Facebook CEO Mark Zuckerberg is a big advocate of personalized learning—using technology to tailor learning to kids’ interests, abilities and pace of work. Now his philanthropic investment company is funding a major initiative to help struggling readers by applying personalized learning to children’s literacy.
The Chan Zuckerberg Initiative (CZI), which Zuckerberg co-founded with his wife Priscilla Chan, is giving a $30 million grant to the Harvard Graduate School of Education and Massachusetts Institute of Technology’s Integrated Learning Initiative to launch Reach Every Reader. The program is a five-year initiative to build a web-based screening tool that diagnoses reading problems before kids can even read, and to develop a set of home and school interventions that personalize literacy support for kids, parents, and teachers.
According to research, a child who fails to read in first grade has a 90% probability of reading poorly in fourth grade and a 75% probability of reading poorly in high school. Literacy programs to help struggling readers are abundant, but it’s not clear why some work and others don’t, or which kids respond best to what program.
The problem, City says, is that a lot of kids struggle with reading early on, and then they keep struggling. “It doesn’t work itself out—we act like it does,” says Elizabeth City, executive director of the program and a senior lecturer at Harvard’s Graduate School of Education. “Don’t wait for them to fail. Try to predict who will struggle to read and then learn enough about what works for which kids in what context to match interventions to learning profiles.”
How it works
Reading is the basis upon which a lot of learning is built. So it’s bad news that, according to the Progress in International Reading Literacy Study, an assessment given to fourth-graders in schools around the world every five years, the US fell from fifth in the world in 2011 to 13th in 2016, with 12 places outscoring the United States by statistically significant margins. The worst-performing students posted the largest losses, with the average score for the bottom 25% of students falling nine points over the five years.
The first step for Reach Every Reader will be to better diagnose the problem, beyond just identifying that a child is having problems. This can be done even before a child can read, testing things like phonological awareness, or sound awareness, which are good predictors of whether a child will have reading problems, says John D. E. Gabrieli, director of MIT’s Integrated Learning Initiative, who will focus on making the screening tool.
“We have increasing evidence that we can identify kids who will struggle to learn to read and overwhelming evidence that the earlier you help a child with a reading difficulty the more they benefit form that help,” he says.
The group’s will create a web-based assessment that will initially be given to about 2,000 students, identifying issues like poor phonological processing (related to memory and retrieval), vocabulary problems, and listening comprehension. Struggling readers will be identified and tracked into the future to see whether the interventions they get actually help, and whether the original diagnosis was correct. Non-struggling readers will also be followed to see if they start having problems later on, a sign the screening test potentially missed them. By doing this every year, the researchers should start to identify—with data—what works and what doesn’t, and make improvements along the way.
Once a child has been diagnosed, the personalized learning kicks in. The group will develop and deliver two interventions: one for kids and caregivers who are reading at home, and one for kids and educators. Both interventions will have an app that’s personalized for children and adults.
Children can choose the content they want—reading about athletes, musicians, or scientists, for example. The app will also deliver content tailored to the reading problems identified in their screening.
Adults get personalized content too, focused on how to support children’s learning. These could include tips on dialogic reading (asking children questions and letting them help tell the story), as well as tips on the importance of pointing to words and images or making noises. They can choose to access the information in the way they prefer, whether through reading, listening, or video.
These actions are important because learning is social, especially so in the early years. The app aims to help the teacher or parent, as well as the child, to bolster that social experience. “The more we understand about the developing brain, the clearer it becomes that children need interaction; they are constantly learning, but they need adults and voices and interactions for that learning to take place,” writes Perri Klass in the New York Times.
City says their plan tries to address two key problems which happen once a child is identified as having problems with reading. First, parents often don’t know how to help their kids. Second, educators receive little support on how to implement interventions they are meant to use.
Yet another literacy program?
Not all literacy efforts succeed. The state of Michigan spent $80 million to improve third-graders’ reading abilities, which have continued to decline, according to the state’s standardized test, M-STEP. And another key reason kids may struggle to read is the persistent stress that comes with poverty, which impairs developing brains and can harm a child’s ability to learn. Unfortunately, no app can fix that.
In addition, the evidence on personalized learning remains thin, and plenty of critics say its benefits are overblown. Some researchers who study it say their work is being overhyped by advocates, and people who actually work in education bristle at Silicon Valley’s repeated efforts to reinvent education with more tech, especially since there’s limited evidence that it actually helps students. And more money is not always the answer to the problem: Zuckerberg’s $100 million investment in Newark’s public schools, for example, produced decidedly mixed results.
But City says you do not have to be a personalized learning advocate, or an ed-tech diehard, to get behind Reach for Readers. “We lead with the idea that the best way for kids to learn to read is with their parent and teachers,” she says. “It’s tech as tool, and not solution.”
This reporting is part of a series supported by a grant from the Bernard van Leer Foundation.