What does it mean to personalize education?

Here at the University of Michigan, we are busy educating more than 28,000 full-time undergraduate students. Every single one has a unique background, current state, and set of interests and goals. These factors evolve constantly from admission to graduation. To educate each of these students well, we must be able to attend and respond to these differences at every point: to personalize their experience.

There are several ways to make personalization happen. They all face real challenges: recognizing difference requires empathy and care, acting on it wisely requires understanding and experience. Analytics can help us all face these challenges.

One way to personalize is to let each student do it for themselves – give them agency and choice. At Michigan we offer thousands of courses which students may assemble into hundreds of undergraduate majors across fourteen undergraduate colleges. All this choice gives each student a chance to personalize: to choose courses and a major well matched to their background, interests, and goals, making sensible adjustments along the way.

Choice is great, but it’s not perfect. While students understand themselves well, they lack the experience and expertise to put that information to use optimally. Choice enables personalization, but doesn’t ensure it will happen well. We need to support their decision making, helping them to use evidence wisely as they make their choices.

Another way to personalize is for us, the faculty and staff of the university, to recognize difference, then change the way we interact with each student in response. This too presents challenges. For one thing, instructors are often ill-positioned and unprepared to personalize. It takes imagination and effort for a professor like me – 52 years old, tenured, long-settled in Ann Arbor – to appreciate the circumstances of an 18 year old away from home for the first time. Even when I can connect and understand a student as an individual, knowing how to respond well is often beyond me. Like the students, I lack the experience and expertise to really know what to advise. I haven’t walked in their shoes, or studied systematically those who have.

Then there’s the problem of scale; our students spend a lot of time in large classes. In small classes, there may be good opportunities to personalize. But it’s an especially grave challenge to look out at hundreds of faces in a lecture hall and remember that each brings together a unique lifetime of experience: more than a decade of schooling, a heritage drawn from family and community, an array of interests, passions, fears, and goals. Too often, large course instructors retreat to an industrial response, treating all students as if they were the same.

Students know themselves, but lack the experience or expertise to know what to do. Faculty and staff have some experience, but are stretched thin trying to understand and attend to every student as an individual.

How can we tackle this problem, let everyone learn from the experience of all, while responding to the unique circumstances of each? In 1995, when I started teaching, this was impossible. Today, analytics and information technology give us a chance.

What does analytics have to do with personalization?

At Michigan, we have extensive, accessible records of the experience of every student, from admission to graduation, going back twenty years. This is an incredible archive of experience. With the right tools, students, faculty, and staff could use this data to learn from the choices and outcomes of 100,000 students. They could see what pathways prior students have followed, learning from their successes and failures. They could explore course taking patterns within majors, see what courses successful students chose and when, and how they decided to meet breadth requirements. It’s not just typical pathways that count: you can learn a lot about what’s possible when you can examine this experience of this many students.

In principal, this sounds great. But there are challenges. We have to provide access to the lessons from this experience while protecting the privacy of our former students. This can be done, but needs care. Then too, most members of our community are not used to learning from extensive data like this. If we want people to learn from this experience both wisely and well, we need to provide support – experts ready to help them analyze and interpret all this complex information.

It’s worth noting that this kind of approach isn’t entirely new. University communities have always helped students make decisions based on the experience of others. Social networks have always connected beginning students with older peer mentors. Experts like academic advisers have always digested the experience of many and presented it as guidance for new students. What’s new here is the ability to dramatically increase the scale and rigor of the process.

Now every student can learn from the experience of all former students, not just the few they meet. Now we can ensure that inferences drawn by everyone from this information meet the highest possible analytic standards.

This approach, helping everyone learn from the experience of all, does a good job of empowering individuals through agency and choice. But how can we use information technology to personalize the ways faculty and staff support individual students, even in large classes?

Imagine a campus where every student receives support from a team of expert coaches, each fully aware of student backgrounds, interests, goals, and real-time status, every coach armed with the ability to learn from the experience of all former students and supported by experts in motivational communication and behavior change. Sounds good, no?

Pioneered for digital health coaching, computer-tailoring systems like ECoach can make this vision a reality, helping experts provide thousands of individuals with feedback, encouragement, and advice tailored to respond to their specific background, goals, and current state. ECoach lets someone like me decide what I would say to a student, if they visited me in office hours, then deliver those messages, whenever I like, to all of my hundreds of students. More than that, tools like this allow teams of students, faculty, and staff to bring together their collective experience around a topic like completing a challenging class, making the transition from high school to college, choosing a major, exploring an internship, or seeking a job and deliver it to every student who might benefit, whenever they need it.

This is what personalization is about: building a system where every student is able to learn from the experience of all, using their own agency and choice to optimize their experience, where experts from the community can share what they know in tailored, effectively communicated ways, even at scale.

Information technology and analytics make this all possible. Personalizing education for the 21st Century is one of the main goals of the University of Michigan’s Digital Innovation Greenhouse. If you have thoughts or ideas about how to do this better, please get in touch.