On Determining Replication Probabilities: Comments on Posavac (2002)