If the views expressed by leading edtech companies are any indication, the race to become the first 'Netflix of Education' is gathering steam. Comments from both established learning companies like McGraw-Hill, as well as more recent entrants like D2L and Udemy, reveal a strong push among edtech companies seeking to position themselves as "education's answer to Netflix."
One feature of Netflix in particular seems to have captured the imagination of edtech companies. The future of learning—we are repeatedly told—is found in personalized instruction, a term often interpreted as delivering educational content tailored to learners' preferences, interests, and schedules. This is a view of personalized learning that is strikingly similar to the approach Netflix has popularized with movies.
But despite the compelling nature of this "Netflix of Education" vision, it can encourage a misleading view of personalized learning. Examining several key areas where the educational Netflix analogy breaks down can help us better understand the qualities of an effective personalized learning solution. Failure to appreciate these points will result in personalized learning products more likely to hinder than help students.
Do Learners Really Know Best?
Netflix primarily uses movie viewers' preferences and engagement data to drive its movie personalization, but effective personalization of instruction cannot rely on learners in the same way. Unlike movie enjoyment, which is an inherently subjective evaluation, learning is an objective outcome. And unfortunately, evidence reveals students are often poor judges of what instructional techniques are most likely to promote optimal learning. And I'm being generous here.
In fact, decades of research reveal that students:
- are poor judges of the efficacy of their learning efforts
- prefer instructional formats that produce inferior learning outcomes
- make suboptimal decisions about when/where/how often to study
- overestimate how much they will remember or how well they will perform
- believe that things like learning styles and brain hemispheres influence learning
And I don't mean to pick only on learners. Instructors are equally susceptible to the deeply counterintuitive nature of learning. Thus, relying on commonsense intuitions about how we learn best is fraught with peril. While the romanticized view that each learner knows how he or she learns best is alluring, it is not supported by the evidence.
Because learning is a deeply counterintuitive process, where the ideal instructional decisions will often be ones that appear to make learning slower, more effortful, and intentionally difficult, we can't put learner preferences at the center of the learning process if successful learning is the goal.
Taking the Measurement of Learning Seriously
Learning isn't subjective like movie preferences, which leads to a second critical difference when applying the Netflix model of personalization to education. While Netflix's algorithms can rely on viewer ratings and behavior to determine how successful its recommendations are, effective educational personalization requires objective measures of student learning.
And the problem is that education companies often seem to be measuring everything but learning--e.g., satisfaction, engagement, grades, mindsets, attitudes, persistence, views, logins, clicks, etc. I don't disagree that these factors can influence and are often related to learning, but these are poor proxies in assessing learner competence.
Considering these issues from a medical perspective, how ought we evaluate a medical treatment that was promoted primarily on the basis of patient satisfaction, self-report surveys about whether the treatment worked, and patient improvement as measured immediately after taking a drug? I'm not saying measuring learning is easy, but it's everything in education.
Keeping the Person in Personalization
Let's get this out of the way: Netflix doesn't care about you as a person.
Netflix is not looking out for your best interests or trying to help you grow. It will not push you to go read a book, even if it is better than the movie. It will not recommend you watch a movie that it doesn't have in its collection, even if you might like it more. It doesn't encourage you to expand your movie watching habits by suggesting movies you may not like but are seminal in cinematic history. Netflix just wants you happy and watching more movies.
Supporting student learning, however, necessitates a very different approach. An effective personalized learning system, rather simply recommending content likely to keep students happy and engaged, will frequently need to do the opposite. Optimal instructional experiences will often be difficult, frustrating, and effortful.
Keeping learners motivated and confident when faced with these cognitive challenges means connecting with learners on a personal level. It requires understanding how to inspire students to persevere when discouraged or facing feelings of self-doubt; it requires trust and empathy. This is not something a content recommending algorithm can imitate.
Thus personalized learning can never be reduced to the recommendation of instructional content, no matter how tailored or 'correct' it is. A personalized learning system must enhance the emotional and personal connection between learners and teachers, rather than obviate it.
Beyond the Netflix of Education
From a technology and usability perspective, Netflix is deservingly worthy of imitation from edtech companies. But from a learning perspective, Netflix's approach to personalization hardly scratches the surface.
Personalized learning is about much more than recommending the next instructional activity or creating content that keeps learners engaged. It's about creating a learning environment that empowers each student to reach their full potential through learning.
This requires aiming to be much more than the "Netflix of Education."
Jay Lynch is Senior Academic Research Consultant for Course Design, Development, and Academic Research (CDDAR) at Pearson.
References
Bjork, E. L., & Bjork, R. A. (2011). Making things hard on yourself, but in a good way: Creating desirable difficulties to enhance learning. Psychology and the real world: Essays illustrating fundamental contributions to society, 56-64.
Bjork, R. a, Dunlosky, J., & Kornell, N. (2013). Self-regulated learning: beliefs, techniques, and illusions. Annual Review of Psychology, 64, 417–44. doi:10.1146/annurev-psych-113011-143823
Dunning, D. (2005). Self-insight: Roadblocks and detours on the path to knowing thyself. New York, NY: Psychology Press.
Karpicke, J. D., & Blunt, J. R. (2011). Retrieval Practice Produces More Learning than Elaborative Studying with Concept Mapping. Science, 331, 772–775. doi:10.1126/science.1199327
Karpicke, J. D., Butler, A. C., & Roediger, H. L. (2009). Metacognitive strategies in student learning: Do students practise retrieval when they study on their own? Memory, 17(4), 471–9. doi:10.1080/09658210802647009
Kirschner, P. A., & van Merriënboer, J. J. G. (2013). Do Learners Really Know Best? Urban Legends in Education. Educational Psychologist, 48, 169–183. doi:10.1080/00461520.2013.804395
Lalley, J. P., & Gentile, J. R. (2009). Adapting instruction to individuals: based on the evidence, What should it mean? International Journal of Teaching and Learning in Higher Education, 20(3), 462–475.
