58-64feature_duckworth.indd

Penn psychologist Angela Duckworth Gr'06 argues that character— not intelligence, quality of instruction, family situation, or income level—is the crucial determinant of achievement in school.
Now she just has to figure out how to measure character— and influence it BY K E V I N H A RT N ET T
Every year large percentages of American elementary- school students fail to learn basic math skills like how to add fractions with unlike denominators. The situation is even worse among students from the poorest American neighborhoods, despite the fact that from fourth grade on their teachers drill them in these simple steps: find a common denominator; add the numerators; reduce.
There are many explanations for why such a simple proce- dure proves to be so hard to convey. Reformers and policy-makers point to subpar teachers and inadequate principals; to single-parenthood and other demographic drags; to health, nutrition, and the intangible handicaps of poverty.
Assistant Professor of Psychology Angela Duckworth Gr'06 has another explanation. Before she entered graduate school at Penn in 2002 she spent five years teaching math and sci-ence in poor urban neighborhoods across the United States. In that time she concluded that the failure of students to acquire basic skills was not attributable to the difficulty of the material, or to a lack of intelligence, or indeed to any of the factors mentioned above. Her intuition told her that the real problem was character.
"Underachievement among American youth is often blamed on inadequate teachers, boring textbooks, and large class sizes," she wrote in a paper titled "Self-Discipline Outdoes IQ in Predicting Academic Performance in Adolescents," which served as her first-year graduate thesis and was published in Psychological Science in 2005. "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 sac-rifice short-term pleasure for long-term gain, and that programs that build self-discipline may be the royal road to build-ing academic achievement." Effortful practice; persistence though boredom and frustration; gritty deter-mination in pursuit of a long-term goal. In Duckworth's view these are the qual-ities that separate more and less suc-cessful students, and in recent years she's emerged as one of the most influ-ential voices in American education reform, where she argues that cultivating "achievement Charter-school networks like the Knowledge is Power character" in kids may be the last, best way to narrow educa- Program (KIPP) have embraced Duckworth's work.
tional inequality in America.
"Schoolwork is not hard in the way that electromagnetism long-term goals." Today grit is a buzzword in the hallways of is hard. It is hard because it's aversive and not fun to do," charter schools around the country, where teachers, princi- Duckworth, who joined the faculty at Penn in 2007, explains. pals, and deep-pocketed board members have all come to "So the straightforwardness of the material combined with believe that inculcating grittiness in students is every bit as the abject failure of students to learn it made me think there important as building academic skills. must be something besides IQ holding them back. That's maybe Duckworth's view, if correct, would have dramatic implica- more obvious for teachers than it is for policymakers." tions for the way policymakers and educators think about The intuitive appeal and expansive application of Duckworth's student achievement. It also raises provocative questions research has earned her increasing popular recognition (a about the limits of research in the social sciences and the New Yorker profile is in the works) as well as a privileged posi- malleability of human character: Is it possible to design tion at the crossroads of basic research and public policy. This measurements to quantify character with the same preci- past fall U.S. Secretary of Education Arne Duncan invited sion that researchers quantify intelligence? And if so, are Duckworth down to Washington to share her policy recom- self-control and persistence amenable to cultivation, let mendations. There she cautioned Duncan about the useful- alone on the scale of public policy? ness of standardized tests as an accountability tool, arguing that performance on those tests tends to be more a function of native intelligence (IQ) than of how well students are actu-ally learning in their classrooms. She also urged Duncan to Angela Lee Duckworth grew up in Cherry Hill, New Jersey, the daughter of well-educated parents who'd immigrated from China. She's quick to say her mom and dad were throw the full weight of the Department of Education behind not Tiger Parents, though academic success was always assumed. initiatives to use "the hard-fought insights of psychological What was less expected of her, she says, was that she'd throw science" to improve the way schools teach students.
herself into community service. "When I was in high school I was Duckworth's experience as a classroom teacher has also sort of spontaneously and not very reflectively drawn to public helped her build strong ties in the education reform com- service activities," she says. "I don't know where that impulse munity, where leaders like Dave Levin, the co-founder of the came from. You could do some retrospective reconstruction, but KIPP (Knowledge is Power Program) charter-school network, that is always a dangerous game to play." consider her a kindred spirit and are using her work to When Duckworth was 18 she went to Harvard, where she develop strategies for teaching achievement character to majored in neurobiology but continued to perplex her father low-income kids. Duckworth is best known for the study of (who is a color chemist at DuPont) by devoting much of her "grit," which she defines as "perseverance and passion for time to leadership roles in several community service orga- nizations. "I don't think in Chinese culture there is as much For most of her time as a teacher she assumed she'd apply her of a tradition of helping anonymous strangers of a different insights about student achievement by opening her own char- race," she says. "My dad was like, ‘ You have a science degree ter school, which would have allowed her to stay in public edu- from Harvard but instead you want to spend your time help- cation while also giving her significant discretion over exactly ing poor black kids?'" how and what students were taught. But after five years in the Following graduation Duckworth spent two years found- classroom, ending in a stint teaching high school science at ing a summer program for disadvantaged kids and then Mastery Charter in West Philadelphia, she concluded that char- went to Oxford on a Marshall Scholarship. After returning ter schools were not the answer to the achievement gap. from Oxford she consulted with McKinsey for a year and "I looked around at these charter schools and it seemed to spent another year as the chief operating officer of a web me intuitively they weren't the way to reform education," she startup called Great Schools that allows parents to compare says. "I saw these charter schools writing their own curricula, public schools. But the bulk of her time over the next seven creating their own HR departments, and it seemed to me years was spent teaching math and science at public schools intuitively that the diseconomies of scale were working in New York, San Francisco, and Philadelphia.
against them." (Today she says her critique was wrong, owing "It seemed to me that if I was going to work on issues of to "a lack of imagination" when it came to foreseeing the equity I should start earlier in the life course rather than development of national charter organizations like KIPP and later," she says, explaining why she was drawn to the class- Mastery, which provide their network schools with the effi- room. "The earlier you start the bigger bang for your buck ciencies of scale that early charters lacked.) you get in terms of closing the gap between the privileged Duckworth wasn't going to open her own school, but she still and the non-privileged. I also just enjoyed working with wanted to make a bigger impact on public education than she young people, so that kind of led naturally to teaching." could as a teacher—at first, she just didn't know how. "I thought Duckworth jokes that the job-hopping she did in her twen- about the biggest problem that needs to be solved in K-12 educa- ties was a case study in "how not to be gritty," but it seems tion and then I listed out all the things I'm good at: I like to write, more a function of the intensity and dynamism of her person- I like analysis, I like math, I like to think hard about problems," ality. In the course of reporting this article I heard colleagues she says. "So I sort of put the Venn diagrams together and in a call Duckworth the most extroverted person, the quickest very top-down deductive way I concluded I should go into psy- learner, and the fastest thinker (and talker) they'd ever met. chology and become a researcher in order to understand these On the day I visited she had a half-dozen bubble gum con- character competencies, and then go back into these schools and tainers on her desk, suggesting an atmosphere of restless help them solve their achievement problems." activity and a need to replenish the saliva that's lost through One night in July 2002 Duckworth was up late with her infant such rapid-fire speech. She also uses expletives in a way that daughter (she and her husband, the president of a Philadelphia- might impress even high-powered cursers like Rahm Emanuel. area real-estate investment fund, now have two girls), and In the course of a 90-minute conversation she called a princi- researching psychology programs online when she came across pal she knew "an asshole," described the opinion of a leading the website of Martin Seligman Gr'67, the Zellerbach Family education foundation as "fucking idiotic," and did a spot-on Professor of Psychology and director of Penn's Positive impression of a teenager with attitude when explaining the Psychology Center (not to mention the founder of the discipline). challenge of conducting experiments with adolescents: "When Duckworth, who'd never taken a psychology class before, was you pay adults they always work harder but sometimes in such a neophyte in the field that she didn't recognize Seligman's schools when I've done experiments with monetary incen- name. "I didn't know he was famous. I was like, he writes very tives there's this like adolescent ‘fuck you' response. They'll well, I like his website, so I emailed him," she says. be like ‘Oh, you really want me to do well on this test? Fuck As it happened, Seligman was up late too, playing bridge you, I'm going to do exactly the opposite.'" online ["Passion Play," Mar Apr 2011]. He replied to Duckworth's Duckworth also has a degree of entrepreneurial energy that, email within minutes, inviting her to attend a research meet- at first blush, makes her an odd fit for the academy. I asked ing at his house the next day—and ended up being so bowled her whether she sees herself more as a reformer, like Dave over by Duckworth's super-charged demeanor that he would Levin, the KIPP founder—driven by the desire to achieve spe- ultimately convince his colleagues to throw the admissions cific outcomes—or as a scientist—dedicated to asking ques- timeline out the window and allow Duckworth to join the pro- tions and following prescribed methods for answering them.
gram that September. All things considered, it wasn't a hard "I think that Dave and I have always had this passionate sell. "Angela was fast, about as fast mentally as it is possible commitment to children combined with an incredible for a human being to be," Seligman writes in his latest book, amount of energy and optimism about doing," Duckworth Flourish: A Visionary Understanding of Happiness and Well- says. "My dad used to say to me, ‘There are thinkers and being. "She blew us away in the interview. In violation of prec- there are doers.' Very few people are both. And I think to the edent, the admissions committee gave in and accepted her." extent that I'm a professor I'm a thinker and it's my duty to For her part, Duckworth was drawn to the Positive Psychology analyze things and see if I can figure out how the world Center because of the way it encourages students to think works. But Dave and I are both very much doers. We share a about the real-world implications of their research. "Marty is kind of boldness, an attitude of just try it, just do it, and only interested in questions with significant relevance to don't just sit on your hands and think all day." people's well being," Duckworth says. "Marty is a basic scien- tist, of course, but he's always been someone with one foot in After failing to measure persistence Duckworth shifted to the world beyond the lab. He's always cared about how psy- self-control, which researchers had been studying since 1972 chology actually changes people's lives within his lifetime for when the eminent psychologist Walter Mischel conducted his the better. I came in to graduate school very much in that famous "marshmallow test." Mischel presented pre-kindergar- view, and it only got reinforced while I was there." teners with a marshmallow but told them they could have two if Achievement—the main focus of Duckworth's research—is they waited 15 minutes to eat the first. In a conclusion that hov- the last of the five key elements in positive psychology's tax- ers over middle-class parenting across America, Mischel found onomy of well-being, which goes by the acronym PERMA. (The that kids who were able to hold out for the second marshmallow others are Positive Emotion, Engagement, Relationships, and tended to have higher SAT scores years later.
Meaning.) For decades psychologists studying achievement Duckworth conducted her own version of Mischel's experi- focused almost exclusively on intelligence. As Duckworth saw ment with students at the Julia R. Masterman Laboratory and it the problem with this single-minded emphasis was two-fold. Demonstration School, Philadelphia's top magnet school, and For one, IQ scores didn't explain everything about why some drew on additional measurement techniques to make her individuals achieve more than others. In fact, they were not results more robust. In the fall of 2002 she offered 140 eighth even particularly good at predicting something as basic as the graders a choice between receiving $1 immediately or $2 a amount of education adolescent-aged kids would go on to week later, and also administered self-control questionnaires attain later in life.
to the students, their parents, and their teachers. She com- The second problem was more practical: IQ may be easy to bined her survey and experimental data to create a "self-con- measure, but it's hard to change. IQ tests given as early as trol index," which she used to anticipate how well students kindergarten are highly predictive of adult intelligence, mean- would fare on their final report cards. When grades came out ing that if boosting intelligence was the only way to boost that spring she found that self-control and GPA were not only academic achievement, reformers were not going to get far.
strongly correlated—her self-control index was twice as good The story was very different with personality. While IQ sta- at predicting academic performance as IQ scores. bilizes before kids even learn to read, many psychologists Following the Masterman study, Duckworth turned to what including Duckworth point to longitudinal survey data to con- has become her signature topic—grit. Grit is a nebulous con- clude that personality doesn't become similarly fixed until at cept compared to self-control, and the way people pursue least age 50. Along the way average levels of personality traits long-term goals is hard to measure experimentally. So change (most people become more conscientious as they grow instead of a lab task Duckworth developed the "Grit Scale," a older), as do rank-order levels, meaning that the most consci- 12-item questionnaire that asks respondents to rate them- entious 10-year-olds are not necessarily the most conscien- selves on statements like "Setbacks don't discourage me" tious adults four decades later. To Duckworth, this fluidity and "I have difficulty maintaining my focus on projects that suggested a tantalizing possibility: If personality evolves so take more than a few months to complete." much as people get older, why shouldn't schools be able to Duckworth wanted to test the Grit Scale in high-achieving influence the direction in which it changes? populations, to see if it could reveal distinctions even within Before Duckworth can hope to modify personality she groups where everyone was talented. In 2004 she adminis- needs to know how to measure it, which is no small task. In tered the Grit Scale to 1,200 incoming cadets at West Point, fact, according to Seligman, measurement challenges are just before they began "Beast Barracks," the academy's the primary reason that researchers have long shied away intensive summer training program that every year leads 5 from studying personality. percent of admitted freshmen to drop out.
"There's a hegemony of silence in science, things you don't work Admission to West Point is determined in large part by a on because they are too fuzzy," Seligman says. "So, while the impor- "Whole Candidate Score," a weighted index comprised of tance of self-control, self-discipline, grit may be obvious once we've variables like SAT score, class rank, and performance on the said it, that doesn't mean that they become eligible as scientific Army's Physical Aptitude Exam. Duckworth found, however, endeavors until a creative person like Angela comes along and that cadets with the highest scores on the Grit Scale were 60 says, ‘Oh come on, we can measure these things!'" percent more likely to make it through Beast Barracks than Duckworth is naturally optimistic about the potential for cadets of merely average grittiness. What's more, her Grit human ingenuity to solve social problems. She likes to say, Scale was nearly four times as good at predicting which "Rather than curse the darkness, light a candle," but in her very cadets would drop out as any of the Whole Candidate Score first research project as a graduate student at the PPC the dark- indicators that the Army had spent years refining. ness won. Her years in the classroom had convinced her that Despite this and other successes administering the Grit persistence was essential for academic achievement, so she Scale (she's used it to study Scripps National Spelling Bee designed a study in which students were asked to find a pattern Contestants and Ivy League college students, among oth- in what was (unknown to them) a series of non-repeating digits. ers), Duckworth acknowledges that questionnaires have The problem, though, was that no one gave up within the allot- significant pitfalls as a research tool: answers can be faked ted time. "I found, as has been found by many psychologists, and respondents are often biased when evaluating people that any good lab experiment has to end within 60 minutes," they know. Even more significant, people don't have a good Duckworth says. "But I couldn't find a task that was so frustrat- intuitive sense of the appropriate scale to use when assess- ing that people would give up within that time." ing their own personalities. "Let's say I ask you to evaluate yourself on the question, ‘I In the Gates project, Duckworth and a group of high-profile am a hard worker.' What standard would you use?" Duckworth collaborators will conduct several studies that attempt to quan- says. She ran into this problem recently while studying delay tify the personality factors that distinguish college graduates of gratification in students in the US and Taiwan. On objective from college dropouts with similar demographic and academic measures US students showed much less self-control, but on backgrounds. Their goal will be to "provide new insight into surveys they gave themselves higher marks than Taiwanese student factors that predict college persistence and develop students gave themselves.
strategies to cultivate them via school-based interventions." The success of Duckworth's research will hinge ultimately on Duckworth's part of the study is based on the work of Anders whether she and her colleagues can devise measurement tools Ericsson, the Florida State University psychologist whose that produce more replicable and precise results than the ones work on the personality characteristics and training habits of they are using today. Or, as she puts it, "If we cannot figure out truly exceptional performers was popularized in Malcolm how to measure these characteristics in Gladwell's recent best-seller Outliers. some kind of reasonable timeframe, with Duckworth and her research team will "If you wanted to help
some kind of objectivity, the research will conduct an in-depth analysis of a small grind to a halt because we can't measure kids, it's not immediately number of high school students from
what it is that we want to actually study." minority backgrounds who demonstrat- obvious how you'd go
Measurement innovations are at the ed superior academic growth from ninth heart of two large-scale projects that about changing poverty. to 12th grade. These superstar students
Duckworth is launching this year: The will be compared against peers who So the direction I'm
first is a three-year-long study of self- entered high school at similar levels of control funded by a grant from the John achievement but made significantly less more excited about is
Templeton Foundation. Students rang- academic progress en route to gradua- ing from pre-kindergarteners to college the effects that schools
tion. One technique she'll be using to seniors will be asked to complete boring make the comparisons is an innovative can have on kids.
software tasks while resisting the temp- new measure called an "emote-aloud tation to divert their attention to more protocol" in which participants are We might not be able
entertaining pursuits. instructed to narrate their feelings as On the day I visited Duckworth she was to make a family richer, they perform a demanding task.
holding a conference call where the proj- "We're going to be applying the same but maybe we can make
ect team was debating whether to have methodology that Ericsson used to kids use iPads or computers in the exper- their kids grittier or
understand Olympic athletes to under- iment. (The concern was that very young stand kids in school," Duckworth says. kids wouldn't be dexterous enough with "We're going to put the kids under the a mouse; in the end the team decided on microscope to understand what high- iPads and within a couple weeks Duckworth had rounded up an achieving kids do differently. We'll follow them through col- additional $10,000 of funding to pay for them). lege and see whether the habits we identify actually have a "So kids are going to be doing these boring tasks on iPads payoff when they're in a new environment. We want to know while trying to resist the temptation of switching over to play what achievement personality looks like in kids and how we Angry Birds," Duckworth says, referring to the notoriously get more of it—what do you do, what do the parents do, what addictive online game. "We'll measure how well kids exert do the schools do." self-control in the face of temptation to take immediate grati-fication. Once we have these measurements established the question will be identifying strategies kids can adopt that will THE most concerted effort to date to implement Duckworth's research is taking place at KIPP make it easier for them to do well in these tasks." Infinity Middle School in Harlem, New York. The KIPP net- The second study is on college persistence and will be car- work of charter schools was co-founded by Dave Levin and ried out over the next two years with a $1.8 million grant Mike Feinberg C'91 ["Alumni Profiles," Nov Dec 2000] in from the Gates Foundation. College persistence has become Houston in 1994, when they were both working as fifth- a hot topic recently, as evidence has emerged showing that grade teachers through Teach for America. Since then KIPP even when low-income students catch up academically with has expanded to 109 schools in 30 low-income regions their middle-class peers, they still end up dropping out of around the country, and has become one of the most promi- college at disproportionately high rates. Most recently, the nent education-reform organizations in the country.
KIPP Foundation reported that while its intensive approach The KIPP model is based on the premise that a high-per- to academic instruction succeeded in getting 80 percent of forming school by itself can overcome the disadvantage that its students into college over the last decade, only 33 percent poor, typically minority students face in many other areas of of those students ended up with a college diploma. That their lives. The basic tools of a KIPP school are an extended number is above the 8 percent of low-income students school day (often running from 7:30 a.m. to 5 p.m. or later), a nationwide who complete college, but it still falls well short longer school year, intensive math and literacy instruction, of college graduation rates for middle-class students. and a pervasive focus on the goal of graduating from college. From the beginning, character development has been an prepared," "Allows others to speak without interruption," and essential part of a KIPP education, as captured by the orga- "Remembers and follows directions." nization's ubiquitous slogan: "Work Hard. Be Nice." Over the Duckworth is aware that measuring character to two deci- last five years Levin, who serves on the KIPP board of direc- mal places on a report card could be perceived as unduly tors and is superintendent of KIPP's eight schools in the harsh, particularly if it's seen to overlook the role that pov- New York area, has teamed up with Duckworth to formalize erty plays in depressing achievement among low-income the way KIPP NYC teaches character—to measure and moni- students. But in her view character isn't purely innate— tor it, and institute strategies for enhancing it.
instead, she argues, it's just as influenced by environmental The cornerstone of the initiative is the "KIPP Character Report forces as things like reading scores and high school gradu- Card," which teachers use to assess students on character traits ation rates, which most people feel entirely comfortable that KIPP considers intrinsic to high achievement. The idea for quantifying and evaluating.
the character report cards originated in a 2005 meeting at Penn "One of the problems with the word character is that it that included Seligman, Levin, Duckworth, and Christopher carries a lot of baggage," Duckworth says. "People some- Peterson, a psychologist at the University of Michigan.
times think that emphasizing character means not empha- At the time of the meeting Peterson and Seligman had just sizing environmental conditions like growing up in poverty finished collaboration on an 800-page tome called Character or not having good role models. But I think it's a false dis- Strengths and Virtues: A Handbook and Classification. To tinction, because your character is influenced by how you write the book they had scoured essential texts from cul- grew up—it's not like there's character on one side and envi- tures throughout history, looking for character strengths ronmental forces on the other. Given that, if you wanted to that have been considered building blocks of the good life no help kids, it's not immediately obvious how you'd go about matter where or who you are. In the end they came up with changing poverty. So the direction I'm more excited about is 24, ranging from bravery to prudence to self-control. the effects that schools can have on kids. We might not be Peterson and Seligman's book provided Levin with a able to make a family richer, but maybe we can make their framework for thinking about character in a more system- kids grittier or more self-controlled." atic way, and in Duckworth he found the perfect person to There is scattered evidence showing that programmatic help him translate that framework into an assessment tool school-based character interventions work. Some of the he could use at KIPP. "Angela is one of the elite people in the most frequently cited interventions include Tools of the country to combine a deep understanding of K-12 education Mind, a preschool program that helps students develop self- with the highest credentials of a researcher," Levin says. "It regulation tools, and the Chicago School Readiness Project, was a natural fit for us to work together." which trains preschool teachers on how to instruct kids in Levin and Duckworth's first step was to boil the 24 traits self-control. Overall, though, the research in this area has down to those with particular relevance for school. They been limited. It's still unknown whether personality is like a removed traits like modesty, spirituality, and fairness, and person's height—which is measurable, but not modifiable—or settled on a list of seven that seemed particularly essential for whether it's more like blood pressure or cholesterol levels, high academic achievement: zest, grit, self-control, curiosity, which can be measured and modified, and which can be social intelligence, gratitude, and optimism. (Love actually influenced at a population level by public health initiatives.
made the initial cut, but, Duckworth says, "Dave didn't want to At the end of my conversation with Seligman I asked him have to tell a parent, ‘Your kid is low on love,'" so they swapped whether he thinks we'll see the day when schools are teaching it out for curiosity.) kids grit and self-control alongside phonics and fractions. I Once the seven traits had been determined, the next step was expected the father of positive psychology to be bullish, but to figure out how to measure them—to define, for example, what he was surprisingly skeptical. "I think Angela has made some optimism or zest looks like in practice. The criteria for measur- progress in this area," he said. "But it is interesting to me that ing each trait didn't have to produce results that concurred with for 3,000 years at least teachers have been trying to get more some absolute value, because there is no truly objective defini- self-discipline out of kids without figuring out how. So for me tion or measure of something like zest. Instead, Levin and the modification of grit and self-discipline are still hopes and Duckworth's goal was to agree on criteria that matched their promissory notes as opposed to fact." general understanding of the character traits, that were easy For Duckworth, however, the challenge of her research ques- for teachers to observe, and that produced results which corre- tion is part of its appeal. She spent the first decade of her sponded roughly with anecdotal evaluations of which students professional life unsure of how to apply her abundant talent. had more or less of a given trait.
Now she no longer has any doubts. "I have complete convic- In the final KIPP Character Report Card each trait is broken tion that this is an incredibly important scientific question," down into two to four indicators on which students are given she says. "If we can figure out the science of behavior and scores from 1-5. The indicators for optimism are "Gets over behavior change, if we can figure out what is motivation and frustrations and setbacks quickly" and "Believes that effort how to motivate people, what is frustration and how do we will improve his or her future." One of the indicators for zest is manage it, what is temptation and why do people succumb to the relatively easy to quantify "Actively participates," while it—that to me would be akin to the semiconductor."◆ another is the less tangible "Invigorates others." Indicators for Kevin Hartnett, a former teacher, is a freelance writer living in Ann Arbor, self-control are more concrete and include: "Comes to class Michigan. A collection of his work can be found at GrowingSideways.net.

Source: https://sta-faculty-resources.wikispaces.com/file/view/Characters+Content.pdf