Psychometricians call an exam's passing score its "cut score." The score draws a line (or "cut") between candidates who pass the exam and those who fail. An exam's cut score is an essential part of its validity: It tells us not only what knowledge and skills a candidate needs to demonstrate minimum competence, but how much of that knowledge and skills must be demonstrated on the assessment. Does a minimally competent candidate know the answer to every question? To just half of them? Or to just a few?
The cut score also accounts for measurement error. No assessment will ever measure a candidate's knowledge or skills exactly. When choosing a cut score, test makers choose between two types of measurement errors: do they want to guard against overestimating a candidate's ability or underestimating it?
The articles on this page offer additional information about cut scores, as well as critiques of the cut scores on the Uniform Bar Exam.
A majority of US jurisdictions administer the Uniform Bar Exam created by the National Conference of Bar Examiners (NCBE). The jurisdictions, however, each set their own cut score for this uniform exam. A candidate who fails the exam in one jurisdiction, might pass it in a different jurisdiction. As this article argues, "These MBE cut score disparities constitute bad logic because every state is attempting to use the same test to predict exactly the same thing: minimum competence to practice law. They are bad science because setting a cut score is a 'critical step' in assuring the validity of the use of the exam. MBE cut score disparities are also bad public policy, which explains why professions other than law have moved to uniform multiple choice test cut scores in their licensing tests."
Joan W. Howarth, The Case for a Uniform Cut Score, 42 J. Legal Prof. 69 (2017).