|
|  |
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
RE: st: wish for logit output
The way to get the exponentiated constant coefficient with -logit- is to
define explicitly the X-variable of ones corresponding to the intercept,
and then to include this variable explicitly in the model, using the
-noconst- option. As in:
sysuse auto, clear
gene byte baseline=1
logit foreign baseline trunk, or noconst
This will display the baseline odds of non-US origin for cars with zero
trunk space, as well as the odds ratio of non-US origin per cubic foot
of trunk space. This trick is equally useful for defining confidence
intervals for geometric means and their ratios. See Newson (2003), which
can also be purchased together with 32 other useful Stata tips in
Newton and Cox (2006).
I hope this helps.
Best wishes
Roger
References
Newson R. Stata tip 1: The eform() option of regress. The Stata Journal
2003; 3(4): 445. Download post-publication update from
www.imperial.ac.uk/nhli/r.newson/
Newton HJ, Cox NJ. Thitry-three Stata tips. College Station, TX: Stata
Press: 2006.
Roger Newson
Lecturer in Medical Statistics
Respiratory Epidemiology and Public Health Group
National Heart and Lung Institute
Imperial College London
Royal Brompton campus
Room 33, Emmanuel Kaye Building
1B Manresa Road
London SW3 6LR
UNITED KINGDOM
Tel: +44 (0)20 7352 8121 ext 3381
Fax: +44 (0)20 7351 8322
Email: [email protected]
www.imperial.ac.uk/nhli/r.newson/
Opinions expressed are those of the author, not of the institution.
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Richard
Williams
Sent: 04 February 2007 17:43
To: [email protected]
Subject: Re: st: wish for logit output
At 11:51 AM 2/4/2007, Maarten buis wrote:
>Teaching logistic regression and the recent discussion with Diego
>Bassani made me remember a long standing wish for the output after
>-logit- with the or option (which I might forget again at the "wishes
>and grumbles" session at the German Stata Users' Group meeting on April
>2nd): I would like to see the exp(_cons). This parameter represents the
>odds of success versus failure for someone whose xs are all zero. This
>helps me to assess the size of an effect. For instance, the odds ratio
>may tell me that the odds of success is twice as large for women then
>for men, but it makes a difference if the odds of success for men is .5
>or .05.
I agree. SPSS provides this; and it gets around Maarten's labeling
concern by labeling things as Exp(B). I suppose there are concerns
about consistency both across routines and across versions of
Stata. Perhaps some little option could be added for those who want
to see the exponentiated constant.
-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
FAX: (574)288-4373
HOME: (574)289-5227
EMAIL: [email protected]
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/
