The Education Gadfly Weekly

American K–12 education is awash in reforms, nostrums, interventions, silver bullets, pilot programs, snake oil peddlers, advocates, and crusaders, not to mention innumerable private foundations that occasionally emerge from their endless cycles of strategic planning to unload their latest brainstorms upon the land. Yet when subjected to close scrutiny, not much actually “works”—and at the high school level practically nothing seems to. Sometimes the flaw was in the conception itself, sometimes in the implementation, oftentimes in the peerless ability of a vast, entrenched, bureaucratic system to repel, besiege, and ultimately tame or expel disruptive innovations of all sorts.

Against that sorry backdrop, the six-decade old Advanced Placement (AP) program qualifies as a rare and welcome exception, as we find in our new book, Learning in the Fast Lane: The Past, Present, and Future of Advanced Placement, published this week by Princeton University Press. The program’s success is all the more remarkable considering how big it’s become, even while maintaining its curricular quality and intellectual rigor: Some three million high schoolers sat for some five million AP exams in May, spanning thirty-eight separate subjects. Despite stiffening competition from dual enrollment and a few other course providers, it’s by far the largest source of college-level work for American high school students. Yet instead of dumbing it down to attract more kids, the College Board has done a commendable job of preserving the attributes that built its reputation in the first place.

How has AP succeeded—and endured and grown and not gone squishy—where so many other innovations and reforms have failed?

First, AP is private. It’s not a government thing, certainly not a federal thing like Race to the Top or No Child Left Behind. The nonprofit, nonpartisan College Board operates it, and does so with mostly private dollars, though districts incur costs when they offer AP courses and a number of states and districts subsidize kids’ exam fees. Those costs inevitably compete in school budgets with other priorities, yet the private nature of AP means that politicians don’t have to approve what’s taught, what’s tested, or much of anything else.

Second, AP is a demand-driven reform that’s almost always voluntary for districts, schools, teachers, and kids alike. Neither the states nor the feds have much to do with it. AP has spread mostly because parents and students want it and see value in it. It has also thrived because thousands of teachers love the challenge and stimulation of AP classrooms. Try to name five other education reforms that attract students and teachers in this way!

Third, AP isn’t all that disruptive an innovation. Unlike charter schools, voucher programs, and virtual schools, it’s not a newfangled institutional form—and it doesn’t usually lead kids to change schools. Unlike personalized learning and online instruction, it’s not a pedagogical or technological novelty. Unlike so many reforms of recent decades, it’s something teachers generally welcome—and those that want no part of it are generally free to keep doing what they’ve always done.

Fourth, successful completion of AP courses and exams yields actual and valuable benefits for the young people who participate, particularly for those who also score at least 3 on the AP exam scale from 1 to 5. Such a “qualifying score” gives them a good shot at arriving in college with credit already established and/or waiving out of boring freshmen courses and possibly earning their degrees faster and cheaper. Students also acquire a backpack full of study skills and practices that will help them succeed in college and beyond; gain a welcome boost in their admissions odds, perhaps to better colleges than they would otherwise even have applied to. Success on AP may also confer confidence that yes, one is indeed “college material,” while serving as external validation of one’s transcript, class rank, teacher comments, etc.

AP has no monopoly on those attributes of success, which is to say other education initiatives could also make use of them. Imagine, for example, an externally validated CTE program, with a standardized curricular framework, providing AP-level prestige to teachers and schools and an attractive option for kids.

AP faces its own challenges, to be sure. It’s not cheap for either schools or kids. In the hands of a weak instructor, it can resemble “teaching to the test” (though AP exams come closer than just about any others to being “tests worth teaching to”). Small, rural schools can seldom offer much of it, and even the best high schools face big obstacles in trying to get perfectly able kids to succeed at AP-level work when they arrive from mediocre (or worse) middle schools.

But for rapidly increasing numbers of students—including many more low-income and minority youngsters—it’s working pretty well. It’s become a lever to open opportunity for high-ability youngsters from poor and minority backgrounds, not just a cure for senioritis among the privileged kids who were its main users a few decades back.

AP’s major outstanding challenge, as it continues to open its doors to more—and more—diverse students (and that job is certainly not finished), is to ensure that many more of those kids leave its classrooms with scores of 3 or above. That takes significant effort, no small amount of resources, and a tremendous change of mindset by teachers and school leaders alike.

But perhaps the biggest lesson from AP’s tremendous growth and solid reputation is that much good can be done by voluntary—even niche—ventures that meet some kids’ (and teachers’ and schools’) needs without necessarily serving everyone, without overhauling structures, and without getting too tangled up in government, regulation, and standardization.

The latest Education Next poll asked respondents whether they support ability grouping, whereby students take classes with peers at similar academic achievement levels, and for middle school the majority’s answer was no. This is likely bad news for high-ability students, who already suffer from a dearth of quality gifted programming and academic acceleration, but it does open the door for what we at Fordham call “personalized pacing”—a pragmatic reform that has the potential to significantly improve the outcomes of advanced students.

The survey asked respondents this question: “Do you think middle school students should be in their core academic classes with only middle school students of similar ability levels, or should they be in class with middle school students of all ability levels?” Education Next gathered responses for the general public, as well as eleven subgroups, such as parents, teachers, Republicans, Democrats, and adults of different races, ethnicities, and income levels.

Just one subgroup preferred ability grouping for middle schoolers: highly-educated white adults, 51 percent of whom supported the policy. In every other instance, the majority of respondents weren’t in favor. Those most opposed included black and Hispanic adults, among whom 59 percent and 56 percent preferred inclusive classrooms, respectively.

As my colleague Michael Petrilli wrote a decade ago and reaffirmed earlier this year, “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.”

Ability grouping is one response to this problem. But if people don’t want it, there are really just two alternatives. One is whole-group instruction, wherein educators basically aim their instruction at the median achievement or ability level in a given classroom. This solution, which is still common, does little for either low or high achievers.

The other option is some form of differentiation. The classic way asks individual teachers instructing a diverse group of kids to somehow reach each one at precisely the appropriate level using only their curricular materials and expertise. A splendid notion, perhaps, but it doesn't work at scale in heterogeneous classrooms, especially when educators are strapped for time, resources, and perhaps the requisite pedagogical skill set. They’re forced to triage, focusing on those in greatest need and paying less attention to everyone else.

But there’s a more modern form of differentiation that enlists instructional technology to enable overworked and overwhelmed teachers to run classrooms in which each student does receive a unique curriculum. Termed “personalized pacing,” it usually uses digital tools to accelerate student learning by tailoring instruction to address their individual needs and accounting for prior achievements. In some ways, it’s actually more akin to grade-level acceleration—a form of ability grouping with a long track record of success—than differentiation. The digital tools facilitate the same enhanced academic benefit, while allowing students to remain in socially advantageous and less controversial same-age classrooms.

It’s not a silver bullet. Its use among kids struggling to meet standards is debated, so more research is needed to determine the best application for those children. And misty-eyed optimism about the latest hardware and greatest software can cause leaders to over-rely on technology and overpromise on its potential, while discounting the risks associated with privacy and profiteering.

For high-ability students and those performing above standards, however, this form of personalization of instruction is a great solution. Teach to One: Math, a middle and high school math program designed by New Classrooms, is a good example. Petrilli summarized it well in another essay:

At the end of each day, students take a brief assessment to gauge how well they have mastered the math they’re working on. Overnight, an algorithm designed by New Classrooms figures out the exact skill each student is ready to learn next, as well as the “modality” that would be the best fit—like whole group instruction, small group instruction, or online learning. In the morning, kids look up at “airport monitors” to find out what and where they will be learning that day, and off they go. Most of the instruction is done with a teacher, in large or small groups, but those groups are constantly changing, bringing students together who are all ready to learn the same skill.

Innovations such as this can do wonderful things. A smart black sixth-grader languishing in a class that does nothing to maximize his potential can be challenged for the very first time, just as a low-income eighth-grader attending a small rural school can finally access the math content she’s long wanted and deserved.

Deploying technology in this way also enjoys wide public acceptance. Education Next’s 2017 poll, the last time it asked about technology in the classroom, found that just 35 percent of the general public opposed students spending more time with digital devices. So personalized pacing is pragmatic, and there’s a window of opportunity for school systems to adopt and deploy it, albeit with wisdom about the needs of both students and teachers. Done right, it holds immense potential to enhance student learning and do what schools have always struggled to do—fairly focus instruction on each individual student.

Editor’s note: This is the third in a series of posts looking at how two school networks—Rocketship Public Schools and Wildflower Schools—enable their students to meet standards at their own pace. See the first two posts here and here.

Last week, we examined how Wildflower and Rocketship ensure that their efforts to tailor instruction to individual kids don’t end up lowering the bar for their struggling students. Both schools are fully committed to making sure all of their charges meet academic standards so they are well prepared for life and further learning. They just don’t think students should have to march through the curriculum in lock-step.

This week, we’ll look at the other end of the performance spectrum: the schools’ high achieving students. For such kids, personalized pacing seems to come with nothing but upsides—unless you oppose in principle any kind of individualized instruction for “gifted” students, as some in New York City apparently do. At Rocketship the focus is on “acceleration”; Wildflower, a Montessori model. is more about “enrichment.” Let’s take a look.

Rocketship’s Preston Smith:

At Rocketship Public Schools, we serve all students with excellence—no matter how far behind or ahead they are when entering our schools. We meet each Rocketeer’s individual needs through our personalized learning model and see tremendous academic growth at all levels of achievement. In our schools, all students have the ability to move beyond their grade level materials through their individualized learning and small group time in each subject. This happens in online learning programs and individual reading and work time, usually through next-level assignments.

During our “guided reading,” for example, Rocketeers are grouped by reading levels that are sometimes at the “next grade” and led by their teacher. For those Rocketeers who are at even higher levels, we may have them participate in a guided reading group at another, higher grade level with a different teacher who may have greater mastery with particular reading levels. We use STEP (a reading assessment developed at the University of Chicago), which levels students for all reading materials and skills. Furthermore, STEP allows our Rocketeers to take ownership of their reading goals, always knowing which STEP level they are at and how to progress to the next. When students advance, even if the next level is technically aligned to a higher grade, the student can always access books at this level and immediately begin reading. Moreover, by having flexibility in classroom groupings, these Rocketeers can always access a teacher with the expertise to help them continue to master the skills required at their current reading level. This helps to prevent any backsliding in reading skills.

Similarly, in STEM, we have small group instruction time, where a teacher can regularly “extend” a lesson or content area (either in person, via independent assignments, or online programs) and ensure that high achievers are gaining access to and mastery of content at the next grade level.

We strongly believe that it is important to challenge students at all times and help them feel engaged in their learning, so we make sure all students have the appropriate materials for their individual learning level and to extend beyond.

Wildflower’s Matthew Kramer:

There’s pretty compelling anecdotal evidence in the form of the many extraordinary people who credit their time in a Montessori environment as a significant reason for their development. We see this in the choices that people with many options make about the school model for their own children, and in the recommendations conventional teachers frequently make to families looking for more challenging environments for academically successful but bored children. 

One way to think of the Montessori curriculum is as a series of basic introductions to various ideas and skills that lay the groundwork for a wide variety of independent explorations. For example, a group of lower elementary children might be shown how to multiply three-digit numbers by two-digit numbers using the Montessori multiplication checkerboard—an iconic Montessori math manipulative. Doing this multiplication involves meticulously following a detailed, multi-step algorithm. If two children receive the same presentation, one might immediately take to the work and make up a list of twenty-five such problems and spend hours working through them independently, carefully writing their answers, and making a bound book out of it that they share with others. The other might take to it more slowly, spending several hours working through a single problem, making mistakes, realizing those mistakes through the material’s built in control of error, and asking a friend for help.

This pattern plays out across a range of topics. A presentation on early humans could lead to a brief report on the subject, or it could lead to a multi-week investigation involving multiple children, student planned and executed trips to a natural history museum, direct outreach to a working paleoanthropologist, in depth research using the internet or the library, a twenty-page report, and a whole-class presentation.

Because the topics introduced in Montessori represent an age-appropriate tour of the whole of human knowledge, there is limitless potential to take a topic and go deep with it, constrained only by a child’s curiosity and the fact that there are always other topics worth exploring around the next bend.

Charters schools are often criticized for not enrolling enough or not adequately serving special student populations, particularly students with special needs. A new study by Tufts University’s Elizabeth Setren evaluates this claim with a unique dataset in Boston. The report looks at the impact of special education and English language learner classification in charter schools in Beantown. Specifically, the outcome of interest is the impact of charter school attendance on academic outcomes for students relative to their pre-lottery special needs status.

Setren makes use of data from randomly-assigned charter school admissions lotteries from 2003–04 through 2014–15, which includes information on the baseline characteristics of students in the lottery and post-lottery outcomes for lottery winners and losers. (Pre-lottery demographics and test scores are similar for those offered and not offered a slot.) The sample includes admissions lottery data from thirty of Boston’s elementary, middle, and high school charter schools, comprising roughly 90 percent of the sector.

Descriptive data show that, by spring 2014, the prevalence of students with special needs in middle school charter lotteries was similar to Boston Public Schools (BPS)—both were around 22 percent. The same was true for English language learners in both sectors—though at slightly higher percentages. (Notably, legislation was passed in Massachusetts in 2010 that required charter schools to increase their efforts to recruit and retain English language learners and students with special needs.)

The key finding is that charter schools remove special-education classifications and place students with special needs in more inclusive settings at the time of enrollment at much higher rates than do traditional public schools (TPS). Specifically, applicants with special-needs status at the time of the lottery are 11.8 percentage points more likely to have their classification removed in charter schools than in TPS. That pattern is consistent for students with more severe disabilities too: Applicants placed in separate classrooms at pre-lottery time are 17.3 percentage points less likely to keep their special-needs status in a charter school compared to TPS. Related, charters move special-needs applicants to more inclusive classrooms 27.1 percentage points more often than do TPS, meaning they spend more time in a general education setting and less time receiving services outside of the mainstreamed classroom.

A similar pattern played out with English language learners. Charter schools removed the English-language-learner status at the time of enrollment 31.8 percentage points more often than did TPS, although those with intermediate and advanced English proficiency drove the differences—those with beginning English proficiency rarely had their ELL classification removed at the time of enrollment. Since learning gains cannot justify the classification differences—recall that students’ placements are changed at the time they enroll—they attempt to find out what might. They discover one of the main drivers is that parents do not disclose the special-needs classification of their child at enrollment. And because the transfer of records between Boston Public Schools and the charter sector takes much longer than between BPS schools, charters must rely on this voluntary parental reporting.

Now for the test score effects. In a nutshell, charter school attendance has large positive effects on math and English language arts test scores for students with ELL or special-needs status at the time of the lottery. For example, a year of charter attendance increases math test scores by 0.26 standard deviations for special needs applicants and by 0.33 standard deviations for English language learner applicants. (ELA increases are in the same ballpark though slightly lower.) These effects appear to accumulate in the first two years and then level off. Finally, charter school attendance also boosts the chances that both English language learners and students with special needs will reach proficiency on the tenth grade math and ELA exams. And it makes it more likely that ELLs in particular will score above 900 out of 1600 on the SAT and enroll in a four-year college.

In looking at potential mechanisms, Setren concludes that the general charter school environment appears to be driving the academic gains—things like a longer school day, a strict behavior code, and emphasis on high expectations. Her last sentence is direct and on point: “The finding that special education students and English language learners can make large academic gains without specialized services in a high quality general education program calls for greater attention to general practices, in addition to the current focus on specialized supports, to improve special needs students’ outcomes.”

Yep, I’d say so. Especially when we hear so often that the cost of special education is rising. And that school districts tend to do a very poor job of boosting the quality and efficiency of it.

SOURCE: Elizabeth Setren, “The Impact of Targeted vs. General Education Investments: Evidence from Special Education and English Language Learners in Boston Charter Schools,” retrieved from Annenberg Institute at Brown University (June 2019).

A new report by Ulrich Boser and The Learning Agency investigates what K–12 educators know—or mistakenly believe—about effective learning strategies and where they obtain information about learning research.

Boser created a scenario-based survey in which respondents chose whether one approach to teaching or studying is more, less, or equally effective than another approach. Both direct and scenario-style questions focused on empirically supported learning strategies: elaboration, retrieval practice, metacognition, spaced practice, interleaving, and dual coding. Questions asked about several learning neuro-myths—including, for example, the idea that students are either right- or left-brained and have different “learning styles” (e.g., auditory, visual, kinesthetic) and teachers should tailor instruction accordingly, and that intelligence can’t be altered through education. Boser also included questions regarding where educators learn about new teaching approaches.

The survey was distributed to an online panel of 515 educators, and 203 responses were used for the results, comprising 159 teachers, 37 support staff, and 7 administrators. The sample was broadly representative of the racial and ethnic diversity of the teaching workforce in the U.S., and the common income categories reported by participants were comparable to the average teaching salary. The sample does, however, overrepresent male teachers, as 51 percent of respondents identified as male, while about 23 percent of U.S. teachers are male.

Boser found that, on average, respondents performed only somewhat better than chance when answering questions related to views on effective teaching strategies and learning myths. Out of seventeen questions total, they identified 8.34 of the beliefs correctly on average, while random chance would yield an average response rate of 6.63. An overwhelming majority weren’t able to identify interleaving (mixing up problem types or kinds of examples) and integrating text and visuals together as particularly effective strategies. Only 31 percent correctly endorsed retrieval practice (actively trying to recall information one wants to remember) over re-reading when directly asked which would be better for learning—though 59 percent answered correctly when given the classroom scenario version of the question. Encouragingly, about 60 percent of educators could correctly identify the effectiveness of elaboration (meaningfully linking new information in the mind to other information), spaced practice (spacing out practice in time to promote long-term retention), and metacognition (reflecting on one’s own understandings and problem-solving strategies).

Most strikingly, 97 percent of respondents thought that students can be categorized into one of several learning styles, and 77 percent thought that students are either right-brained or left-brained, even though both notions have been definitively debunked by research. As for where teachers learn about new education research and evidence, 67 percent of respondents cited conferences and workshops among their top three sources, 59 percent cited professional development, and 53 percent cited peers.

One potential limitation that Boser doesn’t address is that the study doesn’t take into account the geographic regions of the respondents. These educators predominantly rely on conferences, workshops, professional development events, and their peers for their information on teaching and learning—all of which are based on locale. It is important to know whether most of the respondents, for example, work in different states across the country or come from the same handful of school districts. If the latter holds, the findings may not be nationally representative of the educator workforce and could under- or overestimate how much K–12 educators in the U.S. actually know about the science of learning.

Nonetheless, this study is the first to examine K–12 teacher knowledge of research-supported learning strategies. The key takeaway for everyone involved in teacher training, instruction, and development is straightforward: Provide accurate information on the science of learning principles to combat these widely accepted myths and misunderstandings about teaching and learning. Insofar as educators then use this knowledge to inform their own instruction, teacher effectiveness will improve as a whole.

SOURCE: Ulrich Boser, “What Do Teachers Know About The Science of Learning?” The Learning Agency (September 2019).

On this week’s podcast, Mike Petrilli talks with Checker Finn and Andrew Scanlan about their new book on the past, present, and future of Advanced Placement. On the Research Minute, Amber Northern explores why Texas’s fracking boom was a bust for its schools.

Amber’s Research Minute

Joseph Marchand and Jeremy Weber, “How Local Economic Conditions Affect School Finances, Teacher Quality, and Student Achievement: Evidence from the Texas Shale Boom,” Journal of Policy Analysis and Management (August 2019).