Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Maarten Buis <maartenlbuis@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: Recovering cell values from odds ratios |
Date | Tue, 31 May 2011 15:08:12 +0200 |
On Tue, May 31, 2011 at 2:42 PM, King, Carina <c.king09@imperial.ac.uk> wrote: > The 'p' is the calculation from p=o/(1+o), and all the numbers were calculated, apart from the OR and the number of cases (218) and controls (14748). > > Taking 'unconsciousness' as the example: > > P= 3.54/(1+3.54) > P= 0.77973 The o in the formula refers to odds not the odds ratio. These are not the same thing. As the name suggests the odds ratio is a ratio of odds. To turn an odds ratio into odds you first need to know the baseline odds. An odds is the number of successes per failure An odds ratio is the ratio by which the odds in one group differs from the odds in another group. Say we have two groups, A and B. We know that there are a 100 persons in each group, the odds ratio comparing A with B is 2, and the baseline odds (the odds in group A) is .5. In that case we know the odds in group B is twice the odds in group A. The odds in group A is .5, so the odds in group B is 2*.5=1. So within group A we expect .5 successes per failure, and in group B we expect 1 success per failure. These are the odds. We now know that in group A the probability of success is .5/(1+.5)=1/3 and in group B the probability of success is 1/(1+1)=1/2. Now we know that there were (approximately) 33 successes and 67 failures in group A and 50 successes and 50 failures in group B. Notice that if we did not know the baseline odds (and unfortunately this is often not reported) we could not compute these numbers. -- Maarten -------------------------- Maarten L. Buis Institut fuer Soziologie Universitaet Tuebingen Wilhelmstrasse 36 72074 Tuebingen Germany http://www.maartenbuis.nl -------------------------- * * 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/