Abstract: Current versions of finite mixture models with underlying Beta distributions have the problem that the data y have to be contained strictly between 0 and 1. We resolve that problem by using ...