Video: Education Next talks with Mike Petrilli.
The greatest challenge facing America’s schools today isn’t the budget crisis, or standardized testing, or “teacher quality.” It’s the enormous variation in the academic level of students coming into any given classroom. How we as a country handle this challenge says a lot about our values and priorities, for good and ill. Unfortunately, the issue has become enmeshed in polarizing arguments about race, class, excellence, and equity. What’s needed instead is some honest, frank discussion about the trade-offs associated with any possible solution.
U.S. students are all over the map in terms of achievement (see Figure 1). By the 4th grade, public-school children who score among the top 10 percent of students on the National Assessment of Educational Progress (NAEP) are reading at least six grade levels above those in the bottom 10 percent. For a teacher with both types of students in her classroom, that means trying to challenge kids ready for middle-school work while at the same time helping others to decode. Even differences between students at the 25th and at the 75th percentiles are huge—at least three grade levels. So if you’re a teacher, how the heck do you deal with that?
In the old days, “ability grouping” and tracking provided the answer: you’d break your students into reading groups, with the bluebirds in one corner, tackling advanced materials at warp speed, and the redbirds in another, slowly making their way through basic texts. Likewise for mathematics. And in middle and high school, you’d continue this approach with separate tracks: “challenge” or “honors” for the top kids, “regular” or “on-level” for the average ones, and “remedial” for the slowest. Teachers could target their instruction to the level of the group or the class, and since similar students were clustered together, few kids were bored or totally left behind.
Then came the attack on tracking. A flurry of books in the 1970s and 1980s argued that confining youngsters to lower tracks hurt their self-esteem and life chances, and was elitist and racist to boot. Jeanne Oakes’s 1985 opus, Keeping Track, was particularly effective in sparking an anti-tracking movement that swept through the nation’s schools.
According to Brookings Institution scholar Tom Loveless, this advocacy led to fundamental changes at breakneck speed. In a report for the Thomas B. Fordham Institute last year, he wrote,
An eighth grader in the early 1990s attended middle schools offering at least two distinct tracks in [each of] English language arts, history, and science. Mathematics courses were organized into three or more tracks. The eighth grader of 2008, however, attended schools with much less tracking. English language arts, history, and science are essentially detracked, i.e., schools typically offer a single course that serves students at every level of achievement and ability. Mathematics usually features two tracks, often algebra and a course for students not yet ready for algebra.
One of the reasons that detracking advocates claimed so many victories is that they painted their pet reform as a strategy in which everybody wins. Oakes and others insisted that detracking would help the lowest-performing students (who would enjoy better teachers, a more challenging level of instruction, and exposure to their higher-achieving peers) while not hurting top students. But by the mid-1990s, researchers started to compile evidence that this happy outcome was just wishful thinking.
In 1995, scholars Dominic Brewer, Daniel Rees, and Laura Argys analyzed test-score results for high-school students in tracked and detracked classrooms, and found benefits of tracking for advanced students. They wrote in the Kappan magazine, “The conventional wisdom on which detracking policy is often based—that students in low-track classes (who are drawn disproportionately from poor families and from minority groups) are hurt by tracking while others are largely unaffected—is simply not supported by very strong evidence.”
And this was before the policy incentives shifted sharply to prioritize low-achieving students. In another study for the Fordham Institute, Loveless found a clear pattern in the late 1990s when states adopted accountability regimes: the performance of the lowest decile of students shot up, while the achievement of the top 10 percent of students stagnated. That’s not surprising; these accountability systems, like No Child Left Behind (NCLB) in 2002, pushed schools to get more students over a low performance bar. They provided few incentives to accelerate the academic growth of students at the top.
This dynamic might have been most pernicious for minority students. Earlier this year, an Indiana University study found that the “Excellence Gap,” the racial achievement gap at NAEP’s advanced level, widened during the NCLB era. One possible explanation is that high-achieving minority students are likely to attend schools with lots of low-achieving students, and their teachers are focused on helping children who are far behind rather than those ready to accelerate ahead.
The Power of Peers
The attack on tracking also claimed an innocent bystander: ability grouping, which became suspect in many circles, too. Yet in recent years, the “peer effects” literature has shown the benefits of grouping students of similar abilities together. One clever study, by economists Scott Imberman, Adriana Kugler, and Bruce Sacerdote, looked at the fallout from Hurricanes Rita and Katrina. They wanted to know what happened when students who were evacuated from New Orleans ended up in schools in Houston. They found that the arrival of low-achieving evacuees dragged down the average performance of the Houston students and had a particularly negative impact on high-achieving Houston kids. Meanwhile, high-achieving evacuees had a positive effect on local students. As Bruce Sacerdote told me, “The high-achieving kids seemed to be the most sensitive. They do particularly well by having high-achieving peers. And they are particularly harmed by low-achieving peers.” He added, “I’ve become a believer in tracking.”
In 2006, Caroline Hoxby and Gretchen Weingarth examined the Wake County (North Carolina) Public School System. For the better part of two decades, the district, in and around Raleigh, had been reassigning numbers of students to new schools every year in order to keep its schools racially and socioeconomically balanced. That created thousands of natural experiments in which the composition of classrooms changed dramatically, and randomly, and that, in turn, provided Hoxby and Weingarth an opportunity to investigate the impact of these changes on student achievement.
They found evidence for what they called the “boutique model” of peer effects, “a model in which students do best when the environment is made to cater to their type.” When school reassignments resulted in the arrival of students with either very low or very high achievement, this boosted the test scores of other students with very low or very high achievement, probably because it created a critical mass of students at the same achievement level, and schools could better focus attention on their particular needs.
Does that mean students should be sharply sequestered by ability? Not exactly. Here’s how Hoxby and Weingarth put it in their conclusion: “Our evidence does not suggest that complete segregation of people, by types, is optimal. This is because (a) people do appear to benefit from interacting with peers of a higher type and (b) people who are themselves high types appear to receive sufficient benefit from interacting with peers a bit below them that there is little reason to isolate them completely. What our evidence does suggest is that efforts to create interactions between lower and higher types ought to maintain continuity of types.”
In other words, a little bit of variation is okay. But when the gap is too wide—say, six grade levels in reading—nobody wins.
Enter Differentiated Instruction
So if grouping all students together leads to pernicious effects, but divvying kids up by ability is politically unacceptable, what’s the alternative? The ed-school world has an answer: “differentiated instruction.” The notion is that one teacher instructs a diverse group of kids, but manages to reach each one at precisely the appropriate level. The idea, according to Carol Tomlinson of the University of Virginia (UVA), is to “shake up what goes on in the classroom so that students have multiple options for taking in information, making sense of ideas, and expressing what they learn.” Ideally, instruction is customized at the individual student level. Every child receives a unique curriculum that meets that individual’s exact needs. A teacher might even make specialized homework assignments, or provide the specific one-on-one help that a particular kid requires.
If you think that sounds hard to do, you’re not alone. I asked Holly Hertberg-Davis, who studied under Tomlinson and is now her colleague at UVA, if differentiated instruction was too good to be true. Can teachers actually pull it off? “My belief is that some teachers can but not all teachers can,” she answered.
Hertberg-Davis worked with Tomlinson on a large study of differentiated instruction. Teachers were provided with extensive professional development and ongoing coaching. Three years later the researchers wanted to know if the program had an impact on student learning. But they were stumped. “We couldn’t answer the question,” Hertberg-Davis told me, “because no one was actually differentiating.”
Teachers admit to being flummoxed by this approach. In a 2008 national survey commissioned by the Fordham Institute, more than 8 in 10 teachers said differentiated instruction was “very” or “somewhat” difficult to implement. Even ed-school professors are skeptical. A 2010 national random survey of teacher educators asked them the same question and got the same result: more than 8 in 10 said differentiated instruction was very or somewhat difficult to implement.
But that doesn’t mean it’s impossible. I was curious to see differentiated instruction in action, so I visited my local elementary school in Takoma Park, Maryland. Piney Branch Elementary serves an incredibly diverse group of 3rd, 4th, and 5th graders, from the children of übereducated white and black middle-class families, to poor immigrant children from Latin America, Ethiopia, and Eritrea, to low-income African American kids.
I sat down with the school’s principal, Bertram “Mr. G.” Generlette, who has the friendly, laid-back manner of his native Antigua. I cut right to the chase. I’m wondering if I’d be making a mistake to send my son to a school like Piney Branch. Is it going to slow him down if his classmates are several years behind or still learning the language? (Of course, not all poor or minority children are low-achieving, nor are all white students high-achieving. Still, achievement gaps being what they are, the range of academic diversity does tend to be larger at schools with lots of racial and social diversity.)
It was pretty obvious that Mr. G. had heard these questions before, particularly from white folks like me. I asked him if that was the case. “Parents come in, yes,” he told me. “They are new to the neighborhood. Or their child is in kindergarten, or they are moving from private school. After a few minutes, you get the idea.” However, he said with a sly grin, “they very rarely ask the question directly.”
But he wasn’t afraid to answer me directly. “We are committed to diversity,” he started. “It’s a lens through which we see everything. We look at test scores. How are students overall? And how are different groups doing? It’s easy to see. Our white students are performing high. What can we do to keep pushing that performance up? For African American and Hispanic students, what can we do to make gains?”
Since Mr. G.’s arrival five years ago, the percentage of African American 5th graders passing the state reading test is way up, from 55 to 91 percent. For Hispanic children, it’s up from 46 to 74 percent. It’s true that scores statewide have also risen, but not nearly to the same degree.
And there’s no evidence that white students have done any worse over this time. In fact, they are performing better than ever. Before Mr. G. arrived, 33 percent of white 5th graders reached the advanced level on the state math test; in 2009, twice as many did. In fact, Piney Branch white students outscore the white kids at virtually every other Montgomery County school.
What’s his secret? Was he grouping students “homogeneously,” so all the high-achieving kids learned together, and the slower kids got extra help?
“There’s no such thing as a homogenous group,” Mr. G. shot back. “One kid is a homogeneous group. As soon as you bring another student in, you have differences. The question is: how do you capitalize on the differences?”
Well, that sounds OK in theory. But come on, Mr. G., how are you going to make sure my kid doesn’t get slowed down?
“My job as a principal is to let my parents know that your child will get the services they need,” he answered patiently. “We are going to make sure that every child is getting pushed to a maximum level. That’s my commitment.”
And that’s when I was introduced to the incredibly nuanced and elaborate efforts that Piney Branch makes to differentiate instruction, challenge every child, and avoid any appearance of segregated classrooms.
So how do they do it? First, every homeroom has a mixed group of students: the kids are assigned to make sure that every class represents the diversity of the school in terms of achievement level, race, class, etc. Then, during the 90-minute reading block, students spend much of their time in small groups appropriate for their reading level. (Redbirds and bluebirds are back!) However, in the new lingo of differentiated instruction, the staff works hard to make sure these groups are fluid—a child in a slower reading group can get bumped up to a faster one once progress is made.
For math, on the other hand, students are split up into homogeneous classrooms. All the advanced math kids are in one classroom, the middle students in another, and the struggling kids in a third. This means shuffling the kids from one room to another (a process that can be quite time-consuming for elementary school kids). But it allows the highest-performing kids to sprint ahead; one of the school’s 3rd-grade math classes, for example, is tackling the district’s 5th-grade math curriculum. (Because of large achievement gaps at the school, these math classes are more racially and socioeconomically homogeneous than the student population as a whole.)
The rest of the time—when kids are learning science or social studies or taking “specials” like art and music—they are back in their heterogeneous classrooms. Even then, however, teachers work to “differentiate instruction,” which often means separating the kids back into homogeneous groups again, and offering more challenging, extended assignments to the higher-achieving students.
It sounds like some sort of elaborate Kabuki dance to me, but it appears to succeed on several counts. All kids spend most of the day getting challenged at their level, and no one ever sits in a classroom that’s entirely segregated by race or class.
Reading War
Test scores indicate that the strategy is working, too, but that doesn’t mean all parents have been thrilled. Three years ago, Mr. G. told me, a group of white parents pushed to get the school to move to homogeneous classrooms for reading as well as math. “Parents felt that the only way to get kids to read at a high level was to have other kids around them who read at a high level,” he explained. (That didn’t sound so unreasonable to me.) “We had a lot of meetings. The staff overwhelmingly supported the diverse approach, the heterogeneous approach. That was good for me as an administrator because the staff was behind me.”
I tracked down one of the “troublemaker” parents. Her name is Sue Katz Miller and she personifies much of what makes Takoma Park great: she’s smart, she’s an activist, and she’s committed to helping make the city a welcoming community for families of all incomes and backgrounds. (A neighbor of mine called her “a force of nature.”) A former Newsweek reporter and now a regular columnist for The Takoma Voice, she spent two years as PTA president at Piney Branch and is an enthusiastic booster of the school and its diversity. “My kids have both benefited enormously from being in a Piney Branch social milieu,” she told me.
But the reading decision still sticks in her craw. “Why is it OK,” she asked, “to have homogeneous grouping in math and not have it in reading? The answer you get is: well, we can’t do both, they would be switching classes all the time, it would be like middle school and they won’t be able to handle it…. It’s a huge disservice to the kids who are ready for rigor in the humanities and are not math kids. It’s bizarre. We’ve said we’re going to accommodate kids in math but not in reading. It’s completely insane as far as I’m concerned. It makes me angry.”
She lost that battle, but Mr. G. and his teachers didn’t ignore the parents’ concerns, either. He went out and found reading programs suitable for advanced students, like William and Mary, Junior Great Books, and Jacob’s Ladder. He trained his teachers on these programs, ensuring that the students in the top reading groups would be challenged with difficult material. (The teachers loved it.) He tried hard to live up to his promise to push all students as far as they could go.
Competing for Kids
Mr. G. and Piney Branch face some healthy competition. Montgomery County offers a half-dozen “Centers for the Highly Gifted,” magnet schools that are designed for supersmart kids and located in elementary buildings throughout the district. Pine Crest, just a few miles away from Piney Branch, hosts one such center, and an increasing number of Piney Branch 3rd graders were testing into it for 4th and 5th grades.
A year ago, 25 Piney Branch kids were accepted—more than any other elementary school in the district. If they all took up the offer, Mr. G. said, “That’s a teacher walking out of my building.”
So in 2009–10, in cooperation with the district, Piney Branch launched a pilot program to bring the “Highly Gifted Center” curriculum into its classrooms. This wasn’t easy; there wasn’t a curriculum, per se, at the centers. Teachers had the freedom to do what they wanted. So the district helped the teachers put down on paper everything they were doing in the classroom.
Mr. G. arranged to have a 4th-grade and a 5th-grade teacher trained on the Highly Gifted approach, and formed a “cluster group” of gifted students in their classrooms. This means that, in one classroom in each of these grades, there are 12 or so gifted students, along with another 12 or so “on-level” kids. While they are taught together some of the day, they are frequently broken into small groups, so the gifted kids can learn together at an accelerated pace.
Pulling this off takes an energetic and gifted educator; 4th-grade teacher Folakemi Mosadomi, who has the gifted group in her classroom, appears to fit the bill perfectly. Now in her 5th year of teaching (all of them at Piney Branch under Mr. G.), Ms. M. acknowledged that differentiating instruction in this way requires “extensive planning and training,” not to mention someone who is well-organized and creative. But even that’s not always enough.
In the first year of the pilot, she had four different reading groups in one classroom, from kids still learning English to the highly gifted students. “I went from sounding out the ‘A’ sound with one group, to talking to another group about how the Exxon Valdez oil spill was like the Battle of Normandy.” That range was simply too much for one teacher to handle—remember Caroline Hoxby’s finding about “continuity of types?”—so the next year she had just two groups: the gifted students, and the next level down. “Now it’s easier to do more with both groups of students together,” she told me.
And the strategy seems to be working in one important way: last year, about half of the gifted children chose to stay at Piney Branch.
Fragile Compromise
So with a well-trained and dedicated staff, and lots of support, “differentiated instruction” can be brought to life. But even at Piney Branch, which benefits from the vast resources of a huge, affluent school system in Montgomery County, Maryland, it sure seems rickety, held with lots of duct tape and chewing gum, and subject to collapse without just the right staff and parent support.
If the school community placed its highest value on pushing all kids to achieve their full potential, including its high-achieving students, it would probably organize its classrooms differently. It would embrace “ability grouping” and homogenous classrooms wholeheartedly, and would skip all the gymnastics required to keep classes academically, racially, and socioeconomically diverse throughout the day. But Piney Branch understandably seeks to balance its concerns for academic growth with its interest in maintaining an integrated environment, so this uneasy compromise is probably the best it can do.
Piney Branch and Ms. M. might be able to pull it off. But how many Piney Branches and Ms. M.’s are there?
Technology may someday alleviate the need for such compromises. With the advent of powerful online learning tools, such as those on display in New York City’s School of One, students might be able to receive instruction that’s truly individualized to their own needs—differentiation on steroids.
Perhaps. But until that time, our schools will have to wrestle with the age-old tension between “excellence” and “equity.” And that tension will be resolved one homogeneous or heterogeneous classroom at a time.
Michael J. Petrilli is executive editor of Education Next, research fellow at Stanford University’s Hoover Institution, and a vice president at the Thomas B. Fordham Institute. He is working on a book for parents considering diverse public schools like Piney Branch.
