Retrieval Practice in Undergraduate Medical Education
by Cindy Nebel
(Cover photo by Bruno on Pixabay)
About eight months ago, I took a career change and started a new role as Director of Learning Services for a medical school. In the first few weeks on the job, I happily told the students about spacing, retrieval practice, concrete examples, and dual coding as often as they would listen. Their eyes glazed over, they smiled and nodded, thanked me for my time, but it was very clear to me that I wasn’t having much of an impact. Eight months later, I’ve learned so much about the tools, language, and strategies that medical students use and how those ideas can be adapted using effective strategies. In what follows, I hope to provide some ideas for any medical students reading this but to also discuss the lessons I’ve learned about the limitations of research and efficient (not just effective) learning.
Background
First, some prior knowledge for the non-medical folks reading. [Med students, you can skip this next paragraph.]
Image by yanalovephoto on Pixabay
Medical education is divided into Undergraduate Medical Education (UME), which includes coursework, supervised clinical rotations, and specialty electives, and Graduate Medical Education (GME) which is residency and like their entry level job. So, UME is graduate-level education (post-bachelors), but called Undergraduate, for reasons I’ve yet to discover. For the purposes of this blog, I’m focused on Phase 1 of UME – coursework. For most medical schools, preclinical coursework involves a series of short (4-8 weeks-ish) courses on foundational topics of anatomy and physiology and organ modules (e.g. cardiology, brain and behavior, hematology, etc.). UME students have an enormous amount of information to learn in each of these courses in an extremely short amount of time. Many medical schools have switched to a pass/fail curriculum such that the goal is not to master 100% of the content, but maybe 80% to stay safely above the passing threshold. At the end of their coursework, most students (this varies some by medical school) take their first Medical Licensing Examination (called Step 1). The exam is 8 hours long and tests their basic science knowledge through high working memory demand questions that ask them to apply their knowledge to a series of vignettes, usually requiring them to understand how information across systems are connected and problem-solving their way to an answer. Again, the goal is not 100% on this exam, but maybe 70% to be in very safe passing range.
The Problem
During a typical week of medical school, students have many required activities. These vary by school, but might include lectures, active learning sessions, simulations, etc. In addition, they have a long list of material that requires self-directed learning. There are simply not enough hours in the day for them to master the material. They need learning strategies that are not only effective, but efficient.
Effective but Inefficient
A lot of the strategies we discuss take time. It takes time (literally) to engage in quality spaced practice. It takes time to review and retrieve the same material multiple times. Creating effective visuals, organizers, writing quality questions for elaborative interrogation – these are all fantastic strategies for increasing understanding and retention (1), but it is impossible for students to engage in all of these strategies for the vast amount of new material presented to them daily.
Because of this challenge, students often turn to pre-created materials. These are some of the favorites among my students:
Anki (or Anking): Anki is a flashcard system that utilizes an algorithm to maximize retention using expanded spaced retrieval (2). When you get a card right, the space by which you will see it again increases. There are many existing card decks that are often even associated with other third-party materials, so students can unlock cards by topic, utilize cards from former students, and connect those cards with their current classes.
Pro: Anki is evidence-based. It is a retrieval practice tool and the algorithm used for spacing cards is based in the science of learning and can maximize retention of those cards.
Con: Anki is surface-level learning. It is a memorization tool that allows students to recall factual information at an item-specific level. That is helpful, but not sufficient. It can also take considerable time and cause considerable anxiety. One student I spoke with said that, on some days, it takes her 6-7 hours to get through all of her cards. There are some ways that you can adjust Anki to avoid that, but clearly novice users aren’t always aware or able to make those adjustments.
Boards and Beyond: This third-party program provides videos on common topics that are found in medical school curriculum and on Step 1. For my purposes, I’m going to put this as the example program, but the pros and cons below are true for most youtube videos (e.g. Dr. High Yield) and podcasts (e.g. Divine Intervention).
Pro: These videos tend to be more efficient than others. Only the most relevant information is included and the videos have high production value, reducing extraneous cognitive load (3).
Con: As they are videos, these involve what most medical students describe as “passive learning” (I don’t love this terminology). Without an effective note-taking strategy or other paired strategies, the videos themselves are not necessarily the most effective way to learn (4).
Amboss & UWorld: These third-party resources are essentially question banks. There are others, but these are the most frequently used by medical students. While each of them has other features, students are most often using them to answer questions that are very similar in style to the word-heavy, problem-solving questions that appear on Step 1.
Pro: There are a lot of pros here. These are retrieval practice tools and represent ideal transfer-appropriate processing (5) in that students are getting practice with the types of thinking that will be required on the actual big exams.
Con: Each of these has the option to be competed on tutor or timed mode. On tutor mode, students receive a question with immediate feedback. This feels like the most effective way to learn for most students, even though we know that delaying feedback (via timed mode) is more effective. Students can also choose to do a mixed block of questions (i.e. all topics included), a targeted block (e.g. just cardio questions), or they can create a custom bank (e.g. ischemic heart disease and arrhythmias only). Choosing the appropriate way to use these options requires metacognitive awareness and regulation, which may be challenging for students.
In addition to those above, I would love to talk about Sketchy and First Aid, but for the sake of brevity I’ll save those for a Part 2 perhaps.
Image by Clkr-Free-Vector-Images on Pixabay
Lessons Learned
While the research is somewhat limited in the practical application for medical students, the combination of my knowledge of cognitive research and my experience has led me to these broad recommendations for efficiency of learning:
1. Surface and Deep Learning
One issue that we have run into is that students have jumped straight to doing the Step 1-style questions that involve application and problem-solving OR they use Anki for all their retrieval needs. Both necessary, neither sufficient. In working with students, it is very clear that they need to know “the language” of medicine (e.g. what does ischemic mean anyway?) and to have certain facts very easily accessible, but isolated facts will only take you so far. Memorization is needed, but so is the understanding of how all of this material fits together. How do a cluster of symptoms translate into different possible diagnoses? How do those diagnoses affect each other? How do comorbidities impact treatment plans? These aren’t single facts that can be memorized, but understanding processes, systems, organization of knowledge are essential basic science skills for future doctors.
Here, we often advise students to start their medical education careers with a little more memorization and a little less application and to steadily shift such that their clinical training involves very little memorization and mostly application. In terms of third-party materials, this might mean going from an Anki-heavy study plan with some extra strategies for organizing material to a question-bank heavy study plan with a little Anki sprinkled in. But that’s just those two programs, and it is much more important to match the strategy to the material being learned and the learner’s prior knowledge. For example, if a student retained very little from their neuroscience unit, they probably need to do more surface-level review when they get to the brain and behavior course than their peers who retained more.
2. Passive and Active Learning
Oy, I hate the phrase “passive learning” – if it’s passive, you aren’t learning. *steps off soapbox* Students use these terms to mean something like “content review” and “retrieval practice” and what we have seen here is that both are needed to make learning as efficient as possible. Learning can happen from retrieval practice alone (6). Frankly, it can happen from a lecture alone. To be most efficient, we have seen students benefitting from the narrative style of content review, where they receive the interconnectedness of the material followed by active retrieval of that information. Regardless of the source of the “passive” and “active” learning, it’s important that for any given material, students are doing both.
Final Thoughts
The Lessons Learned section above contains what are essentially empirical questions. It has been exciting and enlightening to take the theoretical knowledge and see the direct application and success of strategies that I’ve so often talked about in other contexts. But there have also been some limitations on that theoretical knowledge that have made it all the more clear that researchers know much about the science of learning, but the application of that science in real-world contexts is an art not to be disregarded.
References:
(1) Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students’ learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public interest, 14(1), 4-58.
(2) Pumilia, C. A., Lessans, S., Harris, D., & Harris, D. M. (2020). An evidence-based guide for medical students: how to optimize the use of expanded-retrieval platforms. Cureus, 12(9).
(3) Jordan, J., Wagner, J., Manthey, D. E., Wolff, M., Santen, S., & Cico, S. J. (2020). Optimizing lectures from a cognitive load perspective. AEM education and training, 4(3), 306-312.
(4) Haidet, P., Morgan, R. O., O'malley, K., Moran, B. J., & Richards, B. F. (2004). A controlled trial of active versus passive learning strategies in a large group setting. Advances in health sciences education, 9, 15-27.
(5) Morris, C. D., Bransford, J. D., & Franks, J. J. (1977). Levels of processing versus transfer appropriate processing. Journal of verbal learning and verbal behavior, 16(5), 519-533.
(6) Butler, A. C., & Roediger, H. L. (2008). Feedback enhances the positive effects and reduces the negative effects of multiple-choice testing. Memory & cognition, 36(3), 604-616.
