By Zoran PopoviÄ, Enlearn Founder & Chief Scientist–
Technology enhanced learning has delivered successful āpockets of advancementā in schools, but there has been very little success at scale that has made a profound difference. So we need to ask ourselves what needs to change in order identify and replicate success on a national level? Is the data gathered from educational science helping us scale success? Ā And if not, what needs to be changed in our approach to actionable research that will finally move the needle for all students?
In order to fully answer the question of how to positively affect learning through technology enhanced innovations, we have to, as scientists, start by accepting the most fundamentally challenging and interesting problem — Ā analyzing student learning. Ā The key underlying condition is that learning is, in every case for every child everywhere in the world, 100% contextual, while our resulting āresearch-basedā recommendations and solutions are not.
In every school and in every classroom there are contextual changes every single day. Teachers and students come to school with an array of feelings, strongly held beliefs, intentions, skillsets, and agendas, each of which may intersect with one another positively or negatively. Even if we agree that students and teachers arrive at school inspired by the noblest of intentions, they bear both the burden and the gift of being highly variable within their own experiences.
Almost all research in educational science, however, not only ignores context, but isolates factors affecting learning outcomes outside of context intentionally. We take a particular thing that is going to be varied, isolate it, take a very small sample set, find some particular signal, analyze that data, then publish a paper on it. As a result, we have overly specific, strongly held beliefs about what works in practice, because it worked for a few teachers in a small set of schools. In many ways, this system may have done more disservice than benefit to the field of education because lots of time, money, and energy have been spent by well-intended educators trying to implement these research-based recommendations. Ā But the results fall short, often far short, of the expectations because of other contextual variables that were ignored in the research but which combine to profoundly affect outcomes in real classrooms. Ā
We need to invert the scientific approach weāve been taking in education and move towards a life sciences model if we want to bring our knowledge to scale soon and accurately. Ā Our educational ecosystem is not a system of consistently uniform students and teachers and curricula. Rather, it is a highly complex ecosystem of humans and all of the variables that accompany humans, not just in a static sense but in a continuously changing dynamic sense of learning for both students and teachers. The answer to me is to build a comprehensive look at all of the variables, how they mesh together, and discover which things end up being important. Ā In contrast to life science, we also do not want to just understand this ecosystem, but to accurately determine the just-in-time interventions that would lead to inspired learning by students and rapid on-the-fly professional development by teachers. Ā If we invert the science or discovery process, start in-vivo and at-scale, determine key factors, develop theories of variation and optimal intervention, we can change the efficacy of not just technology enhanced instruction, but educational systems as a whole.
My conclusion is this: Instead of doing small experiments that build a theory and then try to apply that theory to scale…we need to start with the general design principles and apply things at scale. We can then determine how different specializations can be made maximally effective, and analyze those different situational elements. Once we have this insight, we can generalize into a theory of how to adjust things for a specific context. This leads to a different kind of science for the educational field that is potentially much more effective. Rather than offering the kind of science that would say, āThe right thing to do with kids is THISā we would offer, āThe right thing to do for this kind of kid, with this kind of instructor, in this school situation, for this particular topic is THIS.ā Which is much closer to true and effective personalized learning.
_____________________________________________
Zoran PopoviÄ is a Professor of Computer Science and the Director of the Center for Game Science at the University of Washington, as well as founder of Enlearn. Trained as a computer scientist, his research focuses on creating engaging environments for learning and scientific discovery. He is one of the leaders in the field of large-scale citizen science. His laboratory created Foldit, a biochemistry game that produced four Nature publications, and showed for the first time that novices can be developed into world-class experts in a science domain through game-based scientific discovery. His laboratory has produced award-winning math learning games played by over five million learners worldwide. He is currently focusing on engaging methods that can rapidly develop mastery and expertise in arbitrary domains with particular focus on scientific breakthroughs and revolutionizing K-12 math education. Zoran is known for conducting large-scale educational campaigns including Algebra Challenges conducted in Washington, Minnesota, and Norway with almost 100,000 students. To maximize the impact of his research, he founded Enlearn to develop the first platform that adapts all aspects of the learning ecosystem by specializing to each curriculum, student, classroom and teacher in real-time, towards maximizing learning outcomes. His contributions to the field of interactive computer experiences have been recognized by a number of awards including the NSF CAREER Award, Alfred P. Sloan Fellowship and ACM SIGGRAPH Significant New Researcher Award.
