No one has studied personalized learning more closely than the RAND Corporation. The group鈥檚 researchers are in the midst of two big studies: an ongoing look at 40 personalized-learning schools that have received grants from the Bill & Melinda Gates Foundation and an examination of 10 high schools that have been redesigned, in part to focus on greater personalization, with support from the Carnegie Corporation of New York. (91制片厂视频 Week receives grant funding from both the Gates Foundation and the Carnegie Corporation.)
For this special report, 91制片厂视频 Week visited RAND鈥檚 Pittsburgh office for a 90-minute face-to-face conversation with Laura Hamilton, the associate director of RAND education; John F. Pane, a RAND senior scientist and the group鈥檚 distinguished chair in education innovation; and Elizabeth Steiner, a senior RAND policy analyst.
The focus of the conversation was on what K-12 educators and policymakers need to know about the research on personalized learning. The six big takeaways:
1. The Research Behind Personalized Learning Is Still Very Weak
RAND has found some positive results, including modest achievement gains in some of the Gates-funded personalized-learning schools.
But overall, 鈥渢he evidence base is very weak at this point,鈥 Pane said.
鈥淚f anything, I worry that the positive results that have come out of our studies are generating a bit too much enthusiasm. I think people see the headlines, but they don鈥檛 see the limitations of the research that鈥檚 happened so far.鈥
For example, all the schools included in the studies received extra funding to implement personalized-learning models, and many are charters or part of large-scale school redesign efforts. RAND has also been unable to form the kind of strong comparison group used in the most rigorous research studies. And the only outcome measures the researchers have been able to analyze so far are select math and reading scores鈥攏ot anything about student behaviors or mindsets, or the so-called 鈥21st-century skills鈥 that are at the heart of many personalized-learning models.
As a result, Pane said, the evidence to date is not sufficient to draw clear conclusions about personalized learning鈥檚 effectiveness or how well it will transfer to other schools.
鈥淪omebody who鈥檚 trying to make evidence-based decisions has to wait,鈥 he said.
2. It鈥檚 Still Hard to Say What Personalized Learning Is (And Isn鈥檛)
The RAND researchers agreed most personalized-learning models share some broadly common elements, such as a greater focus on meeting individual students鈥 needs and a lesser focus on keeping students on pace with grade-level standards.
The Gates-subsidized schools also share a commitment to such practices as building profiles that capture what each student knows and emphasizing competency- or mastery-based progressions in which students only advance to new material after they have mastered what comes before.
But beyond that, Steiner said, there鈥檚 a tremendous diversity of approaches. 鈥淲hat鈥檚 happening in the field right now is a lot of innovation and a lot of schools building their models, building their curriculum, and inventing new systems,鈥 she said.
That means tremendous variation from one personalized-learning school to the next, including the curricular materials being used, how classrooms are organized, the role of the teacher, how data are used to group students, and how a concept like 鈥渕astery鈥 is defined.
The result, Steiner said, is that 鈥渋t can be difficult to be precise about what exactly a [personalized-learning] school is doing, what its model is, and what particular instructional practices are being implemented.鈥
3. On the Ground, Personalized Learning Also Faces Real Practical Concerns
The biggest is a lack of time.
鈥淲e heard from a number of teachers that [personalizing education for every student] was time-consuming,鈥 Hamilton said. 鈥淭eachers like having the opportunity to be creative and to have some autonomy over what they do in the classroom, but they also want to be supported in that.鈥
It also doesn鈥檛 help that the building blocks of personalized learning aren鈥檛 fully in place yet. A particular challenge: the lack of high-quality curricular resources needed to customize every lesson and activity to a wide range of grade levels and preferences.
And the adults in personalized-learning schools report being torn between competing priorities.
How do teachers, for example, encourage collaboration when every student is working on a personalized lesson, at his or her own pace?
And there鈥檚 an inherent tension between letting each student progress at his or her own pace and making sure every student learns everything in the curriculum and moves toward graduating in four years.
鈥淧rincipals and teachers really struggle with this,鈥 Steiner said. 鈥淚 don鈥檛 know that anyone has figured out a good solution.鈥
4. Still, There Are Reasons to Be Encouraged
It would be premature to conclude that personalized learning doesn鈥檛 work or should be stifled, Pane said. The theoretical underpinnings of the models currently being tested make sense, he said, and the field needs more time to let the current round of experiments play out.
Hamilton concurred. 鈥淲e鈥檙e seeing thoughtful implementations,鈥 she said, and the current shortcomings in the field can be attributed in large part to a 鈥渓ack of experience鈥 and a 鈥渓ack of adequate supports, particularly at the classroom level.鈥
5. The Worst Fears About Personalized Learning Aren鈥檛 Frequently Realized
Inside the schools that RAND is studying, scenes of children wearing headphones and working alone in front of screens are 鈥渓ess common than people might believe,鈥 Steiner said.
More typical is 鈥渆ver-changing classroom organization where the teacher is sometimes working with large groups and sometimes working with small groups or individuals, and technology plays a role in that.鈥
The rampant collection of student data feared by some critics also does not appear to be the norm in the schools that RAND is studying.
The researchers were careful to say that they couldn鈥檛 speak in detail to the back-end data-collection systems employed by the technologies in use in the classrooms they observed. But 鈥渆ven if systems are recording, say, keystroke information, we haven鈥檛 seen anyone using that,鈥 Hamilton said. 鈥淚t鈥檚 not being assembled into a report that goes back to teachers that they can make instructional use of.鈥
6. The Rand Team鈥檚 Bottom-Line Advice
Address the 鈥減ockets of resistance鈥 to personalized learning now, Pane suggested. Right now, it seems, key stakeholders, especially parents, might feel excluded from the process, which could impede success.
Recognize that the success of personalized learning will hinge largely on teachers, advised Hamilton.
And remember that improving schools is about a lot more than you and your new personalized-learning model, Steiner said, citing other critical factors such as establishing a cohesive schoolwide vision and building an effective team that enjoys working together.
鈥淭here鈥檚 a lot of focus on the shiny new parts [of personalized learning], like the technology,鈥 Steiner said. 鈥淏ut there are a lot of other things that go into making a good school, and those should not fall by the wayside.鈥