The Education Gadfly Weekly

Intel’s recent announcement that it will cease sponsoring and underwriting the prestigious Science Talent Search, which it took over from Westinghouse in 1998, is another nail in the coffin of gifted education in the United States.

Unlike many European and Asian countries, which are awash in academic competitions, Olympiads, and other status-laden contests that bright students vie to win, American K?12 education has relatively few that anyone notices. There is, of course, the National Spelling Bee, which Scripps has valiantly stuck with since 1941. But spelling bees are for middle schoolers. The big deal for high schoolers, especially those with a bent toward STEM subjects, has long been the Science Talent Search, which President George H. W. Bush called the “Super Bowl of science.”

Intel’s turnabout surprised former CEO Craig Barrett and disheartened many of us who care about both STEM and gifted education. It’s another sign of America’s inattention to its high-ability learners, especially those from disadvantaged circumstances. That neglect is what triggered the publication of our new book, Failing Our Brightest Kids: The Global Challenge of Educating High-Ability Students. All sorts of data—from the National Assessment of Educational Progress, from research studies like the 2011 Fordham Institute report “Do High Flyers Maintain Their Altitude? Performance Trends of Top Students,” and from elsewhere—have shown that high-achievers made smaller gains in the No Child Left Behind (NCLB) era than did low-achievers. Policy efforts that raised the floor and eased the achievement gap did so at the expense of strong students, who were already nudging the ceiling. Under NCLB, schools and teachers had scant incentive to work hard with kids who were already “proficient.” And so they didn’t—especially in places full of poor and minority kids, many of whom needed extra help to become proficient.

At the same time, data from international measures such as the Program for International Student Assessment (PISA) and the Trends in International Mathematics and Science Study (TIMSS) make painfully clear that the United States isn’t pushing nearly as many of its young people to higher scores as are our competitor nations. These include not just the “Asian tigers,” but also Canada and a number of European countries (see Figure 1). Although we fare better in the early grades, by the time kids are fifteen (when PISA tests them), our high-end results are anything but high.

Figure 1. Percentage of high scorers on PISA, math and science, 2012

America does an especially bad job of getting poor kids into the high-scoring ranks. The Organisation for Economic Co-operation and Development reports PISA results by socioeconomic quartiles linked to parent occupation and education, family wealth, home educational resources, and more. This “ESCS” index is controversial. Yet the data shown in Figure 2 are both instructive and depressing. Of the twelve countries we studied, the United States surpasses only Hungary in getting lowest-quartile youngsters into the top-scoring echelons. Less than 3 percent of American students in the bottom ESCS quartile reach PISA’s top tiers in math, yet 21 percent of disadvantaged Singaporeans achieve this. And when we compare these kids to those from the top quartile, we see that a high-ESCS test-taker in the United States is eight times likelier than a low-ESCS student to be a top scorer in math. In Canada, that ratio is four to one, and in Korea and Finland, it is three to one. (Comparisons based on parents’ education show very similar gaps.)

Figure 2. Percentage of students at levels 5 or 6, PISA math, ESCS top and bottom quartiles, 2012

America’s poor kids can do so much better than most of them are doing in the today. Consider the recently publicized example of Abdisamad Adan, a Harvard freshman from Somaliland, about as poor a place as can be found on the globe. Abdisamad grew up with eighteen brothers and sisters in a small home without electricity. He caught a break when Jonathan Starr, a former hedge fund manager, started the Abaarso School of Science and Technology, a boarding school in Somaliland meant to serve the country’s brightest boys and girls. Abdisamad thrived there. He gained admission to a college preparatory school in New York—and then, impressively, to Harvard. Talk about upward mobility—this amazing story is the result of ability plus a great school designed to develop the ambitions and intellects of promising students.

Then there’s Lin-Manuel Miranda, who wrote the book, music, and lyrics for Hamilton, the Broadway smash of the moment. He also recently won one of this year’s twenty-four MacArthur Foundation “genius” awards. When Miranda’s father moved to New York as a young man, he spoke no English—but when Miranda was six, he was accepted into Hunter College Elementary School, a public school for intellectually gifted children that seeks to identify able kids from all parts of town. His first theatrical successes occurred within the Hunter College system, which cultivated his genius and created his opportunity.

Looking beyond U.S. borders, most of Failing Our Brightest Kids reports what we learned about how eleven other countries educate gifted students, based on conventional research, site visits, and on-the-ground reports. We parse the data and took advantage of expert analysis by others—especially Eric Hanushek, Paul Peterson, and Ludger Woessman. Here’s some of what we found:

• Although countries with strong overall achievement also tend to have many top scorers, the United States cannot assume that raising the water level will do enough for the gifted education boat; it’s far too small to contain all who belong in it.
• Culture matters. When parents value education and push their kids, it makes a difference. Where competition is appreciated rather than shunned, more kids push themselves. (That’s what’s been so great—and rare—about the Science Talent Search.)
• Policy matters too. No Child Left Behind makes it essentially impossible for states to test kids on “out of grade level” material, a policy intended to prevent the tests from being dumbed down for low-achievers. The notion is benevolent, sure, but this policy also makes it impossible for states to truly track academic gains by high-achievers. A provision in the House and Senate reauthorization bills would solve this problem—at least, it would void the current ban—but this is not yet law.
• Waiting for high school to beef up the offerings for high-ability kids is a mistake, particularly for smart students from disadvantaged circumstances who depend on the school system from day one to help them realize their potential.
• “Differentiated instruction” sounds great, but in the end, it’s no substitute for separate learning opportunities, supplements, acceleration, and enrichment programs. Rare is the teacher who can do right by her ablest pupils at the same time she provides slower learners the attention that they need.

After digesting what we learned, we issued a number of recommendations for how we could do a better job.

Perhaps most important, we need to get this topic back onto the policy agenda and rekindle the debate about a society that prizes excellence as well as equity.

We should conduct universal ability screening. Almost every U.S. public school student now takes state exams in (at least) reading and math from third grade on. Let’s identify, say, the 5 percent with the highest scores in a school, district, or state in third or fourth grade, much as Singapore and Western Australia do. Instead of using IQ or aptitude tests like they do, however, we should use existing—and universal—achievement metrics, then supplement their ranks with kids whose promise is spotted by their teachers.

Student progress should be based on mastery. The single best thing our education system could do for high-ability students (and everybody else) would be to enable them to move through their curricula at their own pace. Instead of age-based grade levels—placing all eleven-year-olds in fifth grade and holding them to the same performance standards—let students proceed on the basis of mastery, one unit or module at a time, subject by subject, with no obligation for all to move at the same rate.

Besides allowing for acceleration, schools and districts should encourage flexible ability grouping within classrooms, multi-age and multi-grade groupings, the use of technology for “blended learning,” and whatever else they can to facilitate the educational individualization that we already demand for disabled children.

We should not be afraid of separate classrooms, after-school and summer programs, and even separate schools for high-achievers. Places like the Bronx High School of Science and Virginia’s Thomas Jefferson High School for Science and Technology are educational treasures. But we don’t have enough of them—and we almost never provide their equivalent for elementary and middle school students, which would do much to launch more high-ability poor kids on a trajectory to success.

It’s unfashionable, even politically incorrect, to push for better education for smart kids. One can easily be labeled elitist and swiftly admonished that attention and resources should focus laser-like on the “truly needy.” But plenty of smart kids are also needy, especially those without middle class families and suburban guidance counselors to help them navigate the turbulent waters of American K?12 education. And the country itself is needy: It’s in need of more top-flight scientists, inventors, and entrepreneurs to ensure its future security, prosperity, and competitiveness—which is exactly why Intel’s suspension of its Science Talent Search sponsorship is so shameful.

Former University of Michigan president Mary Sue Coleman was a finalist in that competition in 1961. “It was the Sputnik generation, when America was competing with Russia to get into space,” she recently told the International Business Times. “It was a national obsession. People in school cheered us on like we were star athletes. I got letters from the heads of corporations.”

Shame on Intel for pulling the plug and making matters worse. But shame on all of us for neglecting millions of young people who deserve better—and who have the potential to do great things for their country as well as for themselves.

Editor’s note: This post originally appeared in a slightly different form at Education Next.

Fuse/Thinkstock

The next batch of results from the National Assessment of Educational Progress (NAEP) is due on October 28. Pundits of all persuasions are gearing up to jump on the news and promote their spin. But that’s for amateurs, I say. Why wait for regular old mis-NAEP-ery when you can practice pre-NAEP-ery?

Rumors abound that the news is going to be bad, with scores down nationally and in a bunch of states. That will be used as fodder to attack Common Core, teacher evaluations, charter schools, or whatever else you happen not to like that’s prominent in today’s education policy conversation.

But let me suggest that journalists and editorialists consider the most likely explanation: It’s the economy, stupid. While those of us in education reform are working hard to make sure that demography does not equal destiny, we must also acknowledge the strong link between students’ socioeconomic status and their academic achievement (a link that some amazing schools are weakening).

In fact, the last time we saw national declines in NAEP scores was in the aftermath of the 1990 recession. That was particularly the case for reading. It makes sense—when families are hurting financially, it’s harder for students to focus on learning.

And let’s be clear: The American family has gotten hammered in recent years. Adjusted for inflation, the national median income for families with children dropped 6 percent from 2006 to 2014. In many states, it was much worse:

Change in Median Income of Families with Children: 2006–2014

Note: Data are from Kids Count; I then used the Bureau of Labor Statistics’s inflation adjuster calculator to determine inflation-adjusted income.

So if I were a betting man, I’d wager that the states with big declines in median income are going to be the ones showing lower NAEP scores this time around. And the handful of places that have weathered the recession—or even thrived—are going to come out looking pretty good. (Washington, D.C.: Here’s looking at you, kid.)

We’ll know in a week if I’m right. 

Nastco/iStock/Thinkstock

• When the Los Angeles Board of Education voted to fire fifth-grade teacher Rafe Esquith last week over allegations of sexual misconduct, its members were doubtless aware of the potential for blowback. Esquith, whose myriad awards and world-renowned classroom productions of Shakespeare have made him the district’s most recognizable employee, can count influential friends both inside and outside the school system. And the circumstances of his termination—decided by a closed-door tribunal following an extensive investigation into his private life—do not argue in its favor. Now Esquith’s attorneys have filed a $1 billion class-action lawsuit against the district on behalf of thousands of teachers, claiming that they were targeted for dismissal because of their age and pending retirement windfalls. The accusations lodged against Esquith are ill defined, but serious. If accurate, they are surely serious enough to merit public examination of his record and methods. But if, as some prominent supporters claim, the investigation was a panicked overcorrection in response to earlier scandals, this story could become more bizarre and tragic than it has already. Either way, this is a disciplinary process in dire need of greater transparency.
• The saddest and most illuminating thing you’ll read all week is this long-form Washington Post profile of the plight of impoverished high school graduates in the Deep South. Reporter Chico Harlan spent months with ninteteen-year-old Jadareous Davis, a recent graduate trapped by the social and institutional pathologies of economically depressed Drew, Mississippi. Davis, who squeaked through one of the worst high schools in the state, watches in horror as the pride of his commencement ceremony gives way to unemployment and family dysfunction. He finds a modicum of hope in vocational training, but odds this steep should never weigh against a kid fresh from his senior prom. The United States as a whole, and this region most especially, must come to grips with its inequities and embrace a full program of upward mobility—before more young people are doomed to the same hell.
• The lecture is dead, long live the lecture! Over the weekend, Molly Worthen, an assistant professor at UNC-Chapel Hill, penned an op-ed for the New York Times’s Sunday Review defending the lecture—the beleaguered teaching method we all skipped or slept through in our riotous college days. The favored alternative is “active learning,” which is so strongly backed that educators like Harvard physicist Eric Mazur have opined that lectures might even be “unethical.” And in the hard sciences, Mazur may have a point. Mixing chemicals in the orgo lab and observing kinematics in the physics lab really does deepen understanding and raise scores. But in the humanities, Worthen argues that lectures—in which experts build arguments instead of reciting encyclopedia entries—are still the best way to educate students, especially when paired with smaller discussion sections. “Absorbing a long, complex argument is hard work, requiring students to synthesize, organize, and react as they listen,” she argues. And in the 140-character, six-second age, students have little experience with such prolonged exchanges. They ought to gain some, and maybe lectures can help.

The French mathematician and philosopher Blaise Pascal famously posited that whether or not one believes in God, it behooves us to behave as if he exists. What have you got to lose? If you’re right, you wind up heaven and spare yourself eternal punishment in hell. And if not, well, what did it cost you apart from a few earthly pleasures here and there? Pascal’s Wager basically suggests that your upside is infinite, while your downside is relatively small. So do the right thing.

We need a Pascal’s Wager of curriculum. Schools are going to teach something, so it behooves us to ensure that the textbooks, workbooks, and software we put in front of students are coherent and of high quality. As this report from the Center for American Progress shows, crappy curriculum costs every bit as much as the good stuff. The authors found “little relationship” between the cost and quality of instructional products. And switching to a more rigorous math curriculum, for example, can deliver far greater returns on investment than other reforms. “The average cost-effectiveness ratio of switching curriculum was almost forty times that of class-size reduction in a well known randomized experiment,” the report notes.

Every opportunity for schools getting more bang for their curricular buck is a function of choosing the right product, not finding a better price. But for these cost-neutral benefits to accrue, we have to start taking curriculum seriously as a reform lever—at least as seriously as teacher quality, chartering, and other pet reform ideas, none of which have the demonstrated upside of simply choosing curriculum wisely and implementing it well. In short, we have a lot more empirical evidence for curricular effects than Pascal had for God. “In education, it is rare for a reform to show strong outcomes and be relatively inexpensive,” the authors conclude. “Curriculum reform is both cost-effective and worthwhile and should become a more central part of the effort to improve the nation’s schools.” Amen.

For starters, better “product research” is needed, including more “randomized experiments that clearly show which curricula produce the largest achievement gains.” Boser, Chingos, and Straus are also correct to argue that “when hard evidence on curriculum quality is available, it should supersede the often vague impressions of stakeholder groups that frequently dominate the [adoption] process.” Finally, let me add a recommendation of my own: Education reformers need to get over their blithe indifference on this issue. Some leading-edge reform outfits like KIPP and Achievement First have been quietly coming around, and Common Core itself is predicated on smart curricular choices. But more is needed to usher in an age of enlightenment. Of course teacher quality matters. It will matter even more when teachers are using materials of proven quality.

What have we got to lose? 

SOURCE: Ulrich Boser, Matthew Chingos, and Chelsea Straus, “The Hidden Value of Curriculum Reform: Do States and Districts Receive the Most Bang for Their Curriculum Buck?,” Center for American Progress (October 2015).

The relationship between teacher experience and quality has been widely studied, as has the relationship between teacher experience and salary. The relationship between experience and total compensation—which includes both salary and retirement benefits—is often overlooked. In a new report, researchers from the Manhattan Institute have thrown open the curtains by calculating the total compensation for teachers with master’s degrees and varying years of experience in the country’s ten largest public school systems. They explain that, although the preponderance of research demonstrates that quality differences between teachers based upon experience tend to plateau after 5–7 years, most public school teachers still earn salaries according to fixed schedules that are based entirely on years of experience and advanced degrees. Retirement benefits are distributed in a similar way. Approximately 89 percent of public school teachers earn retirement benefits under final-average-salary-defined benefit (FAS-DB) pension plans, meaning that teachers earn a lifetime annuity available only after they reach their respective plans’ thresholds. These thresholds, like a salary schedule, are based on a combination of age and years of service. As a result, FAS-DB plans often backload retirement benefits.

The scale of backloading varies across plans. In New York City, for example, a teacher earns an average of only $1,031 in retirement compensation during each of her first fifteen years; in each of the ensuing years, she earns an average of $16,908. In Philadelphia, teachers vest after ten years, which is the same year they reach the “maximum rung” on the salary ladder; however, they earn no positive retirement compensation until year seventeen, and their retirement benefits aren’t maximized until they reach thirty-five years of service. The authors point out that this causes significant problems for the large majority of public school teachers who leave teaching before they reach the sufficient years of service mark (not to mention teachers who are mobile throughout their careers). They also note that backloaded plans harm teachers who remain in the classroom after their specified retirement age. For these teachers, retirement compensation actually turns negative during the later years of their careers, making it financially wise for them to leave public schools. Overall, the report finds that backloaded pension plans “substantially increase the total compensation premium paid to highly experienced teachers who remain in the same school district their entire careers.”

Fortunately, there seems to be a solution: The report outlines an alternative model called a “cash-balance” (CB) pension plan, which allows teachers to more smoothly accrue pension benefits throughout their careers. The CB plan better aligns teacher compensation with teacher quality by “raising compensation for teachers in their early years—a time when teachers improve most dramatically.” The CB plan doesn’t link retirement plans to measurable teacher performance; it simply reallocates teachers’ retirement compensation more smoothly from the beginning to the end of their careers. Furthermore, the researchers found that for the ten school districts they studied, changing from a backloaded plan to a CB plan would be of equal cost to taxpayers and would reduce the teacher experience premium without reducing a teacher’s total expected career compensation. Sounds like a win for everybody.

SOURCE: Josh B. McGee and Marcus A Winters, “Rewarding Experienced Teachers: How Much Do Schools Really Pay?,” Manhattan Institute (October 2015).

Nearly everyone agrees that high-quality pre-kindergarten is a worthy investment. Calls to expand it at public expense stem from a handful of well-known (and very costly) intensive models that appeared to deliver long-term positive effects for poor children: improved school readiness, increased graduation rates, and even the mitigation of risk factors like teen pregnancy and incarceration. These oft-cited outcomes are compelling. So is the urge to level the playing field for children who arrive at school with a thirty million word gap. But an actionable definition of “high quality” remains elusive, and studies of large, scaled-up pre-K programs have shown mixed results.

This study from Vanderbilt University’s Peabody Research Institute adds valuable evidence to the discussion of whether, when, and how pre-kindergarten is a worthy investment. In 2009, in conjunction with the Tennessee Department of Education, the institute launched a rigorous study of the state’s voluntary pre-kindergarten program (TN-VPK). It’s a full-day program that targets exceptionally at-risk four-year-olds; researchers tracked two cohorts of children through the end of their third-grade years. Oversubscribed programs enabled a random design whereby children enrolled in TN-VPK were the treatment group and those waitlisted (and ultimately not admitted) became the control group. The study is one of a few randomized control trials of a large-scale pre-K program (most comparable studies employ less rigorous research designs) and the first ever of a state-funded, targeted program.

Findings reported last month come from a subset of the overall study’s sample (specifically the 1,076 children whose parents agreed to annual data collection: 773 participants in TN-VPK and 303 non-participants). Although this leaves something to be desired from a randomization perspective, the authors report that “the consented children…were generally representative of those in the full randomized sample.” Researchers sought to answer three questions. First, would TN-VPK prepare children for kindergarten (in both academic and non-cognitive ways)? Second, would certain subgroups benefit more than others? And third, would effects on achievement and behavior last beyond kindergarten?

To gauge the program’s immediate impact, children were assessed on several academic indicators (via achievement tests) and non-cognitive behaviors (via teacher ratings), both before and after they took part in the program. At the end of the pre-K year, researchers found that TN-VPK had a significant positive effect on children on six achievement subtests, with especially large effects in two literacy measures. The effect size for the composite achievement score was .32—a figure that, according to researchers, is “larger than average” for the pre-K programs studied in the last three decades. The program had the greatest impact on children who were English language learners and/or whose mothers had attained less than a high school degree.

But don’t uncork the champagne. To see if the program’s effects persisted, the children were also assessed and received teacher ratings at the end of each school year (through third grade). Not only did the positive effects of TN-VPK wear off by the end of kindergarten, but they had turned negative by the end of second grade—that is, TN-VPK children scored lower than non-participants on all achievement measures (with most of those comparisons being statistically significant). This reversal occurred even earlier on behavioral effects: By the end of first grade, TN-VPK participants were rated by teachers as having worse work skills and being less prepared for school. Comparable studies of many pre-K programs have discovered a gradual “fadeout” of positive effects once children enter the K–12 system. But the long-term negative impacts discovered by Vanderbilt’s researchers were quite unexpected, and they prompted the authors to urge caution on those who interpret their findings.

So what do we make of this? It’s important to note that children in the control group didn’t receive no treatment—just treatment that wasn’t TN-VPK. Twenty-seven percent of the control group attended Head Start or a private childcare center; the rest received care at home (the quality and details of which are impossible to track). Their alternative early learning experiences may not have prepared them as well for kindergarten as TN-VPK, but that ended up not mattering by the start of first grade. One way of looking at it is that Head Start, private preschools, home providers, and families are setting children up for K–3 success at least as well as (and possibly better than) the state-funded program, which costs $90 million annually.

It’s also important to consider that the “fadeout” of pre-K observed in this study and many others may not be the fault entirely of the pre-K programs. It’s possible that pre-K participants entering kindergarten ahead of their peers are neglected somewhat by teachers, who focus instead on youngsters in the most dire academic need—an understandable but still damaging method of triage. Teachers must capitalize on and sustain student gains made prior to their entry, and also meet the needs of kids who are kindergarten-ready.

Finally, the study calls into question what constitutes “high quality” pre-K. Responses and reactions (sometimes overreactions) to the report were swift, motivating two of the study’s authors to pen a follow-up article. “The virtual ink on our recently released report was barely dry before pre-K advocates were vigorously building a firebreak around these results,” they wrote, “contending that they are not representative of the effects of state pre-K programs generally and stem entirely from the unusually poor quality of the Tennessee program.” Defending Tennessee, they note that TN-VPK is no different from programs being ramped up in other states. It aligns with nine of the ten benchmarks provided by the National Institute for Early Education Research (NIEER). TN-VPK participating districts had all the right inputs: teachers with licenses in early childhood and education, classrooms with an adult-student ratio of 1:10, maximum class sizes of twenty, and state board-approved curricula. This assertion makes the report’s findings even grimmer and adds new urgency for how we reach the most at-risk early learners. If Tennessee’s statewide pre-K program aligns to the best consensus we have on what matters in pre-K, and it is comparable to rapidly expanding programs around the country, we need to rethink the benchmarks for quality. Moreover, for a program to be of truly high quality for at-risk children, it must go beyond a one-year effort at age four.

SOURCE: Mark W. Lipsey, Dale C. Farran, Kerry G. Hofer, “A Randomized Control Trial of a Statewide Voluntary Prekindergarten Program on Children’s Skills and Behaviors through Third Grade,” Peabody Research Institute at Vanderbilt University (September 2015).

David Sacks/Digital Vistion/Thinkstock