An extended Minimum Description Length (MDL) penalty function for binomial proportions

This article extends the Minimum Description Length (MDL) penalty function B2(n) suggested by P.G. Bryant and O.I. Cordero-Braña [Model selection using the minimum description length principle, The American Statistician 54 (2000) 257-268] to cover an arbitrary range of possible parameter values, and provides approximations for large sample sizes. The results are illustrated using data from a recent clinical trial (The Women's Health Initiative).