ABSTRACT

This article introduces a max, designated, and MPN measure of the rarity of an outcome for serial dilution tests. All three measures divide the likelihood of the specified outcome at its MPN by the likelihood of an outcome with the same design. The max measure divides by the maximum of the likelihood at their MPN of all outcomes with the same number of positive tubes. The designated measure divides by the likelihood of an outcome that maximizes over a single dilution at a time. The MPN measure divides by the largest likelihood of any outcome at the MPN of the specified outcome. These new measures appear to agree well with improb and are much easier to compute.