Learning Pits, Productive Failure and Plan-Do-Study-Act
In a New York Times article this past week, Jenny Anderson summarized recent approaches to more effective education for young(er) learners (“Learning the Right Way to Struggle”, The New York Times, April 5, 2022).
James Nottingham has helped teachers and students with the metaphor of the ‘learning pit’, to describe the challenge of learning something new and difficult. You may be at the bottom of a pit when you are learning but you can build your skill in climbing out, with help from teachers and friends and your own effort.
Manu Kapur has developed the concept of ‘productive failure’ in academic work, “a learning design that entails the design of conditions for learners to persist in generating and exploring representations and solution methods (RSMs) for solving complex, novel problems.”
The teams I work with in healthcare face complex, novel problems every day, so Anderson’s article got my attention.
Kapur offers several instructional design principles, starting with “Create problem-solving contexts that involve working on complex problems that challenge but not not frustrate…”.
That sounds right to me—there needs to be enough challenge for me and the teams I help to be worth our time and energy but not so much challenge that we guarantee failure, with concomitant frustration, stress and risk to patients and staff.
Self-calibration and PDSA ‘ramps’
In our current oral health collaborative, my faculty colleagues and I have convinced most collaborative participants that testing an idea or change to workflow on a small scale is practical and helpful. Small-scale often means a test of size one: test the change with one patient, one staff member. Try the idea during one encounter or one huddle. The test cycle is defined in four steps, Plan-Do-Study-Act (PDSA).
Using Kapur’s term, ‘Productive Failure’ is a good way to characterize initial PDSA cycles. In order to solve a problem in a complex and novel situation, a team needs to know the conditions under which the current version of their solution will fail. A team needs to invite failure, safely, in order to learn. See this post for a discussion of a simple prediction game that illustrates this principle.
We’re now starting to discuss the idea of ‘PDSA ramps.’ Ramps contain several linked test cycles. You often need a series of PDSA cycles so you need to know about ramps: it’s rare that a single test, with one patient, will teach you enough to be able to use the change successfully with the next 25 or 100 patients, over a number of weeks and months.
The slide image at the top of this post shows a sketched ramp with four PDSA cycles; the cycles move from an initial hunch to regular practice. The sketch also shows a wedge to keep the highest cycle in place: we want to prevent the wheel from rolling back down the ramp. The wedge corresponds to standardization, discussed here and here.
Consider one change we’ve asked teams to try in the dental clinic: identify a patient with diabetes on today’s schedule and determine if the patient is overdue for an A1c test. If the patient is overdue, assure the patient gets the A1c test done ‘right away’. Once teams have tested this change with one patient, what is the NEXT test?
I tell teams they can think about the next test in two ways.
Increase the size of the test. If they are confident in their idea and ability to carry it out, try the change for the next FIVE patients. If they are not so confident or stressed by other factors, try the change for the next TWO patients.
Change a key factor in the test system. It’s easy to think of one type of factor, the people factor. How will the idea work if they ask a new dentist and dental assistant to try the change? How will the idea work with a patient who has a tenuous connection to the health center? If you are stuck thinking about other factors that might matter, ask yourself time questions: would day of the week or time of day matter to the change? For example, what happens if the last patient of the day needs an A1c test? Will the change idea work at 5 pm?
Using Kapur’s advice, we need a ‘Goldilocks’ next test, just the right amount of problem-solving challenge to keep learning but not so much we guarantee failure or cause harm. We want the ramp to point upward but not be too steep. Teams can self-calibrate by choosing the pace and difficulty of the next test.
Personal note: It’s been more than four months since my previous blog post. While away from posting, I’ve been reflecting on how I learn as I attended to caring for members of my family. Like today’s post, the next few posts will touch on learning and knowing.
Nottingham and Kapur’s work emphasizes the social and emotional context for effective learning, something I sometimes forget in my enthusiasm for analytic insight.
