Personalizing Education at Scale

Personalizing Education at Scale for the University of Michigan’s Third Century

The modern research university was born amidst the 20th century’s massive industrial expansion. Michigan helped to create this model, with enrollments growing from 3482 in 1900 to 43,683 in post-war 1950. This modern industrial education supports occasional apprenticeship with experts by delivering many classes at astounding scale. Introductory courses across the curriculum have become enormous; typically enrolling hundreds, but occasionally thousands of students. In these courses, instruction often follows the lead of Henry Ford: students are treated as identical, each hearing the same lectures, receiving the same advice, doing the same assignments, and taking the same exams. The educational systems of the 20th century can handle scale, but only in an impersonal, mechanized way.

The 21st century began with an information revolution. It promises to change higher education as dramatically in this century as industrialization did in the last. What happens in the classroom has already shifted, moving away from simple delivery of content. But the real revolution will come when we harness information technology to personalize education. It is now possible to gather rich and extensive information about each student’s background, interests, goals, activity, and current status. Technology supported personalization will enable us to act on this knowledge; to optimize education at scale, create much greater students engagement, and accomplish more with less.

What is the transformation we envision? Today, most of the information, advice, and encouragement that students receive from the university is generic: completely unaware of their personal circumstances. During the next decade, this will change. All of our technologically mediated interactions with students will be individually tailored, continually refined by expert faculty and advisors, and delivered using methods known from behavior change research to be effective. Michigan will lead the nation in using technology to personalize education at scale. Students will recognize this, and will come here in part because we provide this highly personalized experience.

Personalizing education in two spheres

Classroom Education: In classes, we expect instructors to personalize the educational experience, responding sensitively to what they know about each student and their work. This can work magic. But only when an expert, concerned instructor of a small class has adequate information about their students. When these conditions can’t be met, we should use technology to tailor our interactions with students. The goal is to emulate what an expert instructor would do in an artisanal teaching environment, but to do it at scale. This is already happening.

For two years, we have used the E2Coach system to support 1900 students a term in all four large introductory physics courses. Last fall, we expanded our support to include introductory statistics, chemistry, and biology, offering the system to more than 5200 students. Personalized support has a clear impact: E2Coach users significantly outperform their peers in these courses, even after careful accounting for differences in their background and preparation. This project is a unique collaboration between behavior change experts and STEM educators; its existence provides a key advantage in our effort to play a national role in personalized education.

E2Coach was born with support from the Gates Foundation’s Next Generation Learning Challenge. Our recent expansion received support from both the National Science Foundation and the UM Learning Analytics Task Force.

Academic Advising: Beyond the classroom, the University already acknowledges the unique needs of every student by offering one-on-one academic advising both at the college and department level. Personal advising has a huge impact, but here too scale imposes severe limits. General advisors at Michigan oversee more than 300 students apiece. While every student meets an advisor at orientation, many never return, and in practice, advisors spend much of their time on a small subset of students. Just as in the classroom, academic advising will be greatly enhanced when we employ technology to gather a portrait of the background, current state, and goals of each student and use this information to tailor the feedback, encouragement, and advice they receive.

Our Third Century Discovery Proposal: Three Tasks to Prepare for Transformation

1: Build an interdisciplinary team to plan and prepare to spread personalized education
To guide Michigan’s move into personalized education, we will bring together an interdisciplinary team to design a transformation project. During this Discovery phase we intend to develop this team through a series of five monthly workshops facilitated by CRLT. These workshops will bring together faculty and staff from the existing E2Coach, Learning Analytics, behavior change, academic advising, and education research communities. This team will help to review the existing E2Coach classroom tools and assess their impact, develop a strategic and technical plan for using tools like this at a transformational scale, and collectively design an initial advising application. This process of planning and design will be the primary product of this Discovery phase.

2: Make technology decisions needed for spreading personalized education widely
Our existing E2Coach applications are built on the open-source Michigan Tailoring System created in SPH over the last 20 years. This powerful research platform has allowed us to accomplish what we have, but it has become clear that a dramatic expansion requires modernized tools. During this Discovery period we will complete a technology roadmap which will clarify what is needed to go to scale.

3: Pilot the use of E2Coach for Academic Advising
So far, all of our work with E2Coach has focused on supporting students in large introductory STEM courses. But academic advising is an even better target for tools like this. During this discovery phase, we will test E2Coach for academic advising by building a tool to support students during freshmen orientation. We will test our approach on the 500 students entering the Honors Program in summer 2014, with the intent to expand to all incoming students during the summer of 2015.

Project Team:
Tim McKay, Physics, College of Literature, Science and the Arts