A Framework for Evaluating Stopping Rules for Fixed-Form Formative Assessments: Balancing Efficiency and Reliability

Stopping rules for fixed-form tests with graduated item difficulty are intended to stop administration of a test at the point where students are sufficiently unlikely to provide a correct response following a pattern of incorrect responses. Although widely employed in fixed-form tests in education, little research has been done to empirically evaluate the stopping rules in these tests that often have important instructional and/or placement implications for students. In this manuscript, we propose and research a framework for evaluating stopping rules with respect to two important and sometimes conflicting criteria: (1) efficiency, and (2) reliability. Using this framework, we provide an example in which we apply three increasingly complex methods for evaluating efficiency and two methods for examining reliability.