The Palin-Revere-Wikipedia story is actually a nice “teachable moment” in thinking about openness in education. As any educator knows, one of the hardest parts of teaching is managing the fine line between freedom and control. Call it the Bueller Principle: you don’t want your students just galavanting around doing whatever they like, but you don’t want to be this guy either. Openness breeds creativity, student-directed learning, and exploration — and also distraction and chaos.

All educational content — be it produced by a publisher or a wiki — must have some control mechanism to ensure its quality. There’s no one “right” mechanism: textbook publishers use a well-established editorial process in which content is reviewed by subject-matter experts as well as copy editors and proofreaders, who work in exchange for money. Wikipedia’s control mechanism relies on the norms of the community and governance rules which are fundamentally enforceable by Wikipedia itself. Both work well for their intended purposes in creating educational resources.

The job gets harder, however, when you start talking about educational services — for example, automatically-graded homework problems like those on Aplia. Providing an educational service implies a transaction between content provider and student: I promise to grade you appropriately, and I’m willing to stand behind that promise. In an offline world, teachers’ services have a well-defined time period: I’m going to teach you in this class, this semester, and after this class is done our relationship is essentially over. By contrast, in an online world of persistent content, there’s no walking away from the service you’re offering. Transactions will continue to occur, and you need to stand behind your work: you have to fix content or technical bugs, grant credit as necessary, provide access to grades for those with the authority to see them, and keep those grades private from those without that authority. You can’t take five minutes, edit a page, and walk away: you’re undertaking a commitment to stand behind the service you provide.

Think about this famous scene from “Dead Poets’ Society” through this lens. It’s all about a “barbaric yawp,” which sounds like openness and self-expression, and to a certain extent that is what the scene’s about — but the physicality of the scene is all about how the teacher tightly controls the creative experience of the student:

Try doing that online! Seriously, try. Because while educational services are much, much harder to create than static content like videos, the transformative potential of weaving together technology, openness, and control is huge. It’s clear that the folks behind Wikipedia are serious about extending their toolset to education authoring tools, which is fantastic. Crowd-sourcing educational services will be a giant benefit to education — but doing it right will be as difficult as it is important.

There tends to be a great deal of focus on the graduates themselves in these speeches — but to those of us who are educators, we should view this as our own final exam. How have we done our job? How well have we equipped the students who have been under our care to deal with the economic, social, and spiritual challenges ahead of them? And more importantly, what could we have done better — what must we do better for the next cohort?

There is a natural inclination in teachers to excel by trying to replicate their own best educational experiences, or avoid mistakes they feel were made by their own teachers. For example, Sal Khan said recently that “I teach the way I wish I had been taught.” But as we age, and the distance between ourselves and our students increases, this benchmark becomes less and less useful. In my own teaching, I found that the pop culture references I laced my lectures with — to Monty Python, Star Wars, Cheers, Arrested Development — fell flatter and flatter as my students’ cultural experience grew more distant from my own. We would do well to remember the words of George Bernard Shaw: “Do not do unto others as you would that they should do unto you. Their tastes may not be the same.”

The cultural gulf between teachers and students is exacerbated by an increasingly widening technology gulf. The hyperconnected world children grow up in presents them with a new set of challenges and tools: my own children can’t imagine a world in which any piece of information isn’t accessible through a quick search on a device held in our pocket.

What do the educational tools of the future look like? At a fundamental level, they’ll help students with four types of information:

1. Non-time-sensitive information. There is no reason for a student to memorize the height of the Hoover Dam, or the year in which Buchanan was elected, or what the methods on an obscure Java object are. If you can get it from a quick Google search, don’t waste your own internal memory cache storing trivia. The tools students need for this kind of information are a facility with search engines and an ability to parse their results. (These are also key job skills!)
2. Time-sensitive information. While you could look up the Spanish phrase for “Is that train about to pull out of the station the one going to Madrid,” you probably want to have those words at the tip of your tongue. Foreign language vocabulary is just one of a slew of examples of declarative knowledge which is best internalized, rather than retrieved. The tools students need for this kind of information are enhanced flash cards like the new Spanish app published by Brainscape, which uses cognitive research to help acquire and retain this kind of information.
3. Procedural knowledge. There are any number of procedures which are taught in school, especially in the area of math. This is one area where students can “learn by doing” at their own pace: there is nothing gained by having all the students go through a procedure in class at a uniform pace which will inevitably turn out to be too slow for some and too fast for others. The tools students need to learn procedural skills will be self-paced learning videos and exercises from sources like Khan Academy, Guaranteach, Virtual Nerd, or iTunes U.
4. Analytical frameworks. As information becomes easier to retrieve, the need for students to have analytical frameworks within which to place this information becomes more important. In teaching introductory economics, I often told my students that what I was giving them was a lens through which they could make sense of the information stream bombarding them every day. (If you read something on Wikipedia — or for that matter, the New York Times — how do you choose whether to believe it or not?) The tools students need for this are traditional classroom discussions, paired with exercises that encourage critical thinking like those in Aplia.

The role of the teacher in this new world will be more akin to a guide than a lecturer. The existence of easily accessed information means that the teacher doesn’t need to waste precious time explaining facts or demonstrating procedures; technology can take care of that. The teacher can spend that time more effectively helping the students under his or her care make sense of the information stream that’s bombarding them every day.

In this brave new world, teachers, students, and parents will all need to work together to navigate their way through a changing educational and technological landscape. And that is going to require a whole different suite of tools. More on this soon…

Perhaps nothing was so perfect a metaphor for our current educational system as the Colossus, a 70-foot-tall statue in the middle of the Faire. This is my son Theo doing his part to drag the gigantic rocks suspended over his head around in a circle:

Most of the exhibits were much smaller, of course, and many exemplified the kinds of microeducational learning opportunities I like to promote. For example, here’s Theo learning to solder for the first time at the sparkfun.com booth:

This is the kind of intense concentration that shows true learning. Here’s a nine-year-old boy holding a 400°C soldering iron for the first time, going slowly at first, then more fluidly, until he ends up with a complete working simon game.

The Colossus and the soldering workshop represent, in my mind, the two pillars of education: creativity and standards. Too often standards-based curricula stifle creativity, as educators are forced to “teach to the test” and cannot follow their students down a divergent learning path. In one sense the Maker Faire stands as a vibrant refutation of such a closed-minded approach.

But at the same time, certain kinds of standards help to guide learning by constraining the student to a well-thought-out path. The simon game that Theo made was a simple kit: the game itself was included on a self-contained chip, so that all Theo had to do was follow a series of instructions to solder pieces through a board. In short, by constraining his tasks, the kit allowed Theo to learn a very specific skill in a short amount of time, and to create something much more sophisticated that he could have by making the game from scratch.

In a way, the discussion about creativity and standards comes down to the distinction between risk and danger — both present in droves at the Maker Faire. Stifling creativity usually arises out of a sense of risk aversion: meeting standards “ensures” that our children are receiving a “high-quality” education, but can prevent teachers from experimenting with new and innovative ways to interact with their classes. On the positive side, standards do help minimize danger, allowing even a child to use a soldering iron to great effect.

So I’m grateful for the creativity of artists like Zachary Coffin, who made the Colossus. But I’m also grateful for the engineering standards the steel manufacturers whose product, suspending those giant rocks above my son’s head, had to meet. And as I think about how to contribute to a better education for Theo and his peers, I leave the Maker Faire with a renewed passion for fostering creativity, and a renewed respect for the utility of standards in achieving that goal.

Fundamentally, the world of educational content is shifting from a market in which a few players (publishers) dominate the market to one in which many smaller firms create content. Let’s start with a visual and dive into it:

The left-hand picture shows the problem facing a publisher today: most of their costs are fixed. That is, the cost of producing an additional textbook, once it’s been written, edited, digitized, designed, etc. is quite low; but the cost of writing, editing, digitizing, and designing that textbook is what we economists like to call “ginormous.” (Technical term.) In fact, in a digital world, the marginal cost of delivering an additional e-book to a user is basically zero. Therefore the average total cost (ATC) of producing Q textbooks is just total fixed cost of writing the book in the first place, divided by Q. The profit that accrues to the textbook publishers is their total revenue (P*Q) minus their fixed costs (ATC*Q), which is shown in the green box on the left-hand graph.

In such a world, the high fixed costs of production represent a barrier to entry which prevents new publishers from entering the market. Furthermore, the fewer players there are, the more each can charge; hence what we’ve seen in the last few decades is a massive consolidation of players in the publishing industry. You can see this in the list of imprints the major textbook publishers have: Cengage Learning, for example, owns the imprints Brooks/Cole, Delmar, Heinle, Schirmer, South-Western, and Wadsworth, while Pearson owns Addison-Wesley, Allyn & Bacon, Benjamin Cummings, Longman, Merrill, and Prentice Hall. All these smaller imprints used to be their own companies; by consolidating them into larger entities, publishers can simultaneously reduce their fixed costs (like running computer servers or hiring editors) and reduce competition, allowing themselves to raise prices.

Now consider the image on the right, which represents the market for microeducational content: not big textbooks, but smaller things like educational apps, flash cards, even tutoring services. On the demand side of this market, people are willing to pay a lot less for an educational app than they are for a textbook, and they’re much more price-sensitive when downloading something on iTunes than when decided whether or not to buy the book their professor is requiring for their course. Therefore the demand curve is both lower (people aren’t willing to pay as much) and flatter (or technically, more “elastic”, which ISN’T THE SAME THING, ECON 101 STUDENTS but is good enough for a hand-waving graph like this one).

On the supply side of the market, the fact that fixed costs are so much lower means that there’s a market supply just like there is in the market for most competitive goods. As we would expect in a competitive market, we see a lot of competition and close substitutes: for example, Moms With Apps highlighted today two of the many, many ABC apps currently on the market for preschoolers. Most educational apps are available for under $10; if that price doubled, you can be sure that more app creators would enter the market to try to get share. So this looks much more like a traditional marketplace than the textbook market.

On the surface, these seem like two very different markets: one for polished college textbooks, the other for homemade ABC apps. Over the next few years, though, we’re going to see continued rises in both textbook prices and in the quality of microeducational content provided in a competitive marketplace. (Khan Academy is just the beginning.)

Those who profit during this transition won’t be the content providers: competition and free entry, especially from non-profits education providers like Khan Academy and OER, will keep individual micropublishers’ profits low or even negative. Instead, profits will accrue to those firms who can become market makers: those who find ways to harness the value created by microeducational content providers and make it easy for parents and teachers to unlock that value for the children under their care.

Them: So, you basically write homework questions.
Me: Yes and no. I think of what we write as assignable learning activities. Yes, a student does them at home, but the purpose is a little different from what people often think of as homework.
Them: Huh?
Me: What we try to do is to create auto-graded tasks for the student to do that, having done those tasks, they have a deeper understanding of some concept.
Them: Could I use them for a quiz?
Me: You could, but that’s not really the point. The point is that doing these tasks is a way for the student to learn the material, not a way to assess whether they have learned it.
Them: But it’s graded, right?
Me: Right, but grades serve two purposes. One purpose of a grade is to determine who in a class deserves an A, or a B, and so on. Another is to provide feedback to the student: “This is something you need more work on.” Quizzes are used for the first; at least the way I use them when I teach, Aplia problems are best suited for the second.

This last point often caused the most cognitive dissonance, and the conversation would usually end shortly after it got to this point. I think the reason for this is that there’s little distinction in many people’s minds between grades-as-feedback and grades-as-assessment. When instructional designers would look at Aplia, in fact, many would describe it not as a “graded assignment” but rather as “guided practice.” In short, the words practice and graded are considered opposites of one another, at least in this context.

Since leaving Aplia, I’ve realized that I myself was making an implicit assumption about the nature of computer-assisted learning. Specifically, I was assuming that in order to grade something, the student had to provide an answer, which is then graded as being correct or not. There might be multiple correct answers, but the basic workflow was always the same: prompt the student, record their interaction with the question, provide a handy green check or red X based on the correctness of that interaction.

What I’ve found in my explorations since leaving Aplia has been a wide range of activities which are aided by computers, but which don’t fit this workflow. One of the simplest and most brilliant is one used by Brainscape in its line of flashcard apps. Their workflow is this: show one side of a flash card, then the other, then ask the student how well they knew the information. The student grades themselves, and the card goes back into the deck:

Cards that the student have identified as points of weakness appear more often than cards where the student is confident, helping the student to improve their understanding where they themselves have identified their own weaknesses. In other words, technology is still playing an important role and, like the Aplia assignments, supporting a feedback mechanism; but the source of the feedback is the student, not the author of the question.

This relates to the concept of deliberate practice discussed yesterday on the Freakonomics blog. Deliberate practice occurs when you actively think about your own learning process as you seek to master something. This is something which is very, very difficult to do: it requires a level of intentional self-reflection that is outside the normal scope of our daily lives. What’s brilliant about Brainscape is the way in which it integrates the act of self-reflection — “How well did you know this?” — directly into a simple user interface.

Assessments, guided practice, deliberate practice — fundamentally, these amount to different types of content (from flash cards to quizzes to experiments to games) strung together by different workflows (automatic feedback, teacher-generated feedback, or student-generated feedback). The challenge lies in figuring out how to create the right mix of content and workflow for each student, in a way that leverages the incentive value of grades while not dampening the enthusiasm and curiosity of children. It’s a tough challenge, but certainly not insurmountable with the artificial intelligence technologies available to us today.

Here’s why. If you haven’t seen the following wonderful animation of a talk by Sir Ken Robinson, it’s well worth your time:

The video is a perfect rebuttal to both the Eggers/Calegari and Bausell arguments. The problem isn’t that we don’t pay teachers enough (though we don’t) or that we don’t have good enough measures of teacher quality (though we don’t). The problem is the assumption, implicit in both of those arguments, that our system of education is basically a factory; one side argues to pay the factory workers more, and the other side argues for better quality monitoring of factory workers.

In my opinion, we’ll be able to attract and retain better teachers and get higher-quality work out of them when we stop treating them like factory workers. In Robinson’s words:

If you’re interested in the model of learning, you don’t start from this production-line mentality. Essentially it’s about conformity, and increasingly it’s about that as you look at the growth of standardised testing and standardised curricula. And it’s about standardisation. I believe we need to go in the exact opposite direction.

This has particular significance for those of us who want to transform education through technology. There is a great segment of the market which is interested in the ability of technology to improve the existing production-line of schools: to make teachers more efficient at their jobs, or to monitor them more effectively, or to expand the number of students they can teach. In other words, they look at the current factory-like system and say, “This is terrible! These factories are crumbling! We need more modern factories!”

I don’t yet know the recipe for how to get out of this factory mentality. I do believe that social media have a critical role to play in any solution, and that creating educational software which rewards collaboration as well as individual success will be one key ingredient. Another key ingredient will be the capacity of technology to create a personalized experience for all students. There are other ingredients, some of which are around now and others which have yet to be created. What we need most, right now, are great test kitchens. And then we need to get cooking.

As an economist, I tend to gravitate towards the profit motive as second-to-none for allocating resources efficiently. However, the assumption underlying that conclusion is that value is reflected in people’s willingness to pay. In a market economy, if it costs me $10 to make a pair of sneakers and I can sell them for $20, then the market sends me the signal that I should be producing sneakers. Done and done.

But in education, people’s willingness to pay isn’t closely related to the actual value they receive. The main consumers of education — children — have no money of their own, so education is funded by parents or the government, either as an investment in children’s future human capital or out of a sense of social justice. Regardless, there’s no easy way to put a price tag on any individual bit of education (a music class, an economics lecture); therefore a market system to deliver educational services is going to be a very deeply flawed market indeed, and produce all kinds of crazy results. And as in the anecdote Khan shares, the profit motive can lead to private online colleges actually seeking out students who won’t receive a real return on their educational investment.

So there’s a strong argument to be made that educational institutions themselves should be non-profit: they should have a fiduciary responsibility to educational integrity, not their shareholders. But as Khan says later in the video, this clearly doesn’t extend to all the tools that are used in education. At an extreme, the school buses that drive kids to school are made by for-profit companies, as are the light bulbs and the computers in the classrooms. Nobody would want their kids to drive in buses, or learn on computers, that were developed by non-profits.

In other words, it makes sense for educational infrastructure to be supplied by for-profit organizations, even if educational institutions should be non-profit.

The question for education technology companies, then, is this: are we more like schools, or more like computers? Do we seek to deliver educational services ourselves, or do we seek to be a tool that educators can use to deliver those services? The names we choose are a good indication of which side of that divide we see ourselves on: “Khan Academy” sounds like a school (non-profit), while “Blackboard” sounds like infrastructure (for-profit).

Over the next few years, though, the line between educational services and educational infrastructure is going to get blurrier: Aplia and Grockit, for example, are at once educational infrastructure and educational content. In this new world, we’re going to need a new type of organization (possibly like a B Corporation) that blends the best of the non-profit and for-profit worlds, while avoiding the worst excesses of both. As Osman Rashid, founder of Chegg and Kno, puts it later on the same video: “You need the soul of a non-profit, and the mentality of a for-profit.”

The field of economics which is concerned with analyzing phenomena of this sort is called search theory. Rather than assuming that everyone in a marketplace has perfect information about all the opportunities available to them, search theorists argue that most people exist in an information-poor environment and only become aware of potential trading partners idiosyncratically.

The contrast between traditional microeconomic theory and search theory is reflected in differing approaches to education. A vast oversimplification of this complex subject might go something like this:

• One way of looking at education is that children are proceeding through a curriculum which adults have designed to introduce them to The Most Valuable Concepts They Should Know. This approach, like traditional microeconomics, presupposes that adults have perfect information about a checklist of Things To Learn (or “learning outcomes” in edu-speak), and the problem we’re trying to solve as educators is to get through that list in the most efficient way possible.
• Another way of looking at children is to see them exploring a vast and mysterious new world, driven by their own curiosity and constantly testing hypotheses about the way that world works. This approach, like search theory, postulates that the most important skills to teach our children revolve around how to thrive in a world of information scarcity: skills like learning a new programming language quickly, researching a topic for a presentation, or leading a scrum team. In this world, the problem we’re trying to solve as educators is to harness children’s creativity and drive, while keeping them focused and on track.

Both of these approaches have merit, and neither would be good for all kids, especially in their more extreme forms. What I know from my own experience, though, is that my son has learned more about the scientific method from going to the Exploratorium and watching Mythbusters than he has in school. And it makes me wonder what we in the education technology industry can do to harness, rather than suppress, the creative instincts of children.

This is the paradox of online content: the very same technology which increases access to content simultaneously decreases the incentives to produce high-quality content in the first place. Put another way: in a world where high-quality content is available for free online, how can you sustain a business model in which you invest the time, money, and effort it takes to produce high-quality content?

One way to solve this problem is a ”freemium” model in which consumers can choose between getting ad-supported content for free, or paying for ad-free content. Indeed, KQED, a National Public Radio station in San Francisco, now offers a paid pledge-free stream for those who don’t want to listen to pledge drives. However, the “price” in this economic model reflects not the value of the content itself, but the amount someone is willing to pay to not be annoyed by ads; therefore this model cannot achieve an efficient outcome in the economic sense.

Another solution is to give valuable content away for free, but charge for a credential or certificate confirming that users have received the value from the content. This is the business model of MIT, which makes course materials freely available online but charges upwards of $50,000 per year for tuition, room, and board if you want an MIT degree. In short, in this model the content itself is not scarce: MIT’s willingness to confer degrees is.

There are other solutions, and I’ll get into them in later posts. One thing that’s clear, though: if you’re in the business of producing and selling content, you can’t solve this problem just by raising prices on the smaller and smaller pool of people willing to pay for it. Sooner or later — and mostly likely sooner — this strategy just leaves the field open for new entrants who have found a way to provide the same value you do, but at much lower cost.

To my mind, Khan’s most impressive achievement has been to show that a single inspired person can create great content and share it with the world, and that the world can find it useful. But the model of any one person providing this kind of content at scale is at best difficult and at worst counterproductive. On the Khan Academy discussion boards, users are clamoring for videos and exercises on everything from English grammar to music to chemistry. There’s just no way for one person, or even a talented team of people, to create that range of content and maintain quality.

The answer to that conundrum lies in those same discussion boards, where a host of people have expressed their own desire to contribute by creating videos, or writing exercises, or even just sharing their insights from teaching. But any one of those people would have a difficult time reaching a wide audience on their own, and Khan Academy isn’t set up (at least yet!) to manage the contributions of their legions of fans.

So the question in my mind isn’t “How can Khan Academy scale?” Rather, it’s “What can we do to find and empower the next Sal Khan — or the next hundred, or thousand, Sal Khans?” To answer this, we’ll need to overcome three challenges:

• Technological challenge: How can we create a platform in which anyone can create and share microeducational content (videos, exercises, lesson plans, etc.)?
• Economic challenge: What incentives can we provide to encourage people to contribute content, and what mechanisms can we put in place to identify and reward high-quality content?
• Social challenge: How do we drive the adoption and usage of this content in education, either within current institutional framework or through institutional change?