You are not currently logged in.
Access JSTOR through your library or other institution:
If You Use a Screen ReaderThis content is available through Read Online (Free) program, which relies on page scans. Since scans are not currently available to screen readers, please contact JSTOR User Support for access. We'll provide a PDF copy for your screen reader.
On the equivalence of prospective and retrospective likelihood methods in case-control studies
Vol. 97, No. 4 (DECEMBER 2010), pp. 990-996
Stable URL: http://www.jstor.org/stable/29777153
Page Count: 7
You can always find the topics here!Topics: Parametric models, Case control studies, Disease models, Inference, Statism, Modeling, Approximation, Statistics, Epidemiology, Bayesian analysis
Were these topics helpful?See somethings inaccurate? Let us know!
Select the topics that are inaccurate.
Since scans are not currently available to screen readers, please contact JSTOR User Support for access. We'll provide a PDF copy for your screen reader.
Preview not available
We present new approaches to analyzing case-control studies using prospective likelihood methods. In the classical framework, we extend the equality of the profile likelihoods to the Barndorff-Nielsen modified profile likelihoods for prospective and retrospective models. This enables simple and accurate approximate conditional inference for stratified case-control studies of moderate stratum size. In the Bayesian framework, we provide sufficient conditions on priors for the prospective model parameters to yield a prospective marginal posterior density equal to its retrospective counterpart. Our results extend the prospective-retrospective equivalence in the Bayesian paradigm with a more general class of priors than has previously been investigated.
Biometrika © 2010 Biometrika Trust