The performance of both clinical diagnostic and mass screening tests is commonly measured in terms of the two basic parameters, sensitivity and specificity. Selection of cut‐off points, balancing the sensitivity‐specificity trade‐off, and comparison with alternative tests are often a matter of arbitrary decisions and are frequently made without consideration of outcome. Some of the concepts used in the analysis of visual signal detection systems can be helpful in this aspect of screening test selection and evaluation. A simplified version of these concepts is developed and its application to clinical and public health optometry is illustrated.
J. Daubs (Fri,) studied this question.