Personalized learning has a big problem.
Inside America’s schools, the term is used to mean just about anything.
Algorithm-driven playlists? Grouping students based on digital data? Letting teens design projects based on their personal interests? Adaptive software that adjusts to each student’s skill level? Customized activities to help kids develop a growth mindset?
Check, check, check, check, check.
“In the same way that Inuits have lots of different words for ‘snow,’ I think these are all personalized learning,” said Larry Berger, the CEO of ed-tech company Amplify.
For some, such variety is reason for optimism. Common threads, such as a focus on giving students more say over their own education, run through the different versions. In addition, Berger said, the disparate strands of personalized learning are bound together by what they reject: classrooms in which all students listen to the same lecture, complete the same worksheets, take the same quiz, and get the same limited feedback on their learning.
But skeptics warn such loose explanations could lead the movement down a well-worn path toward incoherence and ineffectiveness.
“The history of school reform is dotted with the debris of earlier instructional reforms that varied greatly in definition and practice,” wrote Stanford University education researcher Larry Cuban in a Septemberoutlining personalized learning’s chameleon-like nature.
To get a better sense of just how diffuse personalized learning is in practice, Education Week spoke with more than a dozen leaders in the field, including district administrators, school principals, officials from prominent personalized learning networks, researchers, advocates, and CEOs. The conversations focused on comparing and contrasting various personalized learning environments across three key dimensions: what a typical day looks like, the underlying philosophy of how children learn, and who makes decisions in the classroom.
Those experts described a national landscape in which as many as 1,000 schools are trying to fundamentally reshape how they operate, while tens of thousands more are trying out new strategies within their existing structures.
It’s all being called personalized learning. But there’s tremendous variation, sometimes within a single building.
Take, for example, Chicago International Charter School West Belden, which is managed by Distinctive Schools. Over the past five years, the school’s approach to personalized learning has morphed from a focus on flexible classrooms and multi-age student groupings, to incorporating “competency-based progressions,” to using the Summit Learning Platform, a software program developed with support from Facebook and Mark Zuckerberg.
“We don’t believe that personalized learning is any one thing,” said Principal Colleen Collins. “It’s a mindset.”
That’s as good a summary as any of where the nascent field stands.
But what does such ambiguity mean for K-12 leaders around the country, many of whom are considering whether and how to bring personalized learning to their own schools? Here are three questions that can help educators and policymakers look beyond the buzzwords.
What Does a Typical Day Look Like?
Each day in San Jose, Calif., about 15,000 elementary students use a popular software program called. Math class starts with 20 minutes of whole-class instruction. Then students are split into small groups. For the next 40 minutes, they cycle between time with a teacher, time working collaboratively with peers, and independent work on Dreambox.
Jessie Woolley-Wilson, the company’s CEO, says that if schools are using her product, they’re doing personalized learning. That’s because the software is “intelligently adaptive"—it helps students who are behind catch up and lets students who are ahead move forward. If the software notices that a student is using an inefficient procedure to solve a particular problem, it might cut in with a targeted lesson.
But that’s dramatically different from life inside the country’s 65high schools. There, personalization is about tapping into students’ passions through real-world internships. Relationships are cultivated through small advisory groups that stick together for four years. Tests are replaced with “public displays of learning” tied to students’ own interests.
Personalized Learning: What Is It? (Video)
Ask a dozen educators to define personalized learning and you’re bound to get 12 different answers.
And both differ in fundamental ways from the 39 schools, spread across the U.S., using a middle school math program called. Here, classroom walls are torn down and new bell schedules are created so that 60 or more students can use a single large space for up to 90 minutes a day. Every afternoon, algorithms analyze each student’s performance and abilities, then recommend schedules and instructional plans for teachers and students alike to follow the next day.
The term “personalized learning” is commonly applied to all three approaches. But it doesn’t shed much light on the different activities that are actually occurring in each classroom—or whether there’s evidence to back them up.
“The right question is not, ‘Does personalized learning work?’ ” said John Pane, a senior scientist and the distinguished chair in education innovation at the RAND Corporation. “At this stage, we really need to be looking at the actual components of what people are trying to do and assess them each on their own.”
What’s the Philosophy of How Children Learn?
Longtime observers say these various versions of personalized learning are shaped by a range of underlying philosophies.
At one end of the spectrum are approaches that Berger, the Amplify CEO, describes as an “engineering model” of learning, predicated on the notion of an established body of knowledge that schools should help students master as efficiently as possible.
Here’s how he described this model in apublished by Education Week in February:
You start with a map of all the things that kids need to learn.
Then you measure the kids so that you can place each kid on the map in just the spot where they know everything behind them, and in front of them is what they should learn next.
Then you assemble a vast library of learning objects and ask an algorithm to sort through it to find the optimal learning object for each kid at that particular moment.
Then you make each kid use the learning object.
Then you measure the kids again. If they have learned what you wanted them to learn, you move them to the next place on the map. If they didn’t learn it, you try something simpler.
That philosophy animated Silicon Valley’s early conception of personalized learning. It’s still “quite widespread” today, Berger said, particularly among testing companies trying to sell instructional products to schools.
At the other end of the personalized learning spectrum, meanwhile, are more “student-centered” approaches, rooted in the progressive notion that real learning happens when children are driven by their intrinsic curiosity.
Here’s how Kaleb Rashad, the head ofin San Diego, described the philosophy underlying his school:
Human beings are born to learn and create and make sense of their world. … I want my kid going to school where he’s able to learn alongside other people, pursue important questions, hold a sense of curiosity and exploration, and learn to reflect on what he’s tried. If your version of ‘personalized learning’ leaves kids feeling less imaginative, less willing to take risks, less willing to test their ideas, then I have questions about that.
So where do most examples of personalized learning fit?
Somewhere in the middle, said Cuban, the Stanford researcher.
Take, for example, the 11 charter schools operated by California-based. (The group is also behind the Summit Learning Platform, which is now in use by more than 380 other schools, serving more than 72,000 students around the country.)
Summit students generally take the same courses at the same grade level. All students are expected to master the same standards, and the curricula consist of a set menu of lessons, activities, and projects.
But within each lesson, students get targeted supports. They also have a fair bit of choice—around which topics to pursue within a given project, for example. And because Summit believes “social-emotional” development is key to learning, students work with mentors and teachers to track their individual progress in such areas as “emotional intelligence” and “self-directed learning,” as well as reading and math.
“The philosophy is built on teachers, when they have the right information in front of them, being able to know best what students need,” said Andrew Goldin, Summit’s chief program officer.
Why is that belief system important to know?
The larger questions school and district leaders considering personalized learning should be asking, Cuban said in an interview, are: “To what end? Do you want higher test scores? Or do you want kids to learn to be independent decisionmakers?”
Who Makes Classroom Decisions?
Then there’s the question of “student agency.”
One of the more consistent ideas across the personalized learning field is that children should make more choices about their own learning.
But here, too, there’s tremendous variation on the ground.
Take, for example, Teach to One, the middle school math program in which algorithms play a crucial role in setting daily schedules, determining what students have learned, and selecting lessons and activities for them. For students in these classrooms, greater agency generally means more flexibility to move at their own pace, as Joel Rose, who heads New Classrooms, the nonprofit group behind the program, described in an interview:
Let’s say our diagnostic shows that you don’t know how to factor trinomials, and you’re going to learn it next Tuesday. But you watched the videos and you think you’ve got it. If you take an on-demand assessment, and demonstrate that you do know it, that comes off your [playlist.] We give you the power to own your own learning and to demonstrate what you know, so we can use valuable classroom time to teach you the things you don’t know yet.
Compare that with Patrick Henry Elementary in Chicago, one of more than 120 area schools that has worked with a nonprofit called LEAP Innovations to implement personalized learning.
During 60-minute literacy blocks, each Henry student operates at one of four “levels of autonomy.” At the first level, students are told where to go and what to do. By level four, they have near-free reign to decide where they want to work—including in the hallways or outside the classroom—and flexibility in choosing what they work on. In a, Henry teacher Diane Meloscia describes how this approach altered the types of decisions made by everyone in her classroom:
I want [students] to be able to think about something before they do it. I want them to feel that ownership, to be able to start a task, know why it’s valuable, and then know how to get there. … That’s been the biggest change I’ve seen in our kids. They don’t ask me anymore. They try to figure it out, they ask a friend, they ask someone in their group.
A third approach to classroom decisionmaking can be found back at High Tech High, where each teacher “designs” his or her own curriculum and students pick the big questions they want to explore. There, cultivating student agency is about building in kids the capacity and desire to change the world around them.
“It’s not, ‘Hey, good job, you mastered this skill,’ ” said Rashad, the school leader. “It’s, ‘Whose life have you made better?’ ”
If those approaches to classroom decisionmaking seem to be all over the map—well, that’s because they are, said Andy Calkins, the director of, a prominent competition that awards schools grants for adopting personalized learning. (Calkins is a co-author of Next Gen Learning in Action, a blog hosted on .)
One set of reasons for the variability is structural. Age-graded schools, state standards, high-stakes tests, district pacing guides, and the mix of school improvement strategies any one school is likely pursuing at any given time all work as powerful counterweights to truly letting students shape their own educational experiences, Calkins said.
There’s also the tension between personalized learning proponents’ enthusiasm for boosting student agency and what the research says about it.
“This idea about having the learner make their own decisions about the best way to learn something has not been proven to be productive,” said Pane of the Rand Corp.
Finally, there’s the ongoing push and pull over the roles of humans and technology in personalized learning environments. The debate has shifted considerably over the past few years, but remains unsettled, contributing to the confusion in the field.
Take, for example, AltSchool, founded in 2013 by two ex-Google employees. The company quickly raised more than $175 million in venture capital, largely on the strength of a, using it to create algorithm-driven recommendations and playlists, and merging them with the judgments of teachers.
But recently, said chief impact officer Devin Vodicka, the company has pivoted, embracing a new set of beliefs about how classroom choices—from student goal-setting to assigning lessons—should be made. Now, there’s a big focus on reminders and prompts, rather than automated decisions or algorithmic recommendations.
“What we are doing with technology, which we think is compelling, is more behavioral nudges to encourage the humans to be optimizing their decisionmaking,” Vodicka said.
Maybe that will be the idea around which the nebulous world of “personalized learning” finally crystallizes.
Or perhaps it will be any of the myriad approaches currently playing out in schools.
But for now, personalized learning continues to mean a little bit of everything, and nothing in particular.
And that’s leaving K-12 at the center of a blizzard, with a limited vocabulary for describing the new practices, old philosophies, and big questions swirling around them.
Ask a dozen educators to define personalized learning and you’re bound to get 12 different answers about why this approach to learning has strong benefits and real risks for students.
A version of this article appeared in the November 07, 2018 edition of Education Week as Personalized Learning Still Means Whatever People Want It to Mean