Trying to get through a backlog of journals and magazines this weekend, I came across this excellent synthesis of several big ideas that have garnered a lot of attention in the education technology field lately. Specifically, how do MOOCs, flipped classrooms, adaptive learning (Khan Academy), and brick-and-mortar schools fit together? This article, better than any other I've read, explains a lot about the potential for these technologies, and the role that traditional schools (specifically colleges and higher education) will always play.
Carr begins with a brief background on each of the big MOOC players. They've all been started by computer science folks (and artificial intelligence folks in particular). This is not by accident. First, the subject matter is concrete and easy to break into any-sized pieces. It lends itself well to the audience (anyone taking an online course at this stage is likely a bit tech-savvy). And, most importantly, at least a few of the founders are as interested in the data generated by tens of thousands of students as they are in providing a platform for free knowledge transmission to the global population.
Why do they care about data? Tim Fish has said for a decade now that he's a "data guy." Me too. The data Tim talks about is one kind, what I'll call small data. Often this data is individualized (e.g. we can put together a dashboard to show everything relevant to a particular student on one screen, their attendance, grades, sports teams, parents' phone numbers, and thus DASH was born). Sometimes it has been lightly aggregated (e.g. our professional growth software Folio shows teachers what the average and standard deviation are for "organized" in their classroom). Sometimes it has been heavily aggregated (e.g. Folio shows our headmaster that "knowledgeable" is the number one adjective used by students to describe McDonogh teachers). This is all powerful stuff, and information that would have been hard to synthesize 15 years ago, using paper. However, all of this data is about McDonogh: our students, our teachers, our school. We can use it to help guide individual decisions on a daily basis, but it doesn't apply outside of this school.
Big data is different. Big data is hugely aggregated. Google is a master of big data. What's the difference? Big data doesn't care who you are. It can be anonymous or not, but it doesn't matter. No human being sits in Google's datacenters and types out the results of your search queries. Instead, Google knows that 10,000 people searched for "best cough medicine" or "flu symptoms" in the Maryland area this week, so it's likely that there's a higher-than-average risk of a flu outbreak in the area. For traffic, Google has built a portable data collector and installed them in hundreds of millions of cars around the world, each one sending out a small anonymous ping every few minutes (you may know these spy devices as "smartphones"). When you allow them, Google uses those pings in the aggregate to determine what traffic conditions are; if 90% of the pings on the 695 northside inner loop haven't moved much in the last 15 minutes, there may be a backup, and the computer automatically updates the traffic report for that section of highway. Incidentally, Google doesn't do this out of the goodness of their hearts: when you search for "flu symptoms" or "traffic" they can show you advertisements for cough medicine or luxury cars that make the commute more bearable.
So, imagine you are running a class, Algebra. Your content is discrete and very concrete. 1+1=2, every time. You build yourself a tool to help tutor the students in your class, offering them small quizzes and extra videos. This tool allows you to collect data (individualized or aggregated, doesn't matter). Suddenly, you find that 80% of your students failed a quiz. You investigate, and realize that you could have more fully covered the difficult material on that section. That's small data. Imagine having that for every question, instead of every quiz. A majority of your students missed the questions about negative exponents. Slightly bigger data. Imagine you have five sections of Algebra, and the morning sections got the questions right, but the afternoon sections didn't. Time for more coffee at lunch, but maybe you also taught the material slightly differently in the morning. Bigger data. Add some more orders of magnitude. Suddenly, with 10,000 students, you have a laboratory at your fingertips. Students learning with one model (the morning method) answered the questions right. Students using another model (the afternoon method) didn't. You learn, and adjust your course to use the morning model, but now with two more modifications. Now imagine you run a not-for-profit free online university with 200 courses, 400 faculty, and two million students. You can collect all sorts of data about which types of students are most likely to succeed in which settings doing which kinds of work.
This is part of why so many people have poured so much money into MOOCs in the past year (at least $60 million) — they have the potential to help us learn so much about how we learn. This data has another use, however. Imagine if the big data is a bit more personalized. The MOOC platform discovers that I am a visual learner through my interactions on a number of quizzes and questions. It knows that visual learners have tended to learn the material more quickly when given the tutorial videos in a particular order. Or it knows that a kinesthetic learner needs reminders to take a break or to be assigned different kinds of problems. The software can adapt itself to each individual learner, becoming a more powerful tutor for every student. This is adaptive learning, and this is what Khan Academy is currently focused on. Taking lessons from thousands of students, the platform learns which small morsel of information each student needs next to ensure every student is as successful as possible. Big data is turned around and used to help each student, and the process has come full circle.
So are these MOOCs going to replace universities? Not likely. There is still a powerful experience for learners involved in realtime face-to-face communication (you may call this a "classroom discussion" or a "lecture"). Many subjects can't be broken into discrete any-sized bits (I imagine my ninth grade English teacher Wright Abbot, devotee of everything Faulkner, cringing after being asked to make five-minute videos about The Sound and the Fury). Brick-and-mortar schools (like McDonogh where I work) provide the other half of the equation; learning isn't just about information delivery, mastering the content. Ironically, this is where flipped learning can come in. Flipped allows students to spend time outside of the classroom absorbing content, parsing through material, listening to "boring" lectures. When they come to class, they engage in active discussion and participate in more active learning. English classes have been "flipped" for a century. Nobody expects the students to read the novel in class. Instead, they take in content at home, and then analyze and synthesize in school with a master teacher to help guide them through the process. This is the crux of Salman Khan's new book, that there is still a place for students to gather together face-to-face.
A potential niche for MOOCs in the long term is for lifelong learning, the notion that inquisitive and curious adults are constantly absorbing new knowledge, learning new skills, and challenging themselves. A teacher came to me recently wanting to learn how to program. He wanted to learn to build software and websites. I had a few resources for him (all online tutorials) but what I wish I had offered was a link to a Computer Science 101 MOOC, a free, online course for this teacher to spend time familiarizing himself with the principles behind the art of computer science, while providing him an opportunity to practice with graded quizzes and a community of similarly-interested learners. How many times in the past have you suggested to someone (or thought for yourself) to take a class at the local community college on ... computers, finance, art history, biology, oenology, etc. MOOCs don't need to replace universities — they don't need to compete for the same students.
I believe that all of these developments are incredibly exciting, and that we are just getting started with the potential for these education technologies. Equally important, however, I don't envision these technologies decreasing the importance of master teachers or traditional schools. Instead, I think in the future technology will help with the "heavy lifting," allowing every student to help customize his or her own learning, and allowing teachers better insight into how to reach each student in the most effective way.