A Review of the Measurement Functions of Answer Choices in CSAT Mathematics Multiple-Choice Items and a Proposal for Expanding Short-Answer Items

This study aims to examine the function of answer choices in multiple-choice items used in the mathematics section of the College Scholastic Ability Test (CSAT) from a psychometric perspective, and to propose the expansion of short-answer items as an alternative to address their limitations. Despite criticisms of multiple-choice formats in public discourse, such items can serve as valid and effective assessment tools when well-constructed with appropriate stems and distractors. However, in the case of CSAT mathematics, most items remain valid even without the presence of answer choices, which calls for a reevaluation of the structural role of answer choices in terms of psychometric function. Drawing on concepts such as item guessing, item difficulty, and validity, this study compares the function of answer choices in CSAT mathematics items with those in other domains, thereby identifying their structural psychometric limitations. Furthermore, this study examines the recent development of assessment environments that enable handwriting recognition and automated scoring, and discusses how short-answer items in the CSAT mathematics can function as more valid assessment tools while maintaining efficiency and objectivity in scoring. Based on this discussion, this study proposes a policy to expand the use of short-answer items in the CSAT mathematics, with the aim of more fairly and validly assessing students' abilities required for university-level study. This policy is intended to ensure the stable inclusion of high-difficulty items within the curriculum and to lay the groundwork for the future introduction of constructed-response assessments.