Unlearning Is Hard Too

A recent article in the New York Times Upshot shed light on how hard it is to get physicians to stop performing clinical practices that have proven ineffective.  The situation that they describe where doctors continue to practice medicine in certain ways because of what they have previously learned (regardless of updated information) reminded me of my own experiences in education.  We go to school, a training, or hear from a colleague that a practice is effective, and we integrate it into our practice.  Years later updated research shows that either the practice is ineffective or not as effective as we previously believed, but we continue anyway.  As read the quote below I was struck by how much this reminded me of educational practice.   

“We need to take a more cautious approach to technology adoption and learn from mistakes of early adoption of health care technologies based on little or low-quality clinical evidence.  This way we can prevent the need to ‘break up’ with practice when the high-quality evidence shows that it is ineffective.”

In addition to taking a cautious approach to adopting new ideas, we need to integrate the practices of improvement science and/or iterative cycles of testing solutions.  In education we often scale ideas that appear to be solutions to our problems, but do not monitor the effects of new approaches (do they have the intended outcome – even on a small scale).  We are eager to integrate new technologies (e.g. cell phones, wearables, VR), new paradigms to learning (e.g. personalized learning, student-centered learning), and new designs (e.g. schools without grade-levels).  Unfortunately, we don’t have check-in and measure the type of friction these innovations are experiencing.  We miss opportunities to better understand why solutions are not working well in our context.  Instead, after years we either continue with approaches or we don’t, but it rarely has anything to do with evidence.  We need to take a different approach, where we more rapidly collect data in the field that tests our beliefs about how a solution will work and make changes in response to these data.  In education we need both large scale studies of big ideas, but also a culture of testing local solutions to local problems and customizing (or hacking) these solutions to increase outcomes for students.