Underlying Assumptions

Public schooling in the United States has been disrupted because we are asking it to perform two new jobs: keep the United States competitive and eliminate poverty by enabling every child in every demographic to reach proficiency in all core academic subjects. The metrics used to measure the performance of our schools have changed, and schools must change to improve along these new dimensions or else society and the political system will hire new organizations to do those jobs.

Christensen believes that the only way that public schools can fulfill these two new jobs is if schooling is an intrinsically motivating experience. “Motivation can be extrinsic or intrinsic. Extrinsic motivation is that which comes from outside of the task. Intrinsic motivation is when the work itself stimulates and compels an individual to stay with the task because the task itself is inherently fun and enjoyable.” “We all know that becoming a great athlete or a great pianist requires an extraordinary amount of consistent work. The hours of time required to train the brain to fire the synapses in the correct ways and thus hone the necessary muscle memory and thinking required is no different from that needed to learn to read and process information or think through math and science problems.” “When there is high extrinsic motivation for someone to learn something, schools’ jobs are easier. They do not have to teach material in an intrinsically motivating way because simply offering the material is enough. Students will choose to master it because of the extrinsic pressure. When there is no extrinsic motivation, however, things become trickier. Schools need to create intrinsically engaging methods for learning.”

Christensen argues that U.S. schools perform poorly because our prosperity has reduced our extrinsic motivations to learn. “When Japan was emerging from the ashes of World War II, there was a clear extrinsic motivation that encouraged students to study subjects like science and engineering that would help lift them out of poverty and reward them with a generous wage. As the country and its families prospered, however, the external pressure diminished. Some people who are wired to enjoy science and engineering in the way schools traditionally teach it—and therefore are intrinsically motivated—or those who have other extrinsic motivation in play will study them. But many no longer need to endure studying subjects that are not fun for them. The same downward trend is now beginning in Singapore and Korea.”

So how do we make schooling and core academic subjects intrinsically engaging? By tailoring instruction to match how each student learns. “In summary, the current educational system—the way it trains teachers, the way it groups students, the way the curriculum is designed, and the way the school buildings are laid out—is designed for standardization. If the United States is serious about leaving no child behind, it cannot teach its students with standardized methods. We must find a way to move toward what, in this book, we call a ‘student-centric’ model. We use the word ‘toward’ intentionally here because this is not, at least immediately, a binary choice. A monolithic batch process with all of its interdependencies is at one end of the spectrum, and a student-centric model that is completely modular is at the other. How might schools start down this promising path? Computer-based learning, which is a step on the road toward student-centric technology, offers a way.”

Christensen makes a number of assumptions here. First, he assumes that all students will learn if they are intrinsically motivated. Second, he assumes that “allowing students to learn in ways that correspond with how their brains are wired to learn” creates intrinsic motivation. Third, he assumes that we already know how to tailor instruction for every student. Christensen writes:

“In the 1960s and 1970s, society began requiring schools to customize offerings for students deemed to have special needs. Students who qualify for these designations typically require individual approaches, codified in an individualized education plan (IEP). In another special case, educators place immigrant students from non-English-speaking families into custom-designed English language learner (ELL) programs. Customization is almost surely an important advantage for both these categories of students, but it is also terribly expensive.”

“In the one-room schools that characterized public education during most of the 1800s, teaching was customized by necessity, at least by pace and level. Because the room was filled with children of different ages and abilities, teachers spent most of their day going from student to student, giving personalized instruction and assignments, and following up in individually tailored ways. But as classrooms filled in the late 1800s, this method of teaching changed as larger enrollments forced schools to standardize.”

“The second phase of the disruption we term student-centric technology, in which software has been developed that can help students learn each subject in a manner that is consistent with their learning needs. Whereas computer-based learning is disruptive relative to the monolithic mode of teacher-led instruction, student-centric technology is disruptive relative to personal tutors. Tutors today are largely limited to the wealthy; and for those privileged few, good tutors come as close as possible to helping students learn each subject in ways that match the way their brains are wired to learn. Like all disruptions, student-centric technology will make it affordable, convenient, and simple for many more students to learn in ways that are customized for them.”

Christensen assumes that student-centric models already exist in special education and personal tutoring, but that we don’t yet have the technology to bring these models to scale. That is where computers—and student-centric technologies—come into play.

Christensen acknowledges that, at the moment, computer-based learning is not competitive with teacher-led learning. “If the history of these types of innovations can serve as a guide, the disruptive transition from teacher-led to software-delivered instruction is likely to proceed in two stages. We call the first of these stages computer-based, or online, learning. In this stage, the software will be proprietary and relatively expensive to develop. It will also be relatively monolithic with respect to students’ preferred methods of learning in that the instructional methods in this software will largely mirror the dominant type of learning method in each subject.” “Currently, according to reports, online learning works best with more motivated students; over time, it will become more engaging so as to reach different types of learners.” Right now, students who lack motivation to learn are even less motivated by online learning. Christensen assumes computer-based learning will get better as it is adopted and refined in markets where teacher-led learning is unavailable. These markets target nonconsumers, people whose only option is to use either computer-based learning or nothing at all.

Christensen expects computer-based learning to improve along a specific trajectory. Experts in student-centric instruction will use the technology to take their models to scale. As the technology matures and this instructional expertise is codified, instruction becomes modular. “The level of interdependence found in a product is a function of the underlying technology’s maturity. In the early days of most new products and services, the components need to be tightly woven together to maximize the functionality from an immature technology that is not yet good enough to satisfy customer needs. As products and their markets mature, technology grows more sophisticated, as do customers. They begin to understand their unique needs and to insist on customized products. Technological maturity makes customization possible. Product and service architectures become more modular in this environment.” Modularity enables a product to improve more rapidly and inexpensively because the entire product does not need to be redesigned every time. If instructional expertise then becomes commoditized, then user-generated instruction can be delivered via facilitated networks and the existing value network can be disrupted. “In this second stage of disruption, the existing value chain, which we call a ‘value network,’ is almost always disrupted as well. It is rare for a disruption to appear in just one part of a value network without the rest of the system changing, too. It is the disruption of the full value network that ultimately enables these modular solutions to emerge. Embedding a disruptive product in an entirely disruptive value network is key to achieving a less expensive solution than was possible in the first stage of disruption.”

Unless computer-based learning competes on a different dimension (not just in a different market), it will not be disruptive. If computer-based learning simply does the job that schools do now, only better, it will be adopted by schools as a sustaining innovation and nothing else will change. To become a disruptive innovation, computer-based learning must evolve into a student-centric technology, enabling every student in every demographic to reach proficiency in all core academic subjects. But that can’t happen unless the instructional expertise to create student-centric learning models exists to be codified and commoditized. Christensen assumes it does exist, but I don’t. I don’t think we can afford to make that assumption without some evidence.

A Square Peg in a Round Hole

When analyzing schooling, Christensen observes that the public education system in the United States has been disrupted three times. “But because the United States has been unwilling or unable to facilitate the entrance of new organizations with new business models to disrupt the old, public school districts have had to negotiate this disruptive redefinitions of performance entirely within their existing schools. In our studies of disruptive innovation in the private sector, we are not aware of a single instance in which a for-profit company was able to implement successfully the disruptive innovation within its core business. The few that survived disruption did so by creating, under the corporate umbrella, a new business unit, with a new business model attuned to the disruptive value proposition. Asking the public schools to negotiate these disruptions from within their mainstream organizations is tantamount to giving them a demonstrably impossible task.” In the first two cases, public schools were able to manage the disruption from within their existing organizations. They are grappling with the third case now.

Disruption occurs when performance is redefined. A sustaining innovation improves a product. A disruptive innovation changes what improvement means. Originally, public schools in the United States were established to “preserve democracy and inculcate democratic values” by teaching the basics, instilling sound morals, and preparing “an elite group—selected on merit from this entire pool of students, not just those from the upper class—to lead the country wisely in elected office.” One-room schoolhouses appeared to fulfill this job.

“In the 1890s and early 1900s, competition with a fast-rising industrial Germany constituted a minicrisis; Americans responded in the early twentieth century by handing schools a new job: prepare everyone for vocations. The goal was to produce a sound workforce for jobs ranging from administrative functions to technically demanding manufacturing positions so that America could compete with Germany. The old job of preparing the next generation to lead and participate in democracy did not go away; society simply asked schools to perform both jobs.” One-room schoolhouses had been improving through sustaining innovations, but suddenly, those improvements didn’t matter any more. The direction in which schools needed to improve had changed, and the comprehensive high school appeared to fulfill this new job.

Comprehensive high schools steadily improved, fulfilling their job more and more effectively, until the public education system was disrupted again. “The nation asked its schools to take on the new job of keeping the United States competitive. Although seemingly similar to the previous job, it was actually quite different. No longer could students choose most of their classes or focus on the vocational or general or academic track depending on their interests or talents. Virtually everyone had to focus on the core academic classes and take the same tests. Japan’s disruption of America’s manufacturing industries increased the pressure for all students to attend college, which further ratcheted up the need to focus on the core subjects and tests because postsecondary schools increasingly required them. This was a radically different demand of schools.” Suddenly, schools had to measure themselves against standardized tests, some administered globally. Instead of continuing to expand course offerings, courses were pared back to focus on core academic subjects.

Once again, the public education system responded. It didn’t become an industry leader, but it did re-organize itself and begin improving along these new dimensions. Then disruption struck for the third time. “The No Child Left Behind Act not only federally cemented average test scores as the primary metric for performance improvement, but also arguably once again shifted the goalposts. No longer can public schools simply raise the average test scores in their schools; instead, public schools must see to it that every child in every demographic improves his or her test scores. Now the performance measure for schools is the percentage of students who are proficient in core subjects. The essential motivation for asking schools to make sure that all students are proficient in reading, math, and science is to eliminate poverty.”

Incumbents find it next to impossible to manage disruption from within their existing organizations because they are structured to satisfy existing customers. Imagine that someone develops a car that can hover. For most existing customers, these hover cars are vastly inferior to the cars they can already buy on the dimensions that matter to them. There are a few new customers that prefer the hover car for various niche uses, but their numbers are small. The car companies can’t invest in hover cars because their existing customers don’t want them and the market for new customers is too small to contribute to their growth. By the time the hover car has improved enough to either satisfy existing customers or attract enough new customers to make the market valuable, it is too late.

This is how disruption normally proceeds, but it isn’t how the current disruption is proceeding in the U.S. public education system. The customer of public schooling is society and the political system. Society hires public schools to do a job, such as keeping the United States competitive or eliminating poverty. In order to do that job, public schools need students to hire them also, but the primary customer is society. In the three disruptions described by Christensen, it’s been the existing customer that has redefined the job to be done. Public schools aren’t facing the dilemma of having to choose between satisfying existing customers or pursuing new customers. Sure, some stakeholders are unhappy with the new jobs that have been given to schools; some stakeholders wish that we could return to the days where the goal was to increase participation by increasing course offerings. But the most lucrative customer, the one that controls the purse strings, has spoken. This may explain why public schools have been able to manage these disruptions so successfully in the past where for-profit companies have failed.

The other significant difference I see is the source of the disruption. Typically, there is an innovative technology. If this innovative technology causes performance to be redefined and new business models and value networks to appear, then the innovation is disruptive. If none of that happens, then the innovation is sustaining (or not innovative at all). But note that the technology precedes the disruption. There is no disruption without a disruptive technology.

This was the case when public schooling in the United States was disrupted the first two times. Comprehensive high schools appeared because one-room schoolhouses could not compete with the schooling in Germany. Christensen doesn’t say what it was, but some innovative schooling technology had been developed in the German market, and it was now disrupting the U.S. market. The public education system had to respond or give way to new organizations with new models. The same thing happened when comprehensive high schools were replaced by a focus on core academics and standardized testing. The public education system was disrupted by innovative schooling technologies from other countries, such as Japan. But, as far as I can tell, the third disruption, the one we are currently in, was not triggered by any technology. Is there a schooling model anywhere in the world that has eliminated poverty by enabling every child to reach proficiency in all core subjects? This is as though existing car customers suddenly decided en masse that they wanted hover cars before anyone even knows if it is possible to build a hover car. Schools are now scrambling to invent technologies to respond to this disruption, but there is no guarantee that they will be successful.

I want to make one final observation on the history of disruption in public schooling in the United States. None of the innovative technologies in the first two disruptions were instructional in nature. Instruction has not changed since the days of the one-room schoolhouse. Sure, we almost certainly lecture more often today given the batch processing model of schooling we have now, but the lecture existed in one-room schoolhouses. If you were to observe a teacher from the 1780s working with a small group of students and a teacher from the 2010s working with a small group of students, the instruction would be very similar. The innovative technologies that caused the first two disruptions affected systems surrounding the classroom, but they did not change the instruction that occurs inside of the classroom. In other words, the instructional component of schooling was used to achieve a new purpose, but the component itself did not improve. Schooling was disrupted, but instruction was not.

Christensen believes, and I agree with him, that public schools in the United States will need new instructional technologies in order to fulfill both the new job of eliminating poverty and the old job of keeping the United States competitive. Christensen assumes that these new technologies will appear simply because schools have been disrupted and we need them. I am less optimistic.

Applying Disruption Theory to Public Schooling

Clayton Christensen is the author of The Innovator’s Dilemma and an expert in disruptive innovation. In 2008, he applied disruption theory to analyze public schooling in the United States.

When I first read Disrupting Class, I felt that Christensen made a lot of bold claims about schooling and education without backing them up. But after reading his book for a second time, I realized that Christensen isn’t trying to build a descriptive understanding of schooling from data, but a predictive understanding of disruption by applying disruption theory in a new context.

According to Christensen, “researchers build bodies of understanding in two major stages—the descriptive stage and the prescriptive stage.” In the descriptive stage, researchers “generally follow three steps—observation, categorization, and association—as they do their work.” Christensen has observed, categorized, and identified key relationships in industries undergoing disruption. Describing disruption from data has enabled him to construct a model of disruption that he is now applying to public schooling in the United States. In the second stage, researchers use their models to make predictions and uncover anomalies. “Anomalies are actually good news because they allow researchers to say, ‘There’s something else going on here,’ and that is what leads to better understanding.” “Researchers use the anomaly to revisit the foundation layers in the [descriptive stage] so they can define and measure the phenomena less ambiguously, or sort those phenomena into alternative categories. Only then can researchers explain the anomaly and the prior associations of attributes and outcomes.”

Christensen writes, “Our approach in researching and writing this book has been to stand outside the public education industry and put our innovation research on almost like a set of lenses to examine the industry’s problems from this different perspective.” Instead of taking a deep dive into education, he is taking a step back to look at schooling from a distance. In many ways, he isn’t saying what will happen based on his understanding of schooling and education; he is predicting what should happen based on his understanding of disruption. If his predictions are accurate, then his model of disruption is confirmed; if it isn’t and anomalies are discovered, then there is an opportunity to improve our understanding of both disruption and the public education system.

I think that there is great value in examining schooling through the lenses of disruption theory. It has helped me to categorize some of my own observations and bring certain relationships into focus. I also think it is valuable to take a fresh look at education from the outside. It is very interesting to compare what an outsider thinks should be happening to what an insider thinks is happening. That’s what I’ll be doing. In a series of blog posts, I will be re-examining Christensen’s examination, but doing it from the perspective of someone who has been much closer to the ground in schools.

In A Square Peg in a Round Hole, I examine why disruption theory may not apply to what is happening in public schooling in the United States today; in Underlying Assumptions, I examine some of the assumptions that Christensen makes in predicting how schooling will be disrupted; in Anomalous Readings, I discuss some of the anomalies detected in Christensen’s analysis; in Gazing Into the Crystal Ball, I compare and contrast Christensen’s prediction with my own; and in Removing Our Blinders, I examine how blinders could be preventing us from creating disruptive innovations in schooling and instruction.

Lessons from a Boucherie

In one episode of No Reservations, Anthony Bourdain travels to Cajun country and attends a boucherie, a massive party where families gather to break down and cook an entire hog over the course of a day. The tradition of the boucherie arose because there was no refrigeration, so fresh meat had to be prepared and eaten quickly.

The food and the live music at this particular boucherie is phenomenal. It seems as though every adult in the community is an expert at cooking at least one speciality dish and playing at least one musical instrument. You can see the kids in the community hovering around the action, their faces lit up as they try to absorb and learn as much as they can.

I believe that when my friend Alec talks about “everyday” learning, this is what he means. Learning is informal, immersive, and seemingly effortless. Alec recently shared two YouTube videos with me to highlight what everyday learning can look like. The first is from a documentary, Boxing Gym. The second is from a talk by Alan Kay where he shows footage of a tennis coach teaching a 55-year-old woman to play tennis in about 30 minutes.

Notice that I said, “can look like.” There is nothing everyday about these examples of everyday learning. The reporter who shot the footage of the tennis coach set up the shoot in order to discredit the coach because he was offended by the notion that almost anyone could learn to play tennis in a single afternoon. The reporter had been trying to learn to play tennis for years. And I haven’t seen the documentary, but I’m going to guess that the gym in Boxing Gym is quite a special place.

Alec has repeatedly argued that everyday learning is effective because we get what we need from it; I have repeatedly disputed that. For me, the footage of the tennis coach teaching the woman to play tennis in 30 minutes was not surprising. I have seen what people can do in the right circumstances. I have constructed some of those circumstances myself. When you know what people can do, then the everyday learning that we see everyday is far from “good enough.”

At most boxing gyms, I’m going to receive little tutelage unless I show promise as a boxer. The elite boxers aren’t going to take me under their wing. There is probably a clear pecking order and I will be bombarded with the message: “You're wasting your time and ours. You’re just not that good.” Some young boxers will persevere through that and still get what they need, maybe winning the respect of the other gym members over time. But I’d probably quit and try something else.

At the gym in Boxing Gym, I’m going to guess that, somehow, a culture of mentoring has taken root; everyone helps everyone, regardless of skill level. In fact, the members of the gym probably derive more pleasure from seeing a single unskilled boxer progress than in producing x number of world champions. And that probably isn’t due to self-selection. You start boxing there. You see the elite boxers, the ones you admire and want to emulate, patiently working with that clueless kid. You try it too, and you experience that pleasure. In another episode of No Reservations, an ex-con working at a community kitchen talks about what it was like to find a place for the first time where you checked your ego at the door because there was no need for it. These are life-changing experiences. When you know that these kinds of places exist, it is hard to argue that most of us get what we need.

The boucherie culture witnessed by Anthony Bourdain exists because there is a critical mass of domain and mentoring expertise in the community. That community will be self-sustaining only if that expertise is at a sufficiently high level. If the food and music produced at the boucherie is only mediocre, or if the kids don’t see themselves stepping into the roles of lead cook and lead musician in a few years, then you are going to hear a lot of teenagers asking: “Is it okay if I hang out with my friends at the mall instead?”

I talk a lot about the need to ratchet up our domain and mentoring expertise even higher, beyond what we need to sustain a learning culture. I want to find out what it takes to export a learning culture. It’s not that I want to see neighborhoods with boxing gyms at every corner or that throw boucheries every weekend. It’s that: before we can develop the domain and mentoring expertise we need to coach others effectively, we have to want it; and for most of us, before we can want it, we need to see it and know that it is possible. I want people to experience a boxing gym or a boucherie and wonder: “Gee, why doesn’t this exist for <insert your favorite domain here>?”

I live near Boston and Cambridge where there is a large and vibrant maker community. This maker community is making a determined effort to infuse its DNA into local schools. But the coaching available to kids growing up in the maker community sucks. Not most of it; all that I’ve seen. There is nothing happening that comes close to what I know that people can do given the right circumstances. How is this possible? If you have misconceptions about what people can do, then you will be limited by those misconceptions. The boucherie, the boxing gym, and the tennis coach can’t counter those misconceptions because not enough people have regular firsthand experience with them. If we do learn about them, we marvel at them instead of emulating them.

So here is the challenge as I see it. In order for learning to be effective, whether everyday learning or in-school learning, you need effective coaches. In fact, you need a large number of highly effective coaches so that everyone has regular contact with good coaching. But in order for people to even aspire to be a highly effective coach, they also need to come into contact with good coaching. It’s quite the chicken or the egg problem.

Self-Concept, Pick-Up Basketball, and Clash of Clans

There are a few reasons why I’m skeptical of using everyday learning as the model for all learning. One of them is the issue of self-concept.

After college, I attended graduate school at the University of California at Berkeley, and two of my closest college friends settled in the Silicon Valley area. My two friends loved to play basketball, and we spent one summer working out and going to local courts to play pick-up basketball games. Now, the courts we were playing on didn’t look anything like the courts in the movie, White Men Can’t Jump. Guys were most definitely not dunking all over the place. But games were super competitive and the people playing in them could play.

I’m not really sure why I allowed myself to get dragged to these games. My friends were good; I was not. Growing up, I had shot around with my sisters and younger brother, but I had never played in any games, so my ball-handling skills were next to nonexistent. When I stepped onto the court, my mindset was to avoid making visible mistakes. I didn’t want the ball passed to me, and I dreaded having to take an open shot. I focused on defense and rebounding. During the hours of practice we did between games, I did next to nothing to work on my ball-handling skills. In games, I dribbled exclusively with my right hand. With a little practice, I could have learned to dribble with both hands (drastically improving my game), but I never did.

One day, we were invited to play a full-court game. Courts were usually so crowded that most games were half-court, and only the best players were able to play full-court. They had four, but we only had three. So the four guys we were playing against scanned the sidelines and picked this young, chubby kid to be our fourth. It was obvious that no one wanted this kid on their team. He had probably been waiting all day to get into a game, but no one ever picked him.

He was a chucker. As soon as he touched the ball, if he was open and in range of the basket, he was going to shoot. It didn’t matter if I was wide-open beneath the basket for a lay up, there was no way in hell he was going to pass the ball. He was probably thinking that he never gets to play in any games and everyone already thinks he sucks, so he may as well make the most of his opportunity and shoot as often as he can. Besides which, if he did pass the ball to one of us, he figured we’d never pass it back.

The situation looked dire. We were shorter and slower than the four guys we were playing against (at 5' 8", I was playing center in our 1-3 zone), and we were saddled with a chucker and me, a half player. It should have been a romp for the other team. But my two friends and I had something that the other team didn’t: great teamwork. We clamped down on the other team on defense and managed to build a small lead.

I’m not sure what the kid, the chucker, was thinking during all of this. He was hurting us on both offense and defense, but we kept passing him the ball and trying to integrate him into our system. We treated him as though he was a fully functioning member of our team, as though we expected him to play smart and play hard, and we didn’t get down on him when he didn’t. At some point, he must of recognized that something different was going on because he stopped chucking the ball and started working with us. He hustled, looked for the open man, passed the ball, and covered his zone. When we signaled him to cut across the court or switch on defense, he did.

We ended up losing that day, but the game went down to the wire. It was exhilarating and I remember being in the flow. I stopped being self-conscious and started looking to push the ball up court and get open for baskets. Anything to help the team. Among the spectators, I’d like to think that some of them could appreciate what we were doing (that some of them were rooting for us), but the vast majority were simply incredulous that the other team was playing this poorly and thought that we had nothing to do with it. For most players at this level, individual skill is more highly regarded than teamwork.

The transformation in the kid that day was amazing. He went from a classic chucker to a team player in the course of one game. Along the way, he picked up some skills for playing in a zone and communicating on the court that he could build on with the right teammates. He felt this. If we had decided to hop in our cars and drive to another court, he would have gone with us. If we had shown up at this court again the next day, he would have joined us. He had decent skills as a player (better than me), but his lack of status on these courts had caused him to pick up bad habits and become a chucker. Once he became a chucker, even the more generous players on the court didn’t want to play with him. His own perception of himself, drawn from how everyone around him perceived him, became self-fulfilling. But it could all be reversed given the right opportunities, the right mentoring and coaching.

Fast-forward twenty years and this scenario is playing out again. I’m playing Clash of Clans in a clan with fifteen other guys. About half of the clan are in middle school. There is a clear and fairly rigid hierarchy in the game. At the top are the innovators. These are the players who design new bases and attack strategies. Then there are the players who copy the innovators, but can internalize those designs and strategies well enough to adapt them for different circumstances. Then there are the players who can copy the innovators, but can’t counter when a design or strategy is countered; they run the script the same way every time. Finally, there are the players who can’t even manage to run the script for a given design or strategy properly. These last guys “suck.”

A few months ago, Clash of Clans introduced clan wars. Before clan wars, you attacked for loot in order to upgrade your base and your troops. When attacking for loot, you can look for easy bases to attack. It is common to attack fewer than one in twenty bases you look at. And your attacks are private. You can share the replays of your attacks, but that is optional. A lot of players who are poor attackers end up gemming (using real money to buy upgrades instead of using in-game currency).

Once clan wars started, you started attacking bases at your own level for the clan, and those attacks are public. A lot of clan members were embarrassed by how much they “sucked” and would go onto YouTube to search for better base designs and attack strategies. As one of the innovators in the clan, I’m not a fan of copying ideas from YouTube or other sources. At a minimum, I feel like you should understand those base designs and attack strategies well enough to modify them. Over time, I offered tips and suggestions, and encouraged the guys to analyze each others’ attacks. Most ignored me, but a couple took up my advice and became incredibly strong attackers. Once that happened, more joined in. At this point, we have over a dozen very strong attackers, which is rare. Most clans that we go up against have 2-3 strong attackers and bunch of other guys that “suck.”

Yesterday, Supercell updated Clash of Clans with some significant changes to how troops behave, breaking most existing base designs and attack strategies. Across the internet, players are waiting for the innovators to absorb the changes and post new YouTube videos. In my clan, there was a great deal of unease. I suggested that we use the next clan war to focus on experimenting with new strategies, and to forget about winning for a while. This unleashed an immediate torrent of new attack strategies on online chat. Up and down the clan, guys are analyzing the impact of the changes and discussing how to deal with them. The level of analysis is impressive. There’s something else that I’ve noticed. If a new guy joins the clan and he “sucks,” there used to be a chorus to kick him out of the clan. Now, a new guy receives a steady stream of high quality advice from multiple members of the clan. It is gratifying to see the guys I once mentored mentoring others, and doing a better job at it than I did.

The pick-up basketball game and my clan in Clash of Clans are both examples of everyday learning. But they aren’t typical examples of everyday learning. In everyday learning, that chubby kid typically becomes a chucker, internalizes that he is a chucker, and remains a chucker for the rest of his life… with his basketball skills plateauing or atrophying. In Clash of Clans, only the elite players are innovators and everyone else settles for copying the elite players. In the world that I envision, everyone receives the opportunities to overcome and exceed those self-concepts if they choose to do so. As far as I can tell, that requires a large number of strong on-court and in-game mentors and coaches. How do we develop them? That’s what I’m working on.

One final note. Overcoming your self-concept and shifting what you believe you are capable of doing (the standard you establish as “good enough” performance for yourself), is only the first step. If that kid worked hard at becoming a good team player and developed a solid team around him, he wouldn’t necessarily win many games. And without that positive feedback from winning, it’s really tough to keep going and maintain the belief that you can do it. When we started that pick-up game, we started in a zone because it complemented our skills as team players and we didn’t think we could match the other team’s athleticism. We started in a 1-2-1 zone, but we couldn’t stop them from scoring against us. After huddling and analyzing the situation, we thought that a 1-3 zone might work better, and it did. We were able to perform that analysis because we had played and experimented with zones many times, and we had an understanding of their strengths and weaknesses. We also had strong analytical skills. Those things all take time to develop, which is why I think coaching is so important. It doesn’t have to be formal coaching. It could just be random guys making suggestions and observations here and there. But I think that pervasive high-quality coaching is necessary to achieve what I envision.

Long Division and Everyday Learning

Recently, my friend Alec and I have been exchanging emails on the role and effectiveness of everyday learning. Alec believes that all learning should look like everyday learning. While I don’t necessarily disagree with that, I’m not sure that everyday learning is as effective as Alec makes it out to be. It may be a good place to start, but I feel that we need to do much better than that.

So what is “everyday” learning? I’ve been waiting for Alec to define that for me since he’s the one that brought it up, but let me take a stab at it. For me, everyday learning typically has the following characteristics:

1. It is learning that occurs to meet an immediate need
2. The feedback that you receive while learning is nonjudgmental
3. You have plenty of opportunities to apply what you have learned

When you take on a home improvement project and start going to Home Depot every day, you quickly learn how the store is laid out. Your need is immediate: it’d be a big waste of time if you had to wander the entire store to find one item. The feedback is fairly nonjudgmental: either you can find what you need quickly and efficiently or not. (Okay, I do feel like everyone in Home Depot can tell I’m a noob when I’m wandering around like a lost sheep, but my need almost always outweighs the small discomfort I experience.) And you have plenty of opportunities to apply what you have learned: you make multiple trips and only learn the layout over time.

In terms of its everydayness, how you learn something shouldn’t matter. If you download a map of the local Home Depot and study it for hours the night before, ask a friend who knows the store like the back of her hand to go with you and show you the ropes on your first few trips, or go in blind… it is still everyday learning. In terms of effectiveness, my gut tells me that everyday learning is most effective when there is lots of immediate feedback and when you can take self-concept out of the picture.

As a teacher and curriculum developer, there are things that I do to make school learning more like everyday learning. In general, students are introduced to long division in 4th-grade and master it in 5th-grade. But what if you delivered a truck full of cookies to eight 2nd-graders on a playground and told them that they could keep the cookies only if they could divide the cookies evenly amongst themselves?

When they open up the truck, the kids find ten pallets:

I’m willing to bet that some enterprising kid would suggest that they start by each taking one pallet. When the two remaining pallets are unpacked, the kids find twenty crates:

This time, four of the kids suggest that they each take one, and then two, of the crates. When the four remaining crates are unpacked, the kids find forty cartons:

They quickly divide up the forty cartons, each taking five, leaving them with one pallet, two crates, and five cartons each, with nothing left over:

The kids would immediately recognize this as 125,000 cookies if they had ever played Chocolate Chip Cookie Factory: Place Value, but that’s not the point. They just applied the long division algorithm to divide up a million cookies on their own in under half an hour.

Now, if this was really an example of everyday learning, these 2nd-graders would work at some kind of chocolate chip cookie distribution center and they’d be doing these division problems all day. And they’d get very good at them. At first, their solutions would be ad hoc (each problem would be solved as a unique case), but over time, they would start to generalize and find an efficient solution for all problems. How long would that take? Some kids would start doing it right away; others would only do it subconsciously over months.

As a teacher, I can speed things up by designing and sequencing activities to stretch their thinking and by helping them raise things to a conscious level. For example, if a 2nd-grader had to divide three stacks of cookies two ways, he might immediately start by dividing two stacks and opening the third. This would be so obvious to him that he would do it without thinking about it.

But as his teacher, I could ask: “Why don’t you open up all three stacks before dividing them?” This would cause him to pause, think about it, and reply: “Well, that would be silly because you can already tell that each person is going to get one stack. It is faster and easier to start by dividing the big things first.”

I could then follow up by asking him to divide 4 boxes, 6 stacks, and 2 cookies three ways. He might start by dividing three of the boxes and the six stacks before unpacking the leftover box. We could discuss different strategies and explore “what-if” problems together. We could even talk about the benefits of grouping (place value) and how things would be different if everything at the cookie distribution wasn’t in groups of ten. Finally, we could invent a notation system for doing these division problems out on paper. Why would you ever want to do that? Well, imagine that, instead of being on the floor and moving boxes, crates, and pallets around manually, the distribution center became automated and you had to move things around symbolically in a separate control room. How many 2nd-graders would make that transition on their own instead of being laid off and having to find low-wage jobs in the service sector?

These things could happen in everyday learning if you had a good mentor or the distribution center had a surprisingly good training program. But I would say that these opportunities for generalization, reflection, and extension aren’t typical in everyday learning. Remember how I said that some 2nd-graders would start generalizing on their own right away? How did they get that way? Someone modeled it for them, they liked it and internalized it, and started doing it for themselves all the time. My goal as an educator is to guarantee that everyone, whether learning in school or in everyday life, has this same opportunity. I think this only happens if we collectively raise our expectations for what learning looks like so that these models, mentors, and coaches are everywhere for everybody.

When I think about teaching long division using chocolate chip cookies, I’m not thinking about bringing everyday learning into the classroom. I don’t see this curriculum as an example of everyday learning. To me, it is applying basic learning theory. These 2nd-graders have rich and powerful mental models for working with groups. They’ve been working with groups of things in their everyday lives for years. I’m simply leveraging those mental models because it would be incredibly foolish not to. It would take years of arduous work to build parallel mental models for the same concepts in the classroom, and then you’d have two disconnected ways of thinking: one way of thinking for everyday life and a separate way of thinking for the classroom. Who would want that? Crazy.

The Interplay of Affect and Cognition

I had dinner with my friend Alec again Friday night, and we had an interesting discussion about the roles that affect and cognition play in learning. We agreed that affect and cognition are both essential. To learn something, you have to want to learn it and you need to have the tools to understand what you are learning. But Alec felt that, in general, the cognitive tools are there and it’s a lack of affect that impedes learning in schools, while I felt the opposite.

We went back-and-forth on this topic for most of the evening without settling anything. But when I got home, I suddenly realized that I had been misstating my position. When I was a classroom teacher, I spent a significant chunk of my time thinking about and working on the affect of my students. So I had clearly identified a lack of affect; it’s just that the lever I used to shift their affect was providing them with the most powerful cognitive tools I could.

I watch a number of reality shows (Kitchen Nightmares, Restaurant Takeover, Restaurant Impossible) where expert chefs try to help owners turn around a failing restaurant. Most of the restaurants are failing because the owners lack the tools to run a restaurant, are afraid of change, and/or are too proud to admit they are wrong. Some of the tools they are lacking are fairly basic, such as keeping the restaurant clean and giving your staff clear direction, so that has to come down to affect (fear and ego).

The expert chefs work directly on affect, but they also work on helping the owners develop restaurant management skills at the same time. Most owners won’t change or admit to being wrong until they can envision a realistic chance that they can pivot and be successful. These tools range in scope and impact: from teaching the owner to cook a few new dishes, to focusing on food quality and always cooking from fresh, or being mindful to give your guests something special so that you stand out from the competition.

The more powerful the tool, the more empowered the owners and the greater the shift in affect. If the tools are powerful enough, some owners even report on follow up visits that their passion for the restaurant has only grown since the initial intervention and that they are experiencing a level of passion that they never expected or experienced before. For me, as a classroom teacher, I am always looking for cognitive tools that are easy enough for my students to use within their current level of affect, but powerful enough to shift their affect so that they can begin developing and applying a series of more powerful tools.

Apples and Bananas

So what does this look like in an academic setting? In 7th-grade, students learn to simplify expressions by combining like terms and applying the distributive property. However, instead of presenting it in that way, I give them an alternative cognitive tool that is both easier to understand and more powerful.

3 apples + 5 bananas + 2 bananas + 6 apples + 4 bananas

We start by adding apples and bananas. How many apples and bananas are there in total? Students sit up and take notice at this point because they know that I’m up to something. This is way too easy.

4 ( 2( 7a + b ) + 3( 5b + 6a + 2b ) + 10a )

Then I introduce the concept of baskets and nested baskets. Here we have 4 large baskets. Inside each of the large baskets there are 5 smaller baskets and 10 apples. Inside two of the smaller baskets, there are 7 apples and 1 banana. Inside the other three baskets, there are 5 bananas, 6 apples, and 2 bananas. How many apples and bananas are there in total?

At this point, all I am doing is making sure that students can translate the apple, banana, and basket notation into the real world. After that, it is completely up to them to unpack the baskets, and combine apples and bananas. (The answer is 168 apples and 92 bananas.)

This cognitive tool for thinking about and simplifying expressions is easy because it builds on top of existing mental models. The students can predict what they’ll have when they unpack a basket; they can reason through problems without trying to remember a set of procedures. It’s also powerful because they can simplify complex expressions with nested parentheses to any depth… and they can do that in the first 30 minutes of class.

It’s also more powerful because I can ask students to unpack the large baskets first (some students do that on their own anyways).

4 ( 2( 7a + b ) + 3( 5b + 6a + 2b ) + 10a ) = 8 (7a + b) + 12 (5b + 6a + 2b) + 40a

Most teachers would never do that because they would be afraid of confusing their students with too many rules to remember. The standard cognitive tools for simplifying expressions are more fragile and less flexible, and students experience that.

Now, if this is all we did, students would think it was a fun interlude, but that’s about it. If I taught a few more topics this way, it might shift some of their affects a little bit, but they would still attribute their success more to me than to themselves. In other words, they would still see themselves the same way.

To truly shift their affects, I try to get them to learn A, and then B. Then I try to get them to use A and B to learn C, where learning C doesn’t involve any creative tricks on my part. I want them to see that if they really understand A and B (to a greater extent than the standard cognitive tools allow), then they can apply their understanding to learn something really challenging, such as C… and that’s all them. Take that up another level and students start to internalize that they really are good learners, which shifts their affect and can ignite a virtuous cycle. They become passionate about learning and deeply understanding A and B because they know it will enable them to understand C, and they can see themselves doing that in any field they choose.

Teaching a Robot How to Dance

For apples and bananas, you are going to have to take my word for the impact it can have on students. But you can see for yourself how powerful cognitive tools can shift affect by watching two high school juniors, Amalia and Ariana, figure out how to program a robot with follow-the-leader behavior. The 90-minute session is documented in a series of video clips on the Computing Explorations website.

A group of us designed a five-week introductory computer science course for high school students with no background or prior interest in programming or robotics. The course is project-based: students learn how to program and how robots work in order to “teach” a robot to improvise a lead-follow dance with an unknown robot partner in time with unknown music. One of the intermediate challenges is to figure out how to get your robot to track and follow another robot.

If you watch the video clips, you will see that Amalia and Ariana are captivated by the robot and it’s life-like follow-the-leader behavior. It behaves like a cute pet. But that initial interest won’t take them anywhere unless we can give them the cognitive tools to replicate and understand the behavior themselves. After all, they will have to program much more complex behaviors in order to complete the project.

We start by giving them basic cognitive tools for understanding differential steering and the IR sensors. Then we show them how to link them using behavior-based programming (condition-action rules and behavior arbitration). They explore these concepts until they feel like they have a concrete understanding of how the robot works. Then they try to apply what they know to program the robot to orient itself to and follow the beacon. When unexpected behaviors emerge, they need to debug the robot.

At this point, Daniel (the instructor) models how to use the real-time instrumentation to figure out what the robot is “seeing” and to identify the behaviors being triggered. Amalia and Ariana quickly see the utility of this approach, and begin eagerly and capably debugging the robot themselves. They go through six iterations of the engineering-design process to get a robot with working follow-the-leader behavior.

By the end of the session, Amalia and Ariana are noticeably more engaged by the robot and the task than they were at the start of the session. This happened because of what they were able to do: they understood differential steering, IR sensors, and behavior-based programming (level 1); they applied those three elements to program the robot to do something (level 2); and they used real-time instrumentation and the engineering-design process to debug the robot (level 3), resulting in a rewarding accomplishment. And they did it in 90 minutes.

If we stopped there, Amalia and Ariana would have happily gone on to program more cute life-like behaviors into the robot. To create the behaviors they envisioned, they may even have gone out and acquired more powerful cognitive tools for understanding how the robot works. But in case that doesn’t happen (their affect is such that they stay in their comfort zone and plateau), we designed the final project so that they would be compelled to deepen and expand their understanding in order to complete it. Our goal was to get them to a place by the end of the course where they would be eager and ready to take on new challenges, and ideally transfer their new tools and affects to other domains.

Wrap Up

• Affect and cognition are both essential for learning.
• Helping students develop powerful cognitive tools is an effective and necessary way to shift affect.
• Cognitive tools should be easy enough for students to develop given their current level of affect, but powerful enough to shift their affect, hopefully triggering a virtuous cycle.
• In order for learning to have a significant impact on affect, students must see themselves differently because of what they learned.
• Students will see themselves as more capable learners if they go vertical (rapidly building on top of what they have just learned in order to reach a place that they never thought they could reach).