A review of practices for adjusting exam scores and a proposed nonlinear scaling method

Abstract

Advanced practices for summative exam development and post-exam analysis are proven to be effective but aren’t always practical, and, even when these are applied to some degree, exams remain inherently imperfect measures of student ability. Instructors may thus deem it necessary to adjust overall exam scores to account for aspects of an exam that may have been ill-suited to some or all students, and often these adjustments are made in an ad hoc and/or uninformed manner. This paper reviews reasons and methods for adjusting exam scores and proposes a new method that was developed organically from observation, reflection, and literature consultation. The scaling method considers that some of the underlying reasons for adjusting exam scores may affect certain sets of students more than others and seeks to incorporate this proposition while also avoiding weaknesses of other methods.