A successful intervention to improve conceptual knowledge of medical students who rely on memorization of disclosed...

Background: The mere memorization of isolated facts without the claim of integrating them is detrimental to the desired learning outcomes in medical education. The current study was conducted within an environment where items from summative assessments are regularly disclosed by the university and consequently collected into an item bank that is shared between students. Within this environment, we aimed to quantify 1) to which extent students use disclosed items for their preparation for the summative exam, 2) a putative mismatch between (isolated factual) knowledge regarding distinct questions from the item bank and conceptual knowledge, and 3) to which extent this mismatch can be ameliorated by a project aiming to steer student learning away from the memorization of isolated facts toward the acquisition of conceptual knowledge. Methods: This steering project in the midst of the first semester consisted of the implementation of an oral exam based on selected learning objectives, preceded by two seminars. After their summative exam at the end of semester, 135 students performed a voluntary exam for study purposes. Here, authentic (i.e., presumably preknown) physiology questions taken from the item bank were used to assess students' ability to 1) recognize the correct answer in a multiple choice (MC) question, 2) recall the answer (short answer), or 3) display conceptual knowledge closely corresponding to the question presented in the other formats. Additionally, students received a questionnaire addressing their learning habits and attitudes. Results: The median reported percentage of learning time for the summative exam exclusively spent with this item bank was 80%. The results of the voluntary exam indicate that students frequently recognize and recall correct answers of included items without displaying knowledge of the underlying concept. Compared to recall of the correct answer, the probability of giving a correct answer regarding the corresponding basal physiologic concept was lower by 47 percentage points (p <0.001) for topics not included in the steering project. Regarding topics included in the steering project, this discrepancy was reduced to 25.5% (p <0.001). Conclusion: The results of this study demonstrate the influence of disclosed items on student learning and learning outcomes and suggest that a carefully implemented assessment is able to improve conceptual knowledge in physiology.