Abstract This article describes a novel method to study whether Likelihood Ratio (LR) values proposed by a system are tenable given the validation data. The method is simple and easy to implement, and is based on counting how many times an LR is proposed above (or below) any value k, given both hypotheses in question. This gives the user of the system information on the range in which it is sensible to use the system. Visually, this is easily illustrated and software is provided to do so. An advantage of the method is that no utility functions are necessary to develop it nor ad hoc choices have to be made to test for sensitivity.
Alberink et al. (Tue,) studied this question.