To set a standard on the "beardedness" test (see TM 507 062) the probability that a student with a specific score will be judged as bearded must be estimated for each test score. To get an unbiased estimate of that probability, a representative sample of students at each test score level must be chosen. (BW)