested in the exponentiated 
intercept, e.g. in a logistic regression, an intercept of zero would tell 
you that somebody who scored 0 on all the Xs would have exp(0) = 1 odds of 
success, which corresponds to a 50% chance of success.  If the intercept 
was 1, the probability of success for somebody scoring 0 on all the Xs 
would be 73.1%.  If there are no Xs in the model, e.g. you just do -logit 
y- , then the exponentiated intercept gives you the overall odds of success.

In ordinal logistic regression, however, Stata reports cutpoints, not 
intercepts.  Again, maybe somebody can enlighten me here, but I am not sure 
what an exponentiated cutpoint would mean, at least when there is more than 
1 cutpoint. (If there is only 1 cutpoint, it has the same value but 
opposite sign of what you get when you run a logistic regression.)

But, if you really want exponentiated intercepts or cutpoints, there is 
always your calculator, or the -display- command in Stata.




-------------------------------------------
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/


*
*   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