Assessment for Learning with Ungraded and Graded Assessments

Med Sci Educ. 2022 Sep 14;32(5):1045-1054. doi: 10.1007/s40670-022-01612-y. eCollection 2022 Oct.

Abstract

Introduction: Assessment for learning has many benefits, but learners will still encounter high-stakes decisions about their performance throughout training. It is unknown if assessment for learning can be promoted with a combination model where scores from some assessments are factored into course grades and scores from other assessments are not used for course grading.

Methods: At the University of Utah School of Medicine, year 1-2 medical students (MS) completed multiple-choice question quiz assessments and final examinations in six systems-based science courses. Quiz and final examination performance counted toward course grades for MS2017-MS2018. Starting with the MS2020 cohort, quizzes no longer counted toward course grades. Quiz, final examination, and Step 1 scores were compared between ungraded quiz and graded quiz cohorts with independent samples t-tests. Student and faculty feedback was collected.

Results: Quiz performance was not different for the ungraded and graded cohorts (p = 0.173). Ungraded cohorts scored 4% higher on final examinations than graded cohorts (p ≤ 0.001, d = 0.88). Ungraded cohorts scored above the national average and 11 points higher on Step 1 compared to graded cohorts, who had scored below the national average (p ≤ 0.001, d = 0.64). During the study period, Step 1 scores increased by 2 points nationally. Student feedback was positive, and faculty felt it improved their relationship with students.

Discussion: The change to ungraded quizzes did not negatively affect final examination or Step 1 performance, suggesting a combination of ungraded and graded assessments can effectively promote assessment for learning.

Keywords: Assessment; Performance; Preclinical; Undergraduate medical education.