Joe Mudge wrote a guest post for the blog here in March of this year.  The post dealt with our use of α=0.05 for hypothesis testing.  The following is another guest post by Joe Mudge in which he presents evidence of the benefit of using the optimal alpha level approach.  Please chime in with your thoughts.