As part of the ALiEM Faculty Incubator program, teams of 2-4 incubator participants authored a primer on a key education theory, linking the abstract to practical scenarios. For the third year, these posts are being serialized on our blog, as a joint collaboration with ALiEM. You can view the first e-book here – the second is nearing completion and will soon be released. You can view all the blog posts from series 1 and 2 here.
The ALiEM team loves hearing your feedback prior to publication. No comment is too big or too small and they will be used to refine each primer prior to the eBook publication. (note: the blog posts themselves will remain unchanged)
This is the ninth post of Volume 3. You can find the previous posts here: Bolman and Deal’s Four-Frame Model; Validity; Mayer’s Cognitive Theory of Multimedia Learning; The Kirkpatrick Model: Four Levels of Learning Evaluation; Curriculum Development; Programmatic Assessment; Realist Evaluation; and, Kotter’s Stages of Change.
Authors: Michael Barrie; Shawn Dowling (@shawnkdowling); Nicole Rocca
Editor: Jordan Spector
Main Authors or Originators: Benjamin Bloom; James Block; Robert Burns
Other important authors or works: Carroll 1973 model of school learning; F. Keller; C. Washburne
Part 1: The Hook
As the residency program director in the ED at ALiEM Medical Center (AMC), you notice wide variability amongst your resident-learners as to their performance on the annual in-training exam; even amongst learners in the same post-graduate year. This problem is further magnified when Phil, one of your highly regarded PGY3 residents, performs quite poorly during a practice oral exam – the exam preceptor identified large gaps in Phil’s understanding of basic cardiac physiology and in the use of vasopressors. In your discussions with Phil, it becomes clear that he has a poor grasp on a number of fundamental concepts within critical care management. Phil is aware of this and states “I’ve been meaning to sit down one day and study this stuff but residency is so busy”. It is not clear to you why your current curriculum is sufficient for many of your residents to demonstrate competency in the management of critically ill patients, but Phil, a smart and hard-working resident, lags behind. As you review your didactic curriculum, you note that your learners receive a basic cardiac physiology introductory lesson during intern orientation, and a few other lectures led by a guest speaker (a cardiologist) who focused heavily on advance topics and controversies in the literature. That foundation was sufficient for some learners to do well on the exam, but not for Phil. You wonder if there could be a better curriculum design to ensure that all of your learners achieve mastery on fundamental competencies prior to taking on more advanced topics and lessons.
Part 2: The Meat
The founding principle of Mastery Learning (ML) theory is that the majority of students can attain a high level of achievement if provided proper instruction and sufficient time. Benjamin Bloom, a founder and proponent of this theory, argued that when a cohort of students with a normal distribution of aptitude are provided equal instruction within an equal amount of time, a portion of students will not attain mastery of that subject. Within that model, achievement is directly correlated with individual learner-aptitude.1 However, if each student is provided with as much time as he or she requires within a lesson topic, then the majority of learners could be expected to achieve mastery.2 A central thesis in ML is that “mastery” must be pre-defined (e.g. the criteria required to earn an “A” grade). In addition, ML requires that teachers should provide formative assessments at the end of each learning unit, as well as targeted corrective feedback for those who do not attain the mastery level on the first attempt. Overall, the ML instructional theory espouses the following principles:
- Clear description and delineation of learning objectives within the curriculum.
- Division of the lesson plan into discrete learning units, with sequential provision of the content.
- Instruction of each discrete unit for mastery. As such, all students are taught material in a single unit with standard methods, and then examined for mastery of that unit. Additional instruction is provided to students who have not achieved mastery, until they meet the predefined standard.
- Student evaluation reflects mastery of the curriculum as a whole, (rather than achievement relative to classmates).2
Though ML theory gained popularity in the 1960s with Bloom’s work, its roots can be traced to the work of Carleton Washburne in the 1920’s. Washburne was an educator in Winnetka, Illinois and while there developed the ‘Winnetka Plan’.3 The Winnetka Plan was formulated in response to the extant elementary school grading system that expected all learners to progress identically. The Winnetka plan introduced a curriculum that taught the ‘common essential’ subjects (reading, writing, arithmetic) with individualized curricula for students of different aptitudes. Students progressed to new content only after demonstrating mastery of the level below. The locally implemented curriculum change saw some dissemination, but only transiently.
Some time thereafter, John. B Carroll pioneered his own conceptual model of learning and school development. Carroll was another thought leader who championed the idea that most students can attain a certain criterion level within a subject when given enough time.4 He defined learning as a function of the actual time spent learning relative to the time needed. Time spent learning is related to learner perseverance plus opportunity, whereas the time they required for learning is related to a learner’s innate aptitude, the quality of instruction and the learner’s comprehension of that instruction. Visually, the theory can be represented as:
From there, ML theory was coined based on the work of thought leaders in two distinct domains; Benjamin Bloom in education, and Fred Keller in the domain of psychology. Bloom believed that knowledge acquisition need not correlate directly with learner aptitude. He argued that learners should be provided individualized instruction for a duration of time specific to their needs, so the majority of students could achieve mastery. The ML model would specify that students can succeed within a curriculum regardless of aptitude (i.e. the correlation between aptitude and learning would approach zero).1
Contemporary with Bloom’s work, Fred S. Keller gained publicity as theoretician in the domain of psychology. Keller developed the Personalized System of Instruction (PSI), which built upon B.F. Skinner’s work in operant conditioning, modifying that construct for classroom application.5 PSI described an individualized, learner-paced approach that was not easily applied to the conventional classroom setting, where multiple students learn together within finite time frames. It was not until Keller and his colleagues revamped the PSI system to facilitate teaching to multiple learners simultaneously, and adapted it to include discrete learning units, that the theory gained traction in school curricula. The core elements of the PSI strategy were described by Keller and Sherman as follows:
- Defining Mastery
- Curriculum divided into teaching-learning unit
- Objectives for mastery defined for each unit
- Planning for Mastery
- Educational resources provided by educators within each unit
- Feedback/corrective materials developed for students that require remediation
- Teaching for Mastery
- Student-specific rate of progress through teaching-learning units
- Students take the mastery test when sufficiently prepared, and they move on to the next unit only after they pass the mastery test. If they do not demonstrate mastery, students use corrective materials to address learning gaps
- Grading for Mastery
- Criteria for mastery (i.e. performance on a final exam) is defined by educator policy, and not on performance relative to peers5
While the ML and PSI theories are grounded in different realms, they are similar in that they are both founded on the principle that almost all students can master what they are taught if they understand the learning objectives, are given enough time, appropriate instruction for their specific baseline level, and corrective methods targeting areas of difficulty.
Modern takes or advances
Due in part to advancements in digital technology, principles of ML have been incorporated into various educational programs in recent years. One example is the “flipped classroom” didactic paradigm.6 In this model, educational content is pre-defined and posted in a learning management system so that students may review this material ahead of a lesson, working at their own pace. The flipped model allows students to move quickly through familiar content, and spend more time on the material that they find challenging. This model permits learners to engage in higher order discussions, or work through advanced applications of the learning content when in combined lessons. The flipped model is a modern paradigm based on the mastery learning approach.
Another modern example of ML is ‘gamification’ in medical education.7 Gamification describes the use of competitive ‘games’ for learners to navigate as a means to learn new content. In most iterations of gaming, students may work through educational program and objectives at their own pace. The gamer can only advance to the next unit after achieving defined mastery at the current level.
Many recent educational initiatives set forth by the ACGME align with the ML theory of education. With the advent of competency based education, with the implementation of the ACGME milestones and medical undergraduate entrustable professional activities, national leaders in medical education are setting forth individual standards that all learners must achieve to qualify for graduation.8 The milestone project in particular is well suited to Mastery Learning model, as it has different levels that learners should move through sequentially during training, defining criteria that one learner must achieve to be promoted.
A competency based education curriculum can incorporate the principles of ML, but given the breadth and scope of materials covered in contemporary medical school curricula, a ML approach to the entirety of medical school teaching is logistically challenging and likely impractical. It may be more feasible to incorporate mastery learning into specific aspects of a medical curriculum, such as procedure teaching and/or simulation. Here, tasks can be clearly defined and assessed, with development of a timeline to allow learners to progress at their own rate.
Annotated Bibliography of Key Papers
William C. McGaghie, PhD. Mastery Learning: It Is Time for Medical Education to Join the 21st Century. Academic Medicine, Vol. 90, No. 11 / November 20159
This manuscript offers a nice summary of ML theory within medical education. The authors argue that the traditional medical education model developed by Sir William Osler (termed the natural method of teaching) is a passive educational process, predicated largely on a learner’s cumulative patient-care experience. This manuscript argues that the Osler model is inadequate, as it may leave some students with knowledge deficits. Mastery learning is an educational paradigm that promotes excellent performance from all learners, with minimal variation in measured outcomes amongst students. Unlike traditional educational paradigms where lesson time is fixed and outcomes are variable, in the ML model, the inverse is true. Requirements of a Mastery Learning model include – baseline testing, clear learning objectives, engagement in educational activities, an explicit standard for passing, formative testing, and sequential advancement of skills towards competency. This model is appropriate for both undergraduate and post-graduate medical education trainees, and it helps ensure that all students are meeting the milestones required.
David A. Cook, Ryan Brydges, Benjamin Zendejas, Stanley J. Hamstra, Rose Hatala. Mastery Learning for Health Professionals Using Technology-Enhanced Simulation: A Systematic Review and Meta-Analysis. Acad Med. 2013;88:1178-118610
Mastery Learning (ML) is the educational theory upon which Competency Based Medical Education (CBME) is founded. It is well established that CBME requires individualized instruction. And ML offers the ability to tailor education to learners’ needs. Given technological advances, simulation-based education has been proposed to facilitate the incorporation of ML into CBME curriculum. The quantitative outcomes of ML-based simulation education were reviewed in this study. The findings in several studies found that overall, ML grounded simulation had a large positive effect on improving clinical skills, and moderate effect on improved patient outcomes. The authors describe the significant increase in time spent by both teacher and learner as a drawback to ML-based models. The time varied based on the concept or skill taught with ML based simulation.
Guskey, T. R. (2010). “Lessons of Mastery Learning.” Educational Leadership 68(2): 52-57 11
This review summarizes key concepts of mastery learning, and explains how ML can be applied to other related instructional models and interventions. The author defines a number of terms pertinent to mastery learning. Educators must provide a diagnostic pre-assessment for learners, to define pre-knowledge and describe knowledge and skills requisite prior to the lesson. Mastery learning encourages high-quality, group-based initial instruction. This primary intervention should be able to adapt to the context, to relate to students’ interests, and be specific to the students’ needs. For those that have mastered the content, educators must also offer effective enrichment activities to provide challenging and rewarding learning experiences for students who demonstrate mastery ahead of their peers – this is to be material that does not broach upon the next curricular block. The author suggests that future studies will focus less on the value of mastery learning itself, and more on improving processes of learning, instructional materials, and the home learning environment.11
ML based teaching will often require a significant time investment from both teacher and learner.12,13 Be it for the development of other learning aids/materials to assist the slower learners, or the extra time required for slower learners, ML-based education takes time. And the time invested can be difficult to estimate in advance, as it will by nature, be specific and dependent to both the learner and the task. The logistics of delivering this type of course can also be challenging, as the students who pass a curricular block will need enrichment activities to work on while complete the remediation materials.
The methods intrinsic to Mastery Learning are well suited to procedural learning. Skills such as central line insertion or thoracostomy are amenable to such instruction as they can be considered in discrete, easily observable and correctable tasks.14 However, another criticism of ML theory is that the focus on individual units of teaching may fragment learning and limit comprehension of sophisticated concepts. Inui described this limitation with a ballet analogy; a dancer can perfect each of number of individual ballet positions successfully, but it does not guarantee integration of these basic positions into an artful performance of Swan Lake.15 The author writes “Fragmentary assessment of individual tasks in mastery learning could become a dead end in competency evaluation instead of serving as stepping stone to a more holistic assessment of key competencies in integrated processes of care.”15
Part 3: The Denouement
In part to address knowledge deficits amongst some of your otherwise competent residents (like Phil), you seek to create learning modules within a number of key topics in emergency medicine, each with clear objectives and a structured progression. You do so by developing an asynchronous flipped-classroom curriculum, where every learner may progress at his or her own pace – graduating to the next module only after he or she has met predefined standards in the prior module. Resident-learners are administered a pre-test to highlight the students that may need special attention. Faculty pre-record lectures and provide teaching notes on core content. These include core objectives for each curricular block. During conference, time is allocated for small group discussion of the block content. At the end of each curricular block, students are assessed using simulation and multiple choice questions to test core objectives. Residents that do not attain mastery are provided a remediation assignment. Only those learners who have mastered the content are allowed ‘enrichment activities’ such as research opportunities, advanced study and related quality improvement projects.
Initially there was resistance to this curricular overhaul from faculty and residents. However, the participants start to ‘see the light’ after experiencing the curriculum and realizing that it is student focused with the goal of bringing everyone up to pre-defined expectations.
Don’t miss the 10th and final post of the series, coming out Tuesday, July 16, 2019!
PLEASE ADD YOUR PEER REVIEW IN THE COMMENTS SECTION BELOW
1. Bloom BS. Recent developments in mastery learning 1. Educ Psychol. 1973;10(2):53-57.
2. Block JH, Burns RB. Mastery Learning STOR. Vol 4.; 1976. http://blogs.edweek.org/edweek/DigitalEducation/block_burns_1976.Mastery learning.pdf. Accessed December 17, 2018.
3. Le C, Wolfe RE, Steinberg A. I JOBS FOR THE FUTURE THE PAST AND THE PROMISE: TODAY’S COMPETENCY EDUCATION MOVEMENT ACKNOWLEDGEMENTS. https://www.luminafoundation.org/files/resources/the-past-the-promise.pdf. Accessed December 17, 2018.
4. Carroll JB. The Carroll Model: A 25-Year Retrospective and Prospective View. Vol 18. https://pdfs.semanticscholar.org/3e99/5718cd78dcd62ce3d06051e147750c0e65f0.pdf. Accessed December 17, 2018.
5. Keller FS (Fred S, Sherman JG (John G. PSI, the Keller Plan Handbook : Essays on a Personalized System of Instruction. W.A. Benjamin; 1974.
6. Vogel L. Educators propose "flipping" medical training. Can Med Assoc J. 2012;184(12):E625-E626.
7. White EJ, Lewis JH, McCoy L. Gaming science innovations to integrate health systems science into medical education and practice. Adv Med Educ Pract. 2018;Volume 9:407-414.
8. Holmboe ES, Edgar L, Stan Hamstra C. The Milestones Guidebook. https://www.acgme.org/Portals/0/MilestonesGuidebook.pdf?ver=2016-05-31-113245-103. Accessed December 17, 2018.
9. McGaghie WC. Mastery Learning. Acad Med. 2015;90(11):1438-1441.
10. Cook DA, Brydges R, Zendejas B, Hamstra SJ, Hatala R. Mastery Learning for Health Professionals Using Technology-Enhanced Simulation. Acad Med. 2013;88(8):1178-1186.
11. Gaskey TR. Lessons of Mastery Learning – Educational Leadership. Educational Leadership. http://www.ascd.org/publications/educational-leadership/oct10/vol68/num02/Lessons-of-Mastery-Learning.aspx. Published 2010. Accessed December 17, 2018.
12. Anderson LW. An empirical investigation of individual differences in time to learn. J Educ Psychol. 1976;68(2):226-233
13. Arlin M, Webster J. Time costs of mastery learning. J Educ Psychol. 1983;75(2):187-195.
14. Barsuk JH, Cohen ER, Wayne DB, McGaghie WC, Yudkowsky R. A Comparison of Approaches for Mastery Learning Standard Setting. Acad Med. 2018;93(7):1079-1084.
15. Inui TS. The Charismatic Journey of Mastery Learning. Acad Med. 2015;90(11):1442-1444.