The One World Schoolhouse: Education Reimagined

by Salman Khan

virtual teaching;

We are now still in the early stage of an inflection point that I believe is the most consequential in history: the Information Revolution. And in this revolution, the pace of change is so swift that deep creativity and analytical thinking are no longer optional; they are not luxuries but survival skills. We can no longer afford for only some part of the world’s population to be deeply educated.

I hoped to help students see the connections, the progression, between one lesson and the next; to hone their intuitions so that mere information, absorbed one concept at a time, could develop into true mastery of a subject. In a word, I wanted to restore the excitement—the active participation in learning and the natural high that went with it—that conventional curricula sometimes seemed to bludgeon into submission.

Can watching video lessons or using interactive software make people smart? No. But I would argue that it can do something even better: create a context in which people can give free rein to their curiosity and natural love of learning, so that they realize they’re already smart.

believed, and still believe, that teaching is a separate skill—in fact, an art that is creative, intuitive, and highly personal.

two of my first precepts were these: that lessons should be paced to the individual student’s needs, not to some arbitrary calendar; and that basic concepts needed to be deeply understood if students were to succeed at mastering more advanced ones.

Let me make it clear that I did not discover this fact. I stumbled upon it by a mix of intuition and serendipity. But the truth is that well-credentialed educational theorists had long before determined that ten to eighteen minutes was about the limit of students’ attention spans.

Art is the elimination of the unnecessary. —PABLO PICASSO

It was partly for this reason—and not because of some preexisting theory—that I decided that I would never be pictured in the lessons. I didn’t at the time own a suitable video camera, and I didn’t want to buy one. It seemed like a slippery slope. If I had a camera, I would have to worry about the lighting. If I had good lighting, I would have to give thought to what I was wearing and whether I had spinach in my teeth. The danger was that the whole process would become more like making movies than tutoring students. Tutoring is intimate. You talk with someone, not at someone. I wanted students to feel like they were sitting next to me at the kitchen table, elbow to elbow, working out problems together. I didn’t want to appear as a talking head at a blackboard, lecturing from across the room. So it was determined that students would never see me but only hear my voice, while the visuals would be nothing except my scrawls (and occasional historic images) on the black electronic chalkboard. Students would see the same thing I was seeing.

So if faces are so important to human beings, why exclude them from videos? Because they are a powerful distraction from the concepts being discussed. What, after all, is more distracting than a pair of blinking human eyes, a nose that twitches, and a mouth that moves with every word? Put a face in the same frame as an equation, and the eye will bounce back and forth between the two. Concentration will wander. Haven’t we all had the experience of losing the thread of a conversation because we homed in on the features of the person we were talking with rather than paying steady attention to what was being said?

some people fear that computer-based instruction is all about replacing teachers or lowering the level of skill needed to be a teacher. The exact opposite is true. Teachers become more important once students have the initial exposure to a concept online (either through videos or exercises). Teachers can then carve out face time with individual students who are struggling; they can move away from rote lecturing and into the higher tasks of mentoring, inspiring, and providing perspective. This suggests something that is at the very heart of my belief system: that when it comes to education, technology is not to be feared, but embraced; used wisely and sensitively, computer-based lessons actually allow teachers to do more teaching, and the classroom to become a workshop for mutual helping, rather than passive sitting.

At its most fundamental, mastery learning simply suggests that students should adequately comprehend a given concept before being expected to understand a more advanced one. While this seems straightforward and commonsensical enough, mastery learning has had a rocky and controversial history that is of interest for at least two reasons. First, it constitutes another instance of the education establishment failing to follow up on its own best research and soundest advice. Second, because of advances in technology, it is finally possible—nearly a century after the advantages of mastery learning were first described and tested—to broadly apply its methods and techniques to real schools and real students.

Second, mastery learning structured its curriculum not in terms of time, but in terms of certain target levels of comprehension and achievement. This turned tradition quietly but entirely upside down. In the traditional model, a certain amount of class time is devoted to a particular topic or concept; when the allotted interval is finished, the entire class moves on, in spite of the fact that individual students will have achieved widely varying degrees of mastery over the material. In Washburne’s system, by contrast, students, with the help of self-paced exercises, proceed at varying rates toward the same level of mastery. Those who learn more quickly can move ahead or do “enrichment exercises.” Those who learn more slowly are helped along by individual tutoring, or peer assistance, or additional homework.

Let me emphasize this difference, because it is central to everything I argue for in this book. In a traditional academic model, the time allotted to learn something is fixed while the comprehension of the concept is variable. Washburne was advocating the opposite. What should be fixed is a high level of comprehension and what should be variable is the amount of time students have to understand a concept.

Largely, though, what killed mastery learning, 1920s style, seems to have been simple inertia and resistance to new and threatening ideas. In a truly shocking 1989 study, it was concluded that between 1893 and 1979, “instructional practice [in public schools] remained about the same” (and it really hasn’t changed from 1979 to 2012 either)!3 To be sure, some very innovative groups of teachers and schools have been experimenting with new techniques within their classrooms, but the mainstream model did not change in any appreciable way. Did no one notice how much the world was changing, and how much the educational needs of students were evolving as well?

Mastery learning techniques were soon being applied in various pilot programs around the country. In study after study, mastery learning kicked butt when compared to conventional classroom models. One research paper concluded that “students in mastery learning programs at all levels showed increased gains in achievement over those in traditional instruction programs.… Students retained what they had learned longer under mastery learning, both in short-term and long-term studies.”5 Another study found that “mastery learning reduces the academic spread between the slower and faster students without slowing down the faster students.”6 Shifting the emphasis from students to teachers, yet another study recorded that “teachers who [used] mastery learning… began to feel better about teaching and their roles as teachers.”7

The other thing that has changed—and this is huge—is that technology has radically lowered the expenses formerly associated with mastery learning. No more paper workbooks. No more pricey printings of individualized exercises. Everything needed for self-paced learning is right there in the computer; the cost of delivering it to students is miniscule. The old excuse that newfangled teaching methods are just too expensive—or are only the province of elite schools in privileged communities—just no longer applies.

Taking responsibility for education is education; taking responsibility for learning is learning. From the student’s perspective, only by taking responsibility does true learning become possible; studies of mastery learning dynamics make this clear.

In gradually developing my own approach to teaching, one of my central objectives was to reverse this balkanizing tendency. In my view, no subject is ever finished. No concept is sealed off from other concepts. Knowledge is continuous; ideas flow.

Given that learning involves physical changes in each of our individual brains, and given that knowledge consists not of some linear progression but rather the gradually deepening comprehension of a vast web of concepts and ideas, a surprising corollary presents itself: No two educations are the same. There is a very refreshing irony here. You can standardize curricula, but you can’t standardize learning. No two brains are the same; no two pathways through the infinitely subtle web of knowledge are the same. Even the most rigorous standardized tests demonstrate only an approximate grasp of a certain subset of ideas that each student understands in his or her own way. Personal responsibility for learning goes hand in hand with a recognition of the uniqueness of each learner.

The higher your structure is to be, the deeper must be its foundation. —SAINT AUGUSTINE

First, students should be encouraged, at every stage of the learning process, to adopt an active stance toward their education. They shouldn’t just take things in; they should figure things out. This is an extremely valuable habit to inculcate, since in the modern world of work no one tells you what formula to plug in; success lies in the ability to solve problems in novel and creative ways. Besides, if you think about it, asking kids to be active is nothing more than asking them to be their natural selves. Is it natural for kids to sit quietly for an hour, listening? No, it’s natural for kids to want to do something, to be busy with work or play, to interact. Students are not naturally passive. Perversely, they need to be taught to be passive; the passivity then becomes a habit that makes them more tractable, perhaps, but less alert, less engaged in what they’re doing.

Portability and self-pacing, then, are essential aids to active, self-motivated learning. For a student to truly take ownership of his education, however, there’s another resource that’s required: easy and ongoing access to the lessons that have come before. This is where Internet-based learning offers a huge advantage over textbooks and other conventional materials. The lessons never disappear. Figuratively speaking, the blackboard is never erased, the books are never thrown away or given back. Students are encouraged to review because they can be confident that they will find what they are looking for, right there in their own computers. Even better, if the software knows when the student last visited a topic, it can directly make a review happen.

Nurturing this sense of wonder should be education’s highest goal; failing to nurture it is the central tragedy of our current system.

The apprentice system was also the world’s first version of vocational school. It was a place to learn a trade—though in certain instances the trade in question could be extremely highbrow. Many associate the apprentice system with artisans like blacksmiths or carpenters, but it has also historically been the primary mode of education for future scholars and artists. In fact, even today’s doctoral programs are really apprenticeships where a junior researcher (the PhD candidate) learns by doing research under and alongside a professor. Medical residency programs are also really apprenticeships.

Careers in even very intellectual professions, like law and medicine, were primarily developed outside the universities, through apprenticeships (although a few degree programs did begin to emerge in the eighteenth and nineteenth centuries). A law degree didn’t become a mainstream credential in the United States until the late 1800s, when the completion of postgraduate instruction became a requirement for admission to the bar.2 The idea that a college degree is a prerequisite to any professional career is a quite new one, only about a hundred years old. The idea that college is needed for everyone in order to be productive members of society is only a few decades old.

Let me be clear as to why I raise this point. I’m not suggesting that people shouldn’t go to college. My contention, rather, is that universities and their career-seeking students have a deep-seated contradiction to resolve: On the one hand, our society now views a college education as a gateway to employment; on the other hand, academia has tended to maintain a bias against the vocational. Clearly, our universities are still wrestling with an ancient but false dichotomy between the abstract and the practical, between wisdom and skill. Why should it prove so difficult to design a school that would teach both skill and wisdom, or even better, wisdom through skill? That’s the challenge and the opportunity we face today.

Even more, it now became practical to print and distribute writings that weren’t holy books or great works of classical literature—it is no coincidence that the first modern newspaper emerged in Gutenberg’s Strasbourg roughly 150 years after the printing press.

The mass production of books changed all that—and this is an aspect of education history to which too little attention has been paid. No longer was the teacher the sole source of information and the ultimate authority on a subject. Now there was an expert behind the expert, sharing in the teacher’s prestige as the source of knowledge. The teacher ruled in the classroom but the textbook had standing in the world beyond. What if the teacher and the text disagreed? The legitimizing power of print seemed to give the last word to the book. On the other hand, textbooks empowered teachers to expose their students to the latest thinking from the broader world. They gave students the ability to study at their own pace and come to class ready to be engaged at a deeper level by a master teacher.

The challenge, however—the same challenge in the early days of textbooks as now in the wider world of Internet-based learning—was this: How can we most effectively deploy standardized learning tools without undermining the unique gifts of teachers?

Similarly, according to Gatto, our sacred notion of the “class period” was put in place “so that self-motivation to learn would be muted by ceaseless interruptions.” Heaven forbid that students might delve beyond the prescribed curriculum or have time to discuss possibly heterodox and dangerous ideas among themselves; the bell rang and they had no choice but to break off their conversation or their deeper inquiry and move on to the next episode of approved instruction. By design, order trumped curiosity; regimentation took precedence over personal initiative.

To address this issue, the National Education Association formed the “Committee of Ten” in 1892. This was a group of educators—primarily university presidents—led by Charles Eliot, the president of Harvard, whose mission was to determine what primary and secondary education should be like. It was these ten men who decided that everyone in the United States should—starting at age six and ending at age eighteen—have eight years of elementary education followed by four years of high school. They decided that English, math, and reading should be covered every year, while chemistry and physics should be introduced near the end of high school.

Today’s world needs a workforce of creative, curious, and self-directed lifelong learners who are capable of conceiving and implementing novel ideas. Unfortunately, this is the type of student that the Prussian model actively suppresses.

The answer is that our student has been a victim of Swiss cheese learning. Though it seems solid from the outside, her education is full of holes.

still less for moving beyond memorization to experience the concepts through open-ended, creative projects.

Tests say little or nothing about a student’s potential to learn a subject. At best, they offer a snapshot of where the student stands at a given moment in time. Since we have seen that students learn at widely varying rates, and that catching on faster does not necessarily imply understanding more deeply, how meaningful are these isolated snapshots?

So then, coming back to our original question—what do tests actually test?—it seems that the most that can be confidently said is this: Tests measure the approximate state of a student’s memory and perhaps understanding, in regard to a particular subset of subject matter at a given moment in time, it being understood that the measurement can vary considerably and randomly according to the particular questions being asked.

Were the new tests more reliable than the old tests? I have no idea. And that’s really the point. It’s awfully difficult to appraise the quality of tests except by way of the test results. Are they reasonably consistent? Do they more or less conform to what experts think they should be? What politicians want them to be? It’s all rather circular. Again, I don’t deny the importance of testing, and I’m certainly not suggesting we do away with it. What I’m urging, though, is a measure of skepticism and caution in how much weight we give to test results alone.

increasingly, all around the world, mind workers are what’s called for.

our current system of testing and grading tends to filter out the creative, different-thinking people who are most likely to make major contributions to a field.

Consider an analogy. Imagine if we assessed student dancers purely by their flexibility or their strength. If we judged student painters purely by their ability to mix colors perfectly or draw exactly what they see. If we appraised aspiring writers purely by their mastery of grammar or vocabulary. What would we actually be measuring? At best, we’d be measuring certain attributes and prerequisites that would be helpful or necessary for the practice of these respective crafts. Would the measurements say anything about an individual’s potential for true artistry? For greatness? No.

The danger of using assessments as reasons to filter out students, then, is that we may overlook or discourage those whose talents are of a different order—whose intelligence tends more to the oblique and the intuitive. At the very least, when we use testing to exclude, we run the risk of squelching creativity before it has a chance to develop.

As it was in suburban New Jersey, so it was in school districts around the country and the world. Some people argued for more homework, some for less. Various experimental programs were put in place. Some schools made homework “optional.” Some schools put aggregate limits on homework, which created a nightmarish chore for teachers who had to coordinate how much they assigned. Some school districts essentially played semantic games, now calling after-school assignments “goal work” rather than homework. Other schools banned homework on weekends or before vacation breaks; some took the interesting step of forbidding homework on the evening before major standardized exams, perhaps sending the message that it was okay for kids to be stressed and exhausted except when taking tests that might reflect on the performance of the school itself.

One large survey conducted by the University of Michigan concluded that the single strongest predictor of better achievement scores and fewer behavioral problems was not time spent on homework, but rather the frequency and duration of family meals.11 If

In the United States, for the school year 2008–2009 (the most recent year for which comparative numbers are available), the average cost per student for a single year of secondary public education was $10,499. To put this number in perspective, consider that it is larger than the entire per capita gross domestic product of Russia or Brazil. In New York, the state with the highest education costs, the figure was $18,126 per student, more than the per capita GDP of such wealthy nations as South Korea and Saudi Arabia.

At roughly $10,000 per student per year, the average American school is spending $250,000–$300,000 per classroom of twenty-five to thirty students. Where is that money going? Arguably, most of it should be going to teachers; but that isn’t how it works. Teachers’ salaries are a relatively small part of the expenditure. If we generously put a teacher’s salary and benefits at $100,000 per year—teachers in most of the country make far less—and the cost of maintaining a 1,000-square-foot classroom at $30,000 per year (a figure comparable to leasing high-end office space), we still have $120,000–$170,000 for each classroom to be spent on “other stuff.” This other stuff includes things like well-paid administrators, security guards, and well-manicured football fields—none of which have a direct role in students’ learning.

But isn’t the student-to-valuable-time-with-the-teacher ratio more important? I have sat in eight-person college seminars where I never had a truly meaningful interaction with the professor; I have been in thirty-person classrooms where the teacher took a few minutes to work with me and mentor me directly on a regular basis.

If we move away from the broadcast lecture, students can have more of the teacher’s one-to-one attention; good teachers will get to do more of what led them to teaching in the first place—helping kids learn.

To send a child to a top-tier day school costs around $40,000 a year (or roughly $400,000 to $800,000 per year for a classroom of ten to twenty students). Boarding schools may charge more than $60,000. For affluent families in our megacompetitive culture, the tuition is often just a down payment. When the school day is done, the private tutors take over, sometimes charging as much as $500 per hour; it is not unheard of for parents to spend six figures a year, in addition to tuition, on a child’s tutoring.12 The tutoring these days goes well beyond the standard SAT test prep, and is sometimes tailored to specific courses at specific private schools. In an otherwise terrible job market, high-end tutoring has become a boutique growth industry.

What will make this goal attainable is the enlightened use of technology. Let me stress ENLIGHTENED use. Clearly, I believe that technology-enhanced teaching and learning is our best chance for an affordable and equitable educational future. But the key question is how the technology is used. It’s not enough to put a bunch of computers and smartboards into classrooms. The idea is to integrate the technology into how we teach and learn; without meaningful and imaginative integration, technology in the classroom could turn out to be just one more very expensive gimmick.