Proctored Formative Assessment Improves Medical Student Performance on Future Summative Assessments

Although evidence that supports the utility of formative exams in undergraduate medical education is well-documented in literature, there is very limited data on potential impact of the setting in which formative assessments are administered. This study examines whether administering multiple choice question (MCQ)-based formative assessments in a proctored versus non-proctored setting influences student performance on eventual summative assessment. Throughout seven weeks of the first block course in Year 1 curriculum at the University of Illinois College of Medicine at Peoria (UICOMP), medical students from years 2017 to 2019 were assessed by weekly MCQ formative assessments in either a proctored or non-proctored setting. At the conclusion of the course, a 100-item MCQ cumulative, summative exam was administered to all Year 1 students. All MCQs on this final exam were linked to United States Medical Licensing Examination (USMLE) domains, instructional method, and levels of Bloom’s Taxonomy prior to summative exam administration. Student performance on the subset of summative exam MCQs which had been previously assessed in proctored formative assessments was compared to performance on the subset which had previously appeared on non-proctored formative assessments. The percentage correct (item difficulty) on final summative assessments was 0.89 ± 0.31 SD for items assessed in the proctored setting and 0.78 ± 0.41 SD for items assessed in the non-proctored setting. In addition, our analysis showed that regardless of the Bloom taxonomy levels, Team-based learning is an effective instructional approach to test for all Bloom’s levels. This study suggests that participation in formative assessments administered under a proctored setting leads to better outcomes in final summative exams, exemplified by an 11% increase in correct items on the final summative assessment.