Chan Zuckerberg Backs Personalized Learning R&D Agenda

Jim Shelton leads education at the Chan Zuckerberg Initiative (CZI), a new impact organization that seeks to advance human potential and promote equal opportunity through a combination of grants, investments and advocacy. By making grants and equity investments, CZI is advancing personalized learning.

In education, CZI aims to support models that take a whole child approach to personalized learning – taking into account a child’s social, emotional, mental and physical needs in addition to their academic growth.

With Bror Saxberg, Vice President of Learning Science at the Chan Zuckerberg Initiative, Jim addressed the Fusion conference on Friday. He asked the audience, which included leaders from the League of Innovative Schools, to hold two ideas in tension: the full potential of every student and how much we have to learn about learning. He suggested thoughtful, humble leadership.

“Let’s be thoughtful about and humble about what people have done before, the things that we can stand on, and then how we’re going to continue to get smarter and smarter as we go,” Shelton said.

Shelton is motivated by the two sigma opportunity. Benjamin Bloom reported in 1984 that the average student tutored one-to-one using mastery learning techniques performed two standard deviations better than students who learn via conventional instructional method.

The core question of personalized learning is how to scale that kind of two sigma benefit. “How do we create these kind of learning experiences and these kind of learning environments at a scale, at a cost we can afford?” Shelton asked. The purpose, according to Shelton is to “Help each child develop competencies and skills to have a fulfilling life.”

Personalized learning is about designing tailored lessons, said Shelton, but also being in tune with how students walk into that class given the importance of mindset and motivation. “If a child does not know how to be resilient or persistent or to self-direct or self-manage, how can we possibly expect them to achieve at their maximum potential academically?” Shelton asked.

Saxberg added that we can’t treat students the same way; each brings a unique motivational profile and emotional state–each suggesting different interventions.

CZI is taking a “whole child” approach, said Shelton but we “have to know if our kids can read, write, and compute.”

On testing, Bror recommended evidence centered design–thinking about the evidence that can be gathered about what students know. He supports the use of all kinds of evidence about students’ interactions with their learning environments and each other. The problem is that good performance assessment and evidence-gathering is expensive without technologies to help. Bror sees opportunity for using automated scoring and other technologies to help make performance-based evidence-gathering easier and more affordable.

Shelton added, “We must be able to use assessments to tell us how to actually continuously improve what we do every day with them,” speaking of the need to provide educators information that they can use in real time. He went on to say, “The reality is that we have been trying to please the test and work towards the test for over a decade, and it’s not working for us.”

Personalized learning is about helping students do the best next thing not the easiest. “Novice learners are quite bad at choosing what to do next,” said Bror noting novice learners pick activities that make them feel confident and comfortable. A very important role of teachers is recommending the most valuable next steps, not the easiest. This doesn’t mean removing choice – it can include giving choices to students among things that have equal learning benefit but of different interest to different students. That interest can then fuel the learning effort needed.

Technology, as Shelton and Bror described it, augments and supports the work of customized feedback for students to learn more efficiently. But fundamentally, they underscored the importance of human relationships, based on a radical shift in focus, toward the child. The possibilities can be expanded when we are fundamentally focused on the needs of the child and a new vision of his or her potential.

Shelton mentioned Montessori as an example of preschools with a set of guided choices (see recent research confirming Montessori benefits). At Montessori schools, children are given guided choices and then receive support and guidance as they pursue various learning activities. In the U.S., we just have not been really good at finding that correct balance, and Shelton argues that we definitely haven’t figured out how to do it for every single child.

On EdTech, Saxberg said we should think in terms of a “human technology system” for learning, determining which things expert people are most suited to doing, and which things can best be done by machines. And then look at how else tech can help and make solutions more affordable and scalable.

It is important to have data to inform the practitioner on the next best thing to do said Saxberg. “Responsible use of data includes a tension between privacy and the moral imperative to improve learning for each learner as we gather more, better evidence,” said Bror. It’s not simple, but there may be ways to work these tradeoffs better than we are doing now.

Saxberg decried the lack of learning engineers to address the technical challenges of practice inherent in personalized learning. He suggested reading the NRC study How People Learn and

Ruth Clark and Richard Mayer’s e-Learning and the Science of Instruction: Proven Guidelines for Consumers and Designers of Multimedia Learning.

On what to do next, Shelton suggested broadening the definition of success, sharpening our hypothesis about how we get there, and working differently so that it works for every child. “Try it, look at it, build the discipline of getting better faster.”

More from Bror

After Fusion, I asked Saxberg to expand on his vision for whole child development. “Promoting a multi-dimensional trajectory of learners, including social-emotional learning and mindsets, is needed – it is not enough to only look at disciplinary progress,” said Bror. There will be  a good deal of integration challenges to get there.

Bror said that beyond the specific competencies that will fuel many careers, there are two emerging capabilities that will be key: empathy and metacognition.

As the automation economy eats away at routine work by people, the ability to “give each other meaning” will grow in importance, said Bror. At different life stages, through struggle and in times of choice, Saxberg stressed the value of our ability to contribute to others by mentoring, appreciating, and understanding – things that are less likely to be replaced by machines in the near term.

In addition, for the first time in history, said Bror, humans cannot live out their lives by developing a single expertise. Life with smart machines means every job changes at the speed of Moore’s Law. Bror said that requires becoming metacognitive about our skills and interests and what’s required by a changing economy – and having access to reliable systems for changing our skills, and the will to do so regularly.

“The research on learning is extremely clear, how minds actually work. It’s the practice and feedback in well-designed instruction with good coaching that actually changes your brain,” said Bror. “I mean, literally it alters the neurons. Learning is really about how to get well-designed practice and feedback environments available to motivated learners.”

Because we’ll all be working as part of teams augmented by information-rich appliances, Bror added that it requires us to be metacognitive about our roles on a team and the team’s health and effectiveness. (Read more about the science of teams.)

Developing empathy, metacognition and employment skills implies a rich set of experience integrated with practice and feedback, “with an eye toward human and professional skills for contribution,” said Saxberg.

Young people would benefit from exposure to the full stack of human experience in place-based education (#PlaceBasedEd) that could include a combination of travel, simulation and virtual reality to quickly develop global competencies.

Weaving “next best” student choices into a well coordinated curriculum of rich experiences with frequent feedback points to the critical role that teachers will continue to play in personalized learning.

The CZI R&D Agenda

CZI is focused on four key milestones: kindergarten readiness; 3rd grade literacy; smooth transition to high school; and postsecondary success. They provide significant support to:

• Summit Learning Platform is a free personalized learning platform created by Summit Public Schools. Over 100 teacher teams use the platform and benefit from the training and support. CZI engineers work closely with educators at Summit and schools around the country to understand their needs and help improve the platform.

• The College Board is expanding access to rigorous college prep experiences including customized SAT practice through Khan Academy and AP computer science courses and peer advising through the National College Advising Corps.
• Vision To Learn is a nonprofit that provides free eye exams and glasses to tens of thousands of low income children around the country.

CZI also supports a science agenda including:

• The Chan Zuckerberg Biohub: In September 2016, Mark Zuckerberg and Priscilla Chan committed $600M over 10 years to fund the Chan Zuckerberg Biohub, an independent nonprofit research center that brings together physicians, scientists, and engineers from UC Berkeley, UCSF, and Stanford to encourage collaborations between these universities.
• Human Cell Atlas: The Human Cell Atlas is a global effort to create a reference atlas of all cells in the healthy human body as a resource for studies of health and disease. This tool will help scientists and physicians understand how normal cells function, and what goes wrong when disease strikes.
• Supporting knowledge environments: The Chan Zuckerberg Initiative aims to accelerate the sharing and awareness of scientific knowledge by funding strong existing programs and building tools to advance the field. As part of these efforts, the Chan Zuckerberg Initiative acquired Meta, a platform that uses artificial intelligence to help scientists read, analyze, prioritize, and draw insights across millions of scientific papers. This tool will be made available to researchers around the world at no cost.

For more see:

• The Platform Revolution That Will Power Personalized Learning
• Virtual and Augmented Reality in Personalized Learning
• Advancing Human Potential: The Power of Personalized Learning