Chapter 6

GRIT, Character, and Achievement: A New Theory of Intelligence

The University of Pennsylvania’s Psychology Department has an exceedingly competitive PhD program. Every year, we get several hundred applicants, and we take only ten or so. Positive psychology receives around thirty applicants a year and accepts only one. The model admittee is an undergraduate psychology major with a grade point average close to perfect from a major American or European university, with GRE exam scores all well over 700, and with three letters of recommendation—each of which checks the box that says this candidate is “truly exceptional, the best in years.” The admissions committee is traditional, but it is constipated (I have never served on it), and it has turned down some amazingly fine candidates.

One that comes to mind was one of the first women to win a major championship in poker. In her essay she said that she saved up lots of money, took a plane to Las Vegas, and entered the world championship, and that she won the tournament. Both the president of the university, Sheldon Hackney, and I argued that she should be admitted for having demonstrated not mere potential but actual world-class performance—but to no avail. Her GREs, we were told, were not high enough. I am still grateful to her, however, for spending part of her interview correcting my poker errors, thus saving me thousands of dollars over the next decade. “Courage,” she said, “is the key to high-stakes poker. You must treat the white chip as simply a white chip, whether it is worth just one nickel or one thousand dollars.”

Success and Intelligence

Applications are due by January 1, and, after a grueling series of personal interviews, acceptances go out in late February. This has been the operating procedure for the forty-five years that I have been in this department. As far as I know, there has only been a single exception in all this time: Angela Lee Duckworth.

In June 2002, we received a belated application for the class entering in September 2002. It would have been thrown out summarily but for the intercession of the director of graduate training, John Sabini. John, may he rest in peace—he died suddenly at age fifty-nine in 2005—had always been a maverick. He worked on such unconventional topics as gossip, claiming that it is a legitimate form of moral sanction but at a less punitive level than legal sanction. In whatever he did, he rowed upstream against academic social psychology. I had always been the other department maverick, usually loyal to the unpopular argument, the argument that needed hearers. John and I could smell another maverick a mile away.

“I know it’s intolerably late, but you must read this admissions essay, Marty,” John emailed me. It was written by Angela Lee Duckworth. Here is the content, in part:

By graduation, I had spent at least as many hours volunteering in the classrooms of the Cambridge Public Schools as I did in Harvard’s lecture halls and laboratories. Witnessing in person the reality of failing urban students in failing urban public schools, I chose conscience over curiosity. I made a commitment to pursue public education reform after graduation. During my senior year, I founded a nonprofit summer school for low-income middle school students … Summerbridge Cambridge developed into a model for other public schools across the country, was featured on NPR and in many newspapers, was written up as a case study for the Kennedy School of Government, and won the Better Government Competition for the state of Massachusetts.

I spent the next two years at Oxford University on a Marshall fellowship. My research focused on the magnocellular and parvocellular pathways of visual information in dyslexia … I chose not to pursue a PhD at that point in my career … I spent the next six years as a public school teacher, nonprofit leader, charter school consultant, and education policy writer.

After years of working with students at both ends of the achievement spectrum, I now have a distinctly different view of school reform. The problem, I think, is not only the schools but also the students themselves. Here’s why: learning is hard. True, learning is fun, exhilarating, and gratifying—but it is also often daunting, exhausting, and sometimes discouraging. By and large, students who no longer want to learn, who don’t think they can learn, and who don’t see any point in learning simply won’t—no matter how wonderful the school or teacher …

To help chronically low-performing but intelligent students, educators and parents must first recognize that character is at least as important as intellect.

I have chosen not to exhume my essay for admission to Penn graduate school in 1964 and compare it to this one.

Conventional wisdom and political correctness have for almost a century blamed the teachers, the schools, the classroom size, the textbooks, the funding, the politicians, and the parents for the failure of the students—putting the blame on anything or anyone but the students themselves. What? Blame the victim? Blame the character of the students? What nerve! Character had long since gone out of fashion in social science.

Positive Character

In the nineteenth century, politics, morality, and psychology were all about character. Lincoln’s first inaugural address, appealing to the “better angels of our nature,” was emblematic of how Americans explained good behavior and bad behavior. The Haymarket Square riot in 1886 in Chicago was a turning point. There was a general strike, and someone, unknown to this day, threw a pipe bomb; the police opened fire, and in a five-minute melee, eight policemen and an unknown number of civilians were killed. German immigrants were blamed, and the press condemned them as “bloody brutes,” “monsters,” and “fiends.” The deaths were caused, in popular sentiment, by the bad  moral character of the immigrants, and they were labeled anarchists. Four of them were hanged; a fifth committed suicide before his execution.

There was an enormous reaction against the hangings from the left. Riding on the coattails of this protest was a very big idea: an alternative explanation of bad character. All of the condemned came from the lowest class of workers. They were illiterate in English, desperate, on starvation wages, and overcrowded, with whole families living in one small tenement room. The big idea claimed that it was not bad character but a malignant environment that produced crime. Theologians and philosophers took up this cry, and the end result was “social science”: a science that would demonstrate that environment, rather than character or heredity, is a better explanation of what people do. Almost the entire history of twentieth-century psychology and her sister disciplines of sociology, anthropology, and political science have acted out this premise.

DRAWN BY THE FUTURE, NOT DRIVEN BY THE PAST

Notice the cascade of changes that follow from giving up character as an explanation of human misbehavior in favor of the environment. First, individuals are no longer responsible for their actions, since the causes lie not in the person but in the situation. This means the interventions must change: if you want to make a better world, you should alleviate the circumstances that produce bad actions rather than waste your time trying to change character or punishing bad behavior and rewarding good behavior. Second, progressive science must isolate the situations that shape crime, ignorance, prejudice, failure, and all the other ills that befall human beings, so that these situations can be corrected. Using money to correct social problems becomes the primary intervention. Third, the focus of inquiry must be bad events, not good events. In social science, it makes sense to excuse Sammy’s failure at school because she was hungry, or abused, or came from a home in which learning was not valued. In contrast, we don’t take the credit away from people who do good things, because it makes little sense to “excuse” good behavior by invoking the circumstances that led up to it. It is odd to say that Sammy gave such a good speech because she went to good schools, had loving parents, and is well fed. Finally, and so basic as to be almost invisible, the situation view propounds the premise that we are driven by the past rather than drawn by the future.

Psychology-as-usual—the psychology of victims and negative emotions and alienation and pathology and tragedy—is the stepchild of Haymarket Square. Positive psychology’s take on all of this is very different from psychology-as-usual: sometimes people are indeed victims (I am writing this on the day after the horrific Haitian earthquake, with hundreds of thousands of genuine victims now suffering or dead), but often people are responsible for their actions, and their untoward choices stem from their character. Responsibility and free will are necessary processes within positive psychology. If the circumstances are to be blamed, the individual’s responsibility and will are minimized, if not eliminated. If, in contrast, the action emanates from character and choice, individual responsibility and free will are, at least in part, causes.

This has direct implications for how to intervene: in positive psychology, the world can be bettered not only by undoing malignant circumstances (I do not remotely advocate giving up on reform) but also by identifying and then shaping character, both bad and good. Reward and punishment shape character, not just behavior. Good events, high achievement, and positive emotions are just as legitimate objects of science for positive psychology as are awful events, failure, tragedy, and negative emotion. Once we take positive events seriously as objects of science, we notice that we do not excuse or take credit away from Sammy’s brilliant performance because she was well fed, or had good teachers, or had parents who cared about learning. We care about Sammy’s character, her talents, and her strengths. Finally, human beings are often, perhaps more often, drawn by the future than they are driven by the past, and so a science that measures and builds expectations, planning, and conscious choice will be more potent than a science of habits, drives, and circumstances. That we are drawn by the future rather than just driven by the past is extremely important and directly contrary to the heritage of social science and the history of psychology. It is, nevertheless, a basic and implicit premise of positive psychology.

Angela’s proposal that school failure might stem in part from the character of the failing students, and not just from the system that victimizes them, appealed to the positive psychologist in me and to the nurturing of mavericks that was the cornerstone of John Sabini’s pedagogy. Here was just the right sort of maverick: someone with very high intellectual credentials and a sterling education but not housebroken enough by politics to prevent her doing serious research on the character strengths of students who succeed and the character deficits of students who fail.

What Intelligence Is

SPEED

We interviewed Angela forthwith. My first impression of her triggered a memory I have to recount. In the 1970s, I was one of two professors at Penn to found a college house system; Alan Kors, a professor of modern European intellectual history, believed that a university education was truly about the life of the mind. But when we taught our undergraduates, we saw the chasm that separated the classroom from what they considered their real lives: they could simulate intellectual passion in the classroom in order to get good grades, but once released, it was party, party, party. Alan and I had experienced this animal life firsthand in the dorms of Princeton in the early 1960s, but we were vouchsafed a safe haven, one that changed both our lives: Wilson Lodge, an eating facility at Princeton in the 1960s. To this day, after a lifetime of intellectual feasts, it remains the best intellectual experience of my entire life. The president of the senior class, Darwin Labarthe, whom you will hear more of in the next chapter, inspired by university president Robert Goheen, led a walkout against Princeton’s entrenched, anti-intellectual, anti-Semitic club system. Together they created Wilson Lodge, open to any student or faculty member, and more than a hundred of the most intensely intellectual students joined, along with forty of the most devoted faculty.

Alan and I believed that a system in which such devoted faculty lived in the dorms with the undergraduates would provide the same antidote to the animal life in Penn’s dorms. So we founded Penn’s College Houses in 1976. Van Pelt College House was the first one, and Alan, a bachelor—no faculty we could drum up were devoted enough to give up their family life to live with 180 undergraduates—agreed to become the first housemaster. After my divorce, I succeeded him in 1980. I cannot pretend that this was an easy job; in fact, it was the only job I’ve ever held that I count as a failure. My own inadequacies in dealing in loco parentis twenty-four hours a day with late adolescents, trying to settle the endless squabbles among roommates, the suicide attempts, the date rapes, the meanness of the pranks, the absence of privacy, and, worst of all, an unsympathetic administration that treated the resident faculty not as professors but as hourly employees, made the master’s life an endless hassle.

But the intellectual life we created was an improvement, and it survives to this day. And the parties were great. The students called the parties “master blasters.” The centerpiece of these parties was a student named Lisa, who was an astonishingly graceful dancer. The music was rock, usually with a heavy, very fast beat. Lisa somehow took two steps for every one beat, dancing twice as fast as anyone else, from the opening and late into the night.

This brings me back to my first impression of Angela Lee Duckworth: she was the verbal equivalent of Lisa—talking twice as fast as anyone else I knew, indefatigably, and still making good sense.

Speed is something that both attracts and repels in academic life, and it plays a central role in what I think intelligence really is. Intellectual speed was highly prized by my parents and my teachers: the prototypes to be emulated were Dickie Freeman and Joel Kupperman, two of the prodigies who starred in Quiz Kids, a weekly radio show in the early 1950s. They rattled off the answers to factual questions more quickly than the other contestants, questions such as “What state ends in ut?” I know because I competed in the local radio version of this in fourth grade, getting that one right, and guessing correctly that there were five Little Peppers (an old book series). But I came in second to Rocco Giacomino, and therefore failed to make it to the national show when I was stumped by “Who wrote ‘Flow Gently Sweet Afton’?”

My parents’ and teachers’ bias toward speed was not an accidental social convention. It turns out that speed and IQ have a surprisingly strong relationship. In the experimental procedure called “choice-reaction time,” subjects are seated in front of a panel with a light and two buttons. They are told to press the left button when the light is green and to press the right button when the light is red, and to do this as fast as they can. IQ correlates almost as high as +.50 with how fast people can do this. Being fast at choice-reaction time is not simple athleticism, however, since its correlation with “simple reaction time” (“When I say, ‘Go,’ press the button as fast as you can”) is negligible.

Why should intelligence be so closely related to mental speed? My father, Adrian Seligman, was deputy state reporter of the Court of Appeals of New York State. His job was to take the unwieldy and ungrammatical opinions of the seven high court judges and translate them into a readable legalese that resembled English. He was blazingly fast. According to my mother, Irene, who as a legal stenographer was a keen observer, Adrian could do in one hour what it took other lawyers the entire working day to do. This gave him seven hours to check and refine his work, rewrite and rewrite, so that the finished product was much better than what other reporters could produce.

Any complex mental task—rewriting legal opinions, multiplying three-digit numbers, mentally counting the windows in your childhood house, deciding which blood vessel to suture first or whether the next hilltop is a likely ambush site—has fast automatic components and slower voluntary components that take much more effort. You are an experienced staff sergeant urgently approaching a hilltop in Afghanistan. You scan the approach, and from your previous encounters, you know instantaneously that the freshly disturbed soil, the silence, and the absence of animal sounds are danger signs. The more components of a task you have on automatic, the more time you have left over to do the heavy lifting. You now have two minutes to radio the base and ask for the latest report on the presence of foreign fighters. You are told that people in the nearest village spotted three strangers this morning. All this spells ambush or an improvised explosive device, and so you take the long route around the hill. The two extra minutes left over save lives.

The staff sergeant’s mental speed is a surrogate for what proportion of the task he already had on automatic. I see this in serious bridge every time I play (which averages about three hours a day on the Internet). I have played more than 250,000 hands in my lifetime, and all four-way combinations of thirteen (in bridge, each player holds thirteen cards made up of four suits) are now automatic for me. So if I find out that an opponent has six spades and five hearts, I know—instantaneously—that she either has two diamonds and no clubs, two clubs and no diamonds, or one of each. Less experienced players have to calculate that remainder, and some even have to say it to themselves. I actually had to say “two diamonds and no clubs, or two clubs and no diamonds, or one club and one diamond” silently to myself until about my 100,000th hand. A bridge hand, like most of life’s trials, is a timed event. You get only seven minutes for each hand in a duplicate bridge match, and so the more combinations you have on automatic, the more time you have to do the heavy lifting and figure out if the most likely winning play is the simple finesse or the squeeze or the end play.

What distinguishes a great bridge player or a great surgeon or a great pilot from the rest of us mortals is how much they have on automatic. When the bulk of what an expert does is on automatic, people say she has “great intuitions.” Therefore, I take speed very seriously.

Angela (whose theory makes up this chapter) puts it like this:

Most of us recall from high school physics class that the motion of objects is described in the following terms: distance = speed × time. This equation specifies that the effects of speed and time are interdependent and multiplicative rather than independent and additive. If time is zero, whatever the speed, distance will be zero …

Distance seemed to me an apt metaphor for achievement. What is achievement, after all, but an advance from a starting point to a goal? The farther the goal from the starting point, the greater the achievement. Just as distance is the multiplicative product of speed and time, it seems plausible that, holding opportunity constant, achievement is the multiplicative product of skill and effort. Leaving out coefficients, achievement = skill × effort.

Tremendous effort can compensate for modest skill, just as tremendous skill can compensate for modest effort, but not if either is zero. Further, the returns on additional effort are greater for highly skilled individuals. A master woodworker will get done more in two hours than an amateur will in the same period.

So a major component of skill is how much you have on automatic, which determines how fast you can complete the task’s basic steps. As a youth, I became fast, blazingly fast, and I started my academic career with almost the same rate of speech as Angela. I blazed my way through graduate school, not only talking fast but conducting research fast. I earned my PhD after only two years and eight months from my undergraduate degree, and I received an annoyed note from John Corbit, my former professor at Brown University, about my having broken his old record of three years flat.

THE VIRTUE OF SLOWNESS

There is more to intelligence and high achievement, however, than sheer speed. What speed does is give you extra time to carry out the nonautomatic parts of the task. The second component of intelligence and achievement is slowness and what you do with all that extra time that being fast affords you.

Mental speed comes at a cost. I found myself missing nuances and taking shortcuts when I should have taken the mental equivalent of a deep breath. I found myself skimming and scanning when I should have been reading every word. I found myself listening poorly to others: I would figure out where they were headed after their first few words and then interrupt. And I was anxious a lot of the time—speed and anxiety go together.

In 1974 we hired Ed Pugh, a perception psychologist who worked on exacting questions such as how many photons of light are needed to fire off a single visual receptor. Ed was slow. He wasn’t physically slow (he had been the quarterback of his Louisiana high school team), and it wasn’t just the drawl, it was his rate of speech and his reaction time to a question. We called Ed “thoughtful.”

Ed was Penn’s incarnation of the legendary William K. Estes, the greatest of the mathematical learning theorists, and the slowest psychologist I ever met. Conversations with Bill were agony. I had worked for a couple of years studying dreaming—in particular researching what dreaming accomplished for Homo sapiens, given that we lie there physically paralyzed and vulnerable to predators during rapid eye movement sleep for about two hours a night. I encountered Bill at a convention about thirty years ago and asked him, “What do you think the evolutionary function of dreaming is?”

Bill stared without blinking at me for five seconds, ten seconds, thirty seconds (it was so bizarre, I actually counted). After one full minute, he said, “What, Marty, do you think the evolutionary function of waking is?”

I found myself at a party with Ed, and during a long pause that reminded me of the profundities that could issue from Bill after such a pause, I asked Ed, “How did you become so slow?”

“I wasn’t always slow, Marty. I used to be fast; almost as fast as you are. I learned to become slow. Before my PhD, I was a Jesuit. My socius [the mentor who socializes the Jesuit student, in contrast to the other mentor who grades the student] told me I was too fast. So every day he would give me one sentence to read, and then he made me sit under a tree for the afternoon and think about that sentence.”

“Can you teach me to be slow, Ed?”

Indeed he could. We read Søren Kierkegaard’s Fear and Trembling together, but at the rate of one page a week, and to top it off, my sister, Beth, taught me transcendental meditation. I practiced TM faithfully, forty minutes a day for twenty years. I cultivated slowness, and I am now even slower than Ed was then.

What does slowness accomplish in the equation achievement = skill × effort?

Executive Function

Adele Diamond, a professor of developmental psychology at the University of British Columbia and one of my favorite neuroscientists, slows down kindergarten children. Impulsive children have long been known to do worse and worse as they age: Walter Mischel’s classic marshmallow study demonstrated that children who gobbled up the one marshmallow in front of them, rather than waiting for two marshmallows a few minutes later, did poorly. More than a decade later, they got lower grades in school and lower SAT scores than children who could wait. Adele believes that the failure of children to control their fast emotional and cognitive impulses is the seed crystal around which the cascade of school failure begins. Teachers get annoyed and frustrated with such children, and school becomes less fun for these kids. They have difficulty complying with rules, and they become more anxious and avoidant. The teachers expect less and less from these kids, school becomes more of a misery, and the vicious cycle of failure has begun.

Adele believes it is crucial to interrupt these fast processes and get these children to slow down. Going slow allows executive function to take over. Executive function consists of focusing and ignoring distractions, remembering and using new information, planning action and revising the plan, and inhibiting fast, impulsive thoughts and actions.

Adele uses techniques from Deborah Leong and Elena Bodrova’s Tools of the Mind curriculum to slow down impulsive children. One of their techniques is structured play. When a teacher asks a four-year-old to stand still for as long as he can, one minute is the average. In contrast, in the context of a make-believe game in which the child is the guard at a factory, he can stand still for four minutes. Adele has found that Tools of the Mind kids score higher on tests that require executive function.

What other slow processes, in addition to more use of executive function, does having a large amount of the task done quickly and automatically allow? Creativity is surely one. In the equation achievement = skill × effort, achievement is defined not simply by any motion but by motion toward a specific, fixed goal—a vector—as opposed to sheer distance. There are usually several paths to a goal. Some get you there quickly, some slowly, and some paths are dead ends. Deciding which path to take is the slow process we call “planning,” and beyond this, the invention of new paths captures much of what is meant by creativity.

Rate of Learning: The First Derivative of Speed

Mental speed for any given task reflects how much material relevant to that task is already on automatic. We call this material “knowledge,” how much you already know that is relevant to the task. Speed on a task can change over time, and this parallels “acceleration,” the first derivative of speed in mechanics. Is there such a thing as mental acceleration, the increase of mental speed over time, how fast you can acquire new knowledge—the increase in how much of a given task can be put onto automatic over time and experience? We call this the “rate of learning”: how much can be learned per unit time.

Angela was fast, about as fast mentally as it is possible for a human being to be, and she blew us away in the interview. In violation of precedent, the admissions committee gave in and accepted her. She began work immediately on her grand project of examining the character of good and bad students, but then something embarrassing happened. To explain this, we need to go deeper into the nature of achievement.

While Angela was fast, she was ignorant of psychology, woefully so, probably because of the fact that almost all of her prior education was outside of psychology. To socialize her in positive psychology, I invited her in August 2002 to an elite event. Each summer I held a weeklong meeting that brought together twenty very advanced graduate students and postdoctoral students from all over the world with several leading senior positive psychologists. The competition for invitations was fierce and the level of sophistication very high. Never reluctant to pipe up, Angela participated in the conversations, but the feedback I got about her was disappointing. “Who is this clunker you imposed on us?” was one comment from a senior figure.

One criterion of the quality of a car is its speed. Mental speed is a very good quality because it is a surrogate for how much old knowledge is on automatic. But acquiring new knowledge that is not yet on automatic can be slow or fast. Acceleration, how much speed increases per unit of time, is the first derivative of speed, and it is an additional criterion of the quality of a car. Mental acceleration, the rate at which new stuff is learned per unit of time devoted to learning, is another part of the package we call “intelligence.” It turned out that Angela’s mental acceleration was as breathtaking as her speed.

Everyone learns in graduate school, and it is expected that a graduate student will become an expert in his or her small domain pretty quickly. But no student I have known learned at the rate that Angela did; she became the resident master of the huge and methodologically complex literature on intelligence, motivation, and success. Within months, my other students (and I) were going to Angela for advice about the literature and methodology in the psychology of intelligence. She went from clunker to Ferrari (the Enzo model) in about twelve months.

So far then in our theory of achievement, we have explored the following:

 

• Speed: the faster, the more material on automatic, the more one knows about the task.

• Slowness: the voluntary, heavyweight processes of achievement, such as planning, refining, checking for errors, and creativity. The faster the speed, the more the knowledge, and thus the more time left over for these executive functions to be used.

• Rate of learning: how fast new information can be deposited into the bank account of automatic knowledge, allowing even more time for the slow executive processes.

SELF-CONTROL AND GRIT

The three cognitive processes described above all make up “skill” in our basic equation, achievement = skill × effort. But the big game that Angela was stalking was not the cognitive processes in academic achievement but the role of character and where character enters the equation is as “effort.” Effort is the amount of time spent on the task. As she declared in her essay, what she determined to explore was the noncognitive ingredients. The noncognitive ingredients of achievement are summarized by effort, and effort in turn simplifies to “time on task.” The giant in the field of effort is a tall, shy, but unyielding Swede, Professor Anders Ericsson of Florida State University.

Ericsson has argued that the cornerstone of all high expertise is not God-given genius but deliberate practice: the amount of time and energy you spend in deliberate practice. Mozart was Mozart not primarily because he had a unique gift for music but because from toddlerhood, he spent all his time using his gift. World-class chess players are not faster of thought, nor do they have unusually good memories for moves. Rather they have so much experience that they are vastly better at recognizing patterns in chess positions than lesser chess players—and this comes from the sheer amount of their experience. World-class piano soloists log 10,000 hours of solo practice by age twenty, in contrast to 5,000 hours for the next level of pianist, and in contrast to 2,000 hours for merely serious amateur pianists. The prototype of deliberate practice is one of Ericsson’s graduate students, Chao Lu, who holds the Guinness World Record for the amazing number of digits of pi he memorized: 67,890! The advice that follows is straightforward: if you want to become world class at anything, you must spend 60 hours a week on it for ten years.

What determines how much time and deliberate practice a child is willing to devote to achievement? Nothing less than her character? Self-discipline is the character trait that engenders deliberate practice, and Angela’s first plunge into research on self-discipline was with the students of Masterman High School, the great magnet school in  the center of Philadelphia. Masterman accepts promising students beginning in the fifth grade, but many of them wash out, and the real competition begins in the ninth grade. Angela wanted to find out how self-discipline compares with IQ in predicting who will succeed.

IQ and academic performance are part of a well-worked-over field with lots of established measures, but self-discipline is not. So Angela created a composite measure that encompassed the different aspects of self-discipline that eighth graders show: the Eysenck Junior Impulsiveness Scale (yes/no questions about doing and saying things impulsively), a parent and teacher self-control rating scale (“compared to the average child [4], this child is maximally impulsive [7] to maximally self-controlled [1]”), and delay of gratification (over a range of dollars and times; for example, “Would you rather I gave you one dollar today or two dollars two weeks from today?”). Watched over the next year, the highly self-disciplined eighth graders

• earned higher grade point averages,
• had higher achievement test scores,
• were more likely to get into a selective high school,
• spent more time on their homework and started it earlier in the day,
• were absent less often,
• watched less television.

How does IQ compare with self-discipline in predicting grades? IQ and self-discipline do not correlate with each other significantly; in other words, there are just about as many low-IQ kids who are highly self-disciplined as there are high-IQ kids who are highly self-disciplined, and conversely. Self-discipline outpredicts IQ for academic success by a factor of about 2.

This project was Angela’s first-year thesis, and I encouraged her to submit it for publication—which she did. I am an old hand at publishing journal articles, but this was the first time in my experience that I saw an acceptance by return mail, from a top journal, and with no request for any major revisions. Angela concludes the article with these ringing sentences:

Underachievement among American youth is often blamed on inadequate teachers, boring textbooks, and large class sizes. We suggest another reason for students falling short of their intellectual potential: their failure to exercise self-discipline … We believe that many of America’s children have trouble making choices that require them to sacrifice short-term pleasure for long-term gain, and that programs that build self-discipline may be the royal road to building academic achievement.

This also solves one of the perennial riddles about the gap in school achievement between girls and boys. Girls get higher classroom grades than boys all the way from elementary school to college in every major subject, even though girls do not have higher IQs, on average, than boys do. In fact, boys often outperform girls by a little bit on intelligence and achievement tests. IQ overpredicts the grades that boys get and underpredicts the grades that girls get. Could self-discipline be the missing piece in this puzzle?

Angela used her battery of self-discipline measures with girls and boys from the beginning of eighth grade to predict algebra grades, attendance, and math achievement test scores at the end of the year. Girls indeed received higher classroom grades than boys, but the math achievement scores did not differ significantly. The achievement test underpredicted the girls’ classroom grades, as expected. Importantly, girls were much more self-disciplined than boys on all components of the composite. The question, then, is: Does the girls’ superiority in self-discipline account for their superiority in classroom grades? To answer this, a statistical technique called “hierarchical multiple regression” does the trick: this basically asks, When you remove the difference in self-discipline, does the difference in grades vanish? And the answer is yes.

Angela then repeated this study using IQ the next year at Masterman. Girls once again got higher grades in algebra, English, and social studies, and were much more self-disciplined. Boys had significantly higher IQ scores than this group of girls, and once again the classroom grades of the girls were underpredicted by IQ and standardized tests. Using multiple regression, girls’ self-discipline was again the major factor in their superior classroom grades.

While this solves the question of why women get better grades all the way through college, it assuredly does not tell us why men then go on to more professional and graduate degrees and earn higher salaries than women. Women’s superior self-control does not wane with maturity, but after college, many are swamped by cultural factors that dampen the female self-discipline edge.

Self-control predicts matters academic, but how does it predict other outcomes? Obesity, for instance, may have its roots in a critical period: weight gain during early puberty. Angela looked at the school nurses’ records of the weight of the fifth graders she had measured for self-discipline in 2003 and asked how much weight had they gained by eighth grade. Self-discipline did the same thing for weight gain that it did for grades. The kids with high self-discipline did not put on as much weight as the kids with low self-discipline. IQ had no impact on weight gain.

GRIT VERSUS SELF-DISCIPLINE

If we want to maximize the achievement of children, we need to promote self-discipline. My favorite social psychologist, Roy Baumeister, believes it is the queen of all the virtues, the strength that enables the rest of the strengths. There is, however, an extreme trait of self-discipline: GRIT. Indeed, Angela went on to explore grittiness, the combination of very high persistence and high passion for an objective. A modicum of self-discipline, we have seen, accounts for considerable achievement, but what accounts for truly extraordinary achievement?

Extraordinary achievement is very rare. That probably sounds like it is true by definition, a tautology: “very rare” just means the same thing as “extraordinary,” but this is not a tautology, and why it is not exposes the hidden scaffolding behind genius. Most people believe that “genius,” a term I will use as a synonym for truly extraordinary achievement, is simply being in the very extreme of the positive tail of a bell-shaped or “normal” distribution of success. The bell curve works well for ordinary stuff, such as charm, beauty, school grades, and height, but it totally fails to describe the distribution of achievement.

High Human Accomplishment

Charles Murray, the eminent sociologist, in his magnum opus, Human Accomplishment, starts with sports. How many PGA tournaments does the average professional golfer win in his lifetime? The mean is between zero and one. (The mode, or most frequent value, is zero.) But four professional golfers have won thirty or more, with Arnold Palmer winning sixty-one and Jack Nicklaus winning seventy-one (as has Tiger Woods at this writing). The shape of the distribution of the number of PGA tournaments won by a player is not remotely bell shaped but concave upward and extremely skewed (like a cliff) to the left.

The technical name for this kind of curve is log-normal, which means that the logarithm of the variable is normally distributed. The very same pattern is true of tennis, marathons, chess, and batting championships, and as the accomplishment becomes more demanding, the curve appears more and more clifflike. In each of these domains, there are many fine competitors but only two or three giants. They gobble up all notice and are not on a continuum with the merely fine players. The same is true of wealth in every society: a very few people have vastly more than anyone else. The same is said to be true of business, where it is widely held that 20 percent of the employees are responsible for 80 percent of the profit.

To document this, Murray quantifies the shape of genius in twenty-one intellectual fields, including astronomy, music, mathematics, Eastern and Western philosophy, painting, and literature. In every one of these fields, the citation rate of the leading figures is not remotely bell shaped; rather there are only two or three giants who grab the lion’s share of glory and influence. One in Chinese philosophy: Confucius. Two in technology: James Watt and Thomas Edison. Two in Western music: Beethoven and Mozart. One in Western literature: Shakespeare.

Once described, your reaction, like mine, is probably “Of course, I already knew this—at least intuitively.” But why should this be, and why should it be universal across endeavors?

The shape of genius—with the top performers outdistancing the average excellent performer by a much greater margin than they would in bell-shaped distributions—follows from multiplying, rather than adding, the underlying causes of genius. William Shockley, the Nobel laureate who invented the transistor, found this pattern in the publication of scientific papers: a very few people published many papers, but most scientists published none or only one. Shockley wrote:

For example, consider the factors that may be involved in publishing a scientific paper. A partial listing, not in order of importance, might be: (1) ability to think of a good problem, (2) ability to work on it, (3) ability to recognize a worthwhile result, (4) ability to make a decision as to when to stop and write up the results, (5) ability to write adequately, (6) ability to profit constructively from criticism, (7) determination to submit the paper to a journal, (8) persistence in making changes (if necessary as a result of journal action)… Now if one man exceeds another by 50 per cent in each of the eight factors, his productivity will be larger by a factor of 25. (click here)

This is the underlying rationale for GRIT, the never-yielding form of self-discipline. Very high effort is caused by a personality characteristic of extreme persistence. The more GRIT you have, the more time you spend on the task, and all those hours don’t just add to whatever innate skill you have; they multiply your progress to the goal. So Angela developed a test for GRIT. Take the GRIT test now and give it to your kids as well.

Please respond to the following eight items using the following scale:

 

1 = Not like me at all, 2 = Not much like me,

3 = Somewhat like me, 4 = Mostly like me,

5 = Very much like me

1. New ideas and projects sometimes distract me from old ones.^* _____________
2. Setbacks don’t discourage me. _____________
3. I have been obsessed with a certain idea or project for a short time but later lost interest.^* _____________
4. I am a hard worker. _____________
5. I often set a goal but later choose to pursue a different one.^* _____________
6. I have difficulty maintaining my focus on projects that take more than a few months to complete.^* _____________
7. I finish whatever I begin. _____________
8. I am diligent. _____________

To get your score

1. Add your score on statements 2, 4, 7 and 8.
2. Then add items 1, 3, 5, and 6 and subtract that total from 24.
3. Then add the two steps together and divide by 8.

Here are the norms by sex:

 

DECILE (TENTHS)	MALE (N = 4,169)	FEMALE (N = 6,972)
1st	2.50	2.50
2nd	2.83	2.88
3rd	3.06	3.13
4th	3.25	3.25
5th	3.38	3.50
6th	3.54	3.63
7th	3.75	3.79
8th	3.92	4.00
9th	4.21	4.25
10th	5.00	5.00
Mean, SD (standard deviation)	3.37, 0.66	3.43, 0.68

 

What has Angela discovered about GRIT? The more education, the more GRIT. Not surprising, but which comes first? Does more education produce more GRIT, or more likely, do gritty people persevere through many failures and humiliations and so go on to get more education? This is still unknown. More surprising is the fact that, controlling for education, older people have more GRIT than younger people, with those over sixty-five having much more than any other age group.

GRIT’S BENEFITS

Grade Point Average

The GRIT test was taken by 139 of Penn’s psychology majors. We knew their SAT scores, which are a good estimate of IQ. Angela followed them through their studies and looked at the grades they went on to get. High SATs predicted high grades—that is indeed the only proven benefit of having high SATs—and high GRIT also predicted high grades. Importantly, holding SATs constant, higher GRIT continued to predict higher grades. At every level of SAT, the grittier students got better grades than the rest, and students with lower SATs tended to have more GRIT.

West Point

In July 2004, 1,218 plebes entering the U.S. Military Academy took the GRIT test along with a mountain of other tests. The army does numbers and is very serious about trying to predict accomplishment through psychological tests. Interestingly, GRIT seemed like a unique test, since it did not correlate with the “whole candidate score”: the sum of SATs, leadership potential ratings, and physical aptitude. GRIT predicted which new arrivals completed the grueling summer training (which used to be called “beast barracks”) and which ones dropped out more accurately than any other test and better than all the other tests combined. GRIT also predicted grade point average and military performance scores over the first year, but so did the more traditional tests, and GRIT did not outpredict them. Indeed, a brief self-control scale (a less extreme version of GRIT) outpredicted GRIT for grade point average. Angela replicated this study at West Point in 2006 and then went on to find that GRIT predicted retention in the U.S. Special Forces as well as sales in real estate.

The National Spelling Bee

The Scripps National Spelling Bee involves thousands of kids ages seven to fifteen from all over the world. In 2005, 273 made it to the grueling finals held in Washington, and Angela gave an IQ test and the GRIT test to a large subsample. She also recorded how much time they spent studying the spelling of obscure words. GRIT predicted making it to the final round, while self-control did not. Verbal IQ, the component of IQ about words, also predicted making it to the final round. Finalists who were well above the average GRIT matched for age and IQ had a 21 percent edge in advancing to the final rounds. The statistics showed that gritty finalists outperformed the rest, at least in part because they spent more time studying words. Angela replicated this again the next year and this time found that the additional time practicing accounted for the entire edge GRIT confers.

Building the Elements of Success

Let’s review the elements of achievement that have emerged from the theory that achievement = skill × effort:

1. Fast. The sheer speed of thought about a task reflects how much of that task is on automatic; how much skill or knowledge relevant to the task a person has.

2. Slow. Unlike underlying skill or knowledge, the executive functions of planning, checking your work, calling up memories, and creativity are slow processes. The more knowledge and skill you have (acquired earlier by speed and deliberate practice), the more time you have left over to use your slow processes and, hence, the better the outcomes.

3. Rate of learning. The faster your rate of learning—and this is not the same factor as your sheer speed of thought about the task—the more knowledge you can accumulate for each unit of time that you work on the task.

4. Effort = time on task. The sheer time you spend on the task multiplies how much skill you have in achieving your goal. It also enters into the first factor: the more time spent on the task, the more knowledge and skill that accrete, or “stick” with you. The main character determinants of how much time you devote to the task are your self-discipline and your GRIT.

So if your goal is higher achievement for yourself or your child, what should you do?

Not much is known about how to build the first factor: speeding up thought. What speed accomplishes, however, is knowledge; the faster you are, the more knowledge you acquire and put on automatic for each unit of time you spend practicing. Hence, spending more time on the task will build achievement. So even if your child is not innately gifted, deliberate practice will help enormously by building his knowledge base. Practice, practice, practice.

Building slowness allows space for executive function—planning, remembering, inhibiting impulses, and creativity—to grow. As psychiatrist Dr. Ed Hallowell says to children with attention-deficit/hyperactivity disorder, or ADHD, “You have a Ferrari of a mind, and I’m a brake specialist. I am here to help you learn to apply the brakes.” Meditation and cultivating deliberation—slow talking, slow reading, slow eating, not interrupting—all work. For young children, Tools of the Mind may work. We need to know much more about how to build patience, an unfashionable but critical virtue.

As far as I know, rate of learning—how much is acquired per unit time—is almost never measured in isolation from the amount of knowledge itself. So nothing is known about how to increase your rate of learning.

The real leverage you have for more achievement is more effort. Effort is no more and no less than how much time you practice the task. Time on task acts in two ways to increase achievement: it multiplies existing skill and knowledge, and it also directly increases skill and knowledge. The best news is that effort is very malleable. How much time you devote to a task comes from the exercise of conscious choice—from free will. Choosing to devote time to an endeavor comes from at least two aspects of positive character: self-control and GRIT.

Higher human accomplishment is one of the four components of flourishing and yet another reason that will and character are indispensible objects of the science of positive psychology. My hope (actually, my prediction) is that this decade will see major discoveries in how to increase GRIT and self-control.

Until recently, I thought of positive education as a worthy ideal, but I wondered if it could ever take hold in the real world. Something big has now happened that is an inflection point for positive education, and that is the story of the next two chapters.