Re: Subject: Re: st: Odds ratio

[Marcello Pagano <pagano@hsph.harvard.edu>]

True, except your transformation is linear--and since we usually round 
off to the closest degree, that makes
Fahrenheit approximately twice as precise as Centigrade  :-) --whereas 
the transform from odds to
probabilities is not.  The transformation is one-to-one and onto so the 
information on the two scales is
the same, but the feel is slightly different.

This all changes when you take the ratios.  The biggest difference with 
the ratios, of course, is that the
prior odds ratios are the posterior odds ratios, making case-control 
studies sound, something that cannot
be said of relative risks. (The relative odds of having lung cancer for 
smokers to non-smokers is the same
as the relative odds of being a smoker for those with lung cancer to 
those without lung cancer, for example.) 
Thus it is much more than a matter of taste as in the Fahrenheit versus 
Centigrade comparison; or pounds
versus Kilos, or miles versus ......

m.p.



Clyde Schechter wrote:
> Although not precisely analogous, I think of the odds ratio vs. risk ratio
> issue as being a bit like temperature measurements in different units. 
> It's mostly a matter of what you're accustomed to.  Just as it is easy to
> convert between odds and probabilities, it is also easy to convert
> temperatures between Farenheit and Celsius.  Yet, having grown up in the
> US, I have had, as far back as I can remember, an instinctive
> understanding of what a given Farenheit temperature feels like and its
> implications for getting dressed.  But only after extensive travels to
> Europe was I finally able to intuitively understand temperatures reported
> in Celsius.  And notwithstanding a physical chemistry course in college, I
> still have no intuitions about temperatures reported in Kelvins.
>
>
> Clyde Schechter, MA MD
> Associate Professor of Family & Social Medicine
> Albert Einstein College of Medicine, Bronx, NY, USA
>
>
>
> *
> *   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/