The Distributed Classroom

The Distributed Classroom

Learning in Large-Scale Environments

Justin Reich and Nichole Pinkard

Writers in the Secret Garden: Fanfiction, Youth, and New Forms of Mentoring, Cecilia Aragon and Katie Davis, 2019

Peer Pedagogies on Digital Platforms: Learning with Minecraft Let’s Plays on YouTube, Michael Dezuanni, 2020

A Manifesto for Teaching Online, Siân Bayne, Peter Evans, Rory Ewins, Jeremy Knox, James Lamb, Hamish Macleod, Clara O’Shea, Jen Ross, Phil Sheail, and Christine Sinclair, 2020

The Distributed Classroom, David A. Joyner and Charles Isbell, 2021

The Distributed Classroom

David A. Joyner and Charles Isbell

The MIT Press Cambridge, Massachusetts London, England

© 2021 Massachusetts Institute of Technology

All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher.

The MIT Press would like to thank the anonymous peer reviewers who provided comments on drafts of this book. The generous work of academic experts is essential for establishing the authority and quality of our publications. We acknowledge with gratitude the contributions of these otherwise uncredited readers.

This book was set in ITC Stone and Avenir by New Best-set Typesetters Ltd.

Library of Congress Cataloging-in-Publication Data

Names: Joyner, David, author. | Isbell, Charles, author.

Title: The distributed classroom / David A. Joyner and Charles Isbell.

Description: Cambridge, Massachusetts : The MIT Press, 2021. | Series: Learning in large-scale environments | Includes bibliographical references and index.

Identifiers: LCCN 2020053035 | ISBN 9780262046053 (hardcover)

Subjects: LCSH: Education, Higher—Effect of technological innovations on. | Internet in higher education. | Web-based instruction. | Blended learning. | Open learning.

Classification: LCC LB2395.7 .J69 2021 | DDC 378.1/7344678—dc23

LC record available at https://lccn.loc.gov/2020053035

10 9 8 7 6 5 4 3 2 1

d_r0

Contents

Series Foreword

Preface on COVID-19

Part I Where We Are Now

1 The Classic Dichotomy

2 Place and Time

3 Progress So Far

Part II What We Do Next

4 The Distributed Classroom Matrix

5 Symmetry

6 Practical Considerations

Part III The Places We’ll Go

7 From Stopgap to Snowball

8 The Distributed Campus

9 Fears, Risks, and Other SCARY Words

10 Lifelong Learning for All

Acknowledgments

Notes

Index

Series Foreword

Learning across the life span is more important than ever, and with the wealth of resources and communities available online, there has never been a better time to be a learner. Learners of all ages—in formal and informal settings—are turning to online tools to help them develop new skills and knowledge, for work, school, and leisure. The field of large-scale learning engages in the study of networked environments with many, many learners and few experts to guide them.

Large-scale learning environments are incredibly diverse: massive open online courses (MOOCs), intelligent tutoring systems, open learning courseware, learning games, citizen science communities, collaborative programming communities, community tutorial systems, social learning networks, and countless informal communities of learners on platforms such as Reddit, YouTube, and fanfiction sites. These systems either depend upon the direct participation of large numbers of learners in a single instance, or they are enriched by continuous improvement based on analyzing data generated by many learners over time. They share a common purpose—to increase human potential—and a common infrastructure of data and computation to enable learning at scale.

Technologies for large-scale learning are sometimes built deliberately, as in the case of MOOC platforms, and they are sometimes adapted from technologies originally developed for other purposes, as in the case of video-sharing sites. In some cases, they are used by individual learners around the world, and in other cases, large-scale learning environments are embedded within more traditional and small-scale educational systems such as classrooms and schools. They can be used to foster human capacity and create new opportunities, but they can also be used to teach and spread hateful ideologies. With a capacious enough definition of learning, large-scale learning technologies are implicated in nearly every part of the human experience in the networked world, from schooling to professional learning to politics to health care and beyond.

The Learning in Large-Scale Environments series from the MIT Press seeks to investigate, critique, and explain these large-scale environments and the various ways they are hybridized with residential learning space. Just as large-scale learning environments are diverse, our series includes books with a diverse set of methodological and theoretical perspectives, ranging from learning science to computer science to sociocultural research traditions. The series examines large-scale learning environments at multiple levels, including technological underpinnings, policy consequences, social contexts and relationships, learning frameworks, and the experiences of educators and learners who use them. Our hope is that researchers will find valuable contributions to the scholarly literature and that educators and policy makers will find useful insights as they consider how best to support the learning needs of students of all ages around the world.

Justin Reich

Nichole Pinkard

Series Editors, Learning in Large-Scale Environments

Preface on COVID-19

Over the last year, we have seen an unprecedented rapid shift to remote, online, and hybrid learning driven by the COVID-19 pandemic. Schools and universities around the world have had to quickly figure out how to teach online.

Much has already been made of this emergency change. Skeptics of remote learning have urged the community to treat this as a temporary shift, noting the numerous perceived weaknesses of online education.¹ Advocates have looked at this as an opportunity to speed up the shift to more remote options, noting its benefits to cost, access, and scale.² Students have spoken up as well: many have regarded the online experience as so inferior as to be worth lawsuits seeking the return of tuition dollars.³ While some students have reflected positively on the transition,⁴ the overall impression from students, parents, and teachers has been largely negative; the push to remain remote, partially or completely, throughout fall 2020 and into spring 2021 came largely from health concerns, not faith in online learning.

During the early part of the pandemic (at least in the United States), this emergency shift was been toward exclusively remote education: all students were online, all classes were online, and there was no in-person experience. One complete experience was replaced with another. In some ways, this simplifies the transition: classes are replaced one-to-one with remote analogues. While the world of online education is new to many faculty and students, it is not entirely new on its own: distance learning has existed for over a century, and modern technology has moved it more into the mainstream than ever before, with most colleges offering some online classes and programs in a typical distance-learning model. Under this model, the assumption from the outset is that all students are remote. This is how we in Georgia Tech’s College of Computing run our online Master of Science in Computer Science (OMSCS) program. Launched in 2014 under the leadership of then Dean Zvi Galil,⁵ we have grown to over eleven thousand students from all around the world, and none of them are required to come to campus. In addition to serving as the college’s current Dean (Charles) and the program’s executive director (David), we teach six classes between us, totaling over three thousand students each semester.

So for all the complexity involved in navigating new and unfamiliar platforms and technologies, there was also some simplicity in this transition: entirely in-person courses were transitioned to entirely remote courses. The two structures could be considered alternatives. There are design decisions to be made within each, of course, such as whether an online course should be synchronous or asynchronous, but the distinction between in-person and remote is clear.

As the pandemic dragged on, however, the situation became more complex. As colleges were planning for fall 2020 and beyond, they began facing decisions about remaining all online or pursuing a hybrid model. Under a hybrid model, there would once again be an in-person experience, but with heavy restrictions. Class sizes would be limited. Large classes might progress in small mini-sections, with each student spending a fraction as much time in the classroom. Some students and instructors might be unable to return to campus due to visa issues or medical conditions, requiring their own remote access. The threat of another sudden emergency shift would loom over course designs. Some classes would be offered online. Some students would attend online classes. Some students might exclusively attend online classes, while still others participate on campus. The number of combinations and permutations of the multitude of variables was mind-boggling, and it was difficult to give any individual teacher good advice on how to handle their unique class situation.

At the core of this difficulty, though, was the desire to bring back the in-person experience as much as possible. If the online experience was believed to be universally equivalent to the in-person experience, then students, teachers, and administrators would likely not hesitate to remain online for the foreseeable future. But there is a belief that the classroom experience is superior, and we would argue that belief hangs on that word choice: classroom. There are dynamics and experiences and interactions that occur in person that do not automatically translate to online learning because of the lack of a classroom. This criticism is not specific to remote learning either; the same objection is lodged at large lecture halls, again for lacking interactions we would more commonly attach to classrooms: conversation among peers, communication with teachers, personalized input on one’s work, and peripheral participation with a social learning community. There is an old adage that distance education starts in the third row which reflects this tension: students who are not sitting up front and actively engaged may as well be distance learners.

For years, this issue has plagued large core classes and upper-level classes for in-demand fields, but in the wake of COVID-19, it began to affect everyone: How do you build a classroom experience when you cannot gather all your students together at the same place and time? How do you create meaningful interactions—both direct and peripheral—between students and faculty and among students themselves without synchronous co-location? Why do so many of these remote learning experiments—and distance learning classes before them—lack the feel of a classroom?

This book presents a vision of a distributed classroom: a learning experience that retains that classroom environment but distributed across time and space. Rather than defining a classroom as a physical location used by a co-located group of students and teachers at a shared time, a distributed classroom looks at how characteristics of that environment may be shared with remote and asynchronous learners. It describes a view on learning that is not dependent on the typical same-time, same-place assumption of traditional education, but also one that does not swing all the way to the other extreme of full asynchronous remote learning. It seeks to fill in the entire spectrum of synchronicity and co-location from one extreme to the other, asking students only to make the minimum necessary trade-off needed to achieve a particular level of access to the learning experience. It is not merely a way of teaching remotely, but a way of proactively organizing the classroom experience to exist independent of other constraints. It strives to allow learners to participate regardless of location and time, but without the heavy compromises in interaction and assessment made toward that end by efforts like MOOCs.

More than anything else, the distributed classroom strives to present an optimistic, learner-centric view of remote learning and the future of education, one where every person on earth is turned into a potential learner as barriers of cost, geography, and synchronicity disappear. In the process, it reorients the discussion of online education from a separate mechanism for transmitting content to a new medium for extending an experience.

We developed these ideas in the context of continuity planning for COVID-19, but nothing about this vision is specific to the pandemic; we strongly believe it was coming in the next several decades. This vision provides a plan for designing educational experiences that allow us to survive the present, but also to prepare for the future. In addition, it provides guidelines for how this paradigm can be used to students’ benefit rather than the benefit of other competing demands.

Several parallel trends over the past decade have already laid the groundwork for these developments, from the proliferation of learning management systems to the emergence of new course formats like HyFlex and flipped classrooms.⁶ This vision brings them together into a cohesive, prescriptive blueprint for leveraging these trends to solve both old and new problems. For years, there has been a need for lower-cost, more accessible, and more distributed educational options at all levels, from kindergarten through adult learners. The student loan crisis in the United States, the vast discrepancies in learning opportunities based on socioeconomic status, and the concentration of institutions in a small number of countries all need to be addressed. These problems, however, are vague and distant, and there is enormous inertia behind the status quo. The problems presented by the pandemic, however, were clear and immediate, providing the sort of critical mass to realize major change in a short period of time. Just as mRNA-based vaccines like those developed for COVID-19 are now being created to address older viruses like influenza and HIV, so also progress made to make remote learning more widely available during the pandemic can now be pointed at older problems like inequitable access.⁷

The vision of the distributed classroom presents a way to address immediate concerns, but it is not merely a stopgap for the remainder of the pandemic; it is a vision of how education can and should work in the coming decades. COVID-19 has been a challenge, but also an opportunity to address these issues in durable ways that will retain their advantages after the crisis is over. The distributed classroom encourages us to look excitedly to the future rather than longingly at the past—and not to ask ourselves when things will return to normal but rather how we were ever satisfied with normal in the first place.

I

Where We Are Now

1

The Classic Dichotomy

Students’ enrollment in distance-learning classes and programs has been rising significantly over the past several years, bolstered by the growth of online education. According to data from the National Center for Education Statistics, in fall 2018, 34.7 percent of all postsecondary students were enrolled in at least one distance-education class, and 16.3 percent were enrolled exclusively in distance education classes.¹ At the K–12 level, most states now offer public online virtual academies, while there are also significant for-profit players in the space offering fully online class experiences. At the same time, however, many teachers, students, and employers remain skeptical of online education.²

We do not bring these studies up to highlight the growth of online education, which you likely already know about; nor do we bring them up to let you know that online education has skeptics, which you likely also already know. We bring this up to point out that research on online education frequently assumes a clear dichotomy between online and face-to-face. If your perception is based on the research, the national conversation, and the legal statutes, then it appears that face-to-face and online education exist largely separately. Asking the questions, How is online education perceived? and How have online programs grown? places online education into a separate, largely distinct category.

This distinction has far-ranging implications. For example, the US Department of Education’s guidelines contain dedicated definitions for interaction in online classes, mandating that they be regular and substantive, a term not applied to in-person classes and poorly defined as it pertains to new technologies.³ Financial aid programs often distinguish online classes and degrees from face-to-face programs; for example, the Pennsylvania Higher Education Assistance Agency State Grant (PHEAASG) is not available to students taking more than half of their classes online.⁴ Students may not be eligible to apply for visas to enroll in classes proven to be deliverable online, a point of concern during the rapid shift to remote education in the wake of COVID-19.⁵ Critically, these rules all rely on the existence of a clear distinction between online and face-to-face classes.

That clear distinction exists at different levels. At the highest level, we can juxtapose entirely online universities or school systems—like Ashford University and Florida Virtual School—with traditional universities and school systems. We can also segment that down, finding online degree programs within traditional universities; these may have a corresponding campus program, like our online MSCS program at Georgia Tech, which grants the same degree as the face-to-face program, or they may be exclusively online, like the University of Illinois iMBA program, which led to the discontinuation of its residential MBA program.⁶ We can often scale down even further, finding specific online classes available to students otherwise enrolled in person, as noted by the PHEAASG regulations; many universities offer online classes to high school dual-enrollment students as well, such as David’s own online CS1301 class, Introduction to Computing, at Georgia Tech. But regardless of the level, the distinction exists.

Why the distinction? Implicit in this dichotomy is the need for comparison. How do students’ outcomes differ in online and traditional educational environments? How do the costs differ? How does access differ? How can online education be used to reach communities that lack high-quality traditional education? What are the risks of online education to hidden social functions played by traditional education? These questions are asked over and over, reinforcing the either-or dichotomy between the two areas. More problematic, the implication of the different treatment of online classes by the law, financial aid, and public opinion is that online classrooms are in some way inferior. Rarely do we see incentives for more online classes, but we frequently see regulations limiting the number of online classes that students may take. Universities often attach the word online to degrees given via online delivery mechanisms, insinuating that the degree itself must be differentiated from in-person alternatives.⁷

These broad comparisons are unfair. Neither online nor traditional education is a monolithic entity consistent across all locales, schools, and subjects. To illustrate the range of possible differences, we take the example of an online undergraduate class we launched in January 2017 at Georgia Tech. In developing this class, we paid close attention to the research showing that learning outcomes often tend to lag in online classes compared to traditional classes. We wanted to ensure that the online class—CS1301: Introduction to Computing—could promise comparable learning gains to the traditional version of the same curriculum before rolling it out to a larger audience. The class we produced ended up turning out students who learned as much as or more than students in a traditional version of the class.⁸ Other experiments at MIT and Carnegie Mellon have found similar results.⁹ The class has been offered every semester (fifteen terms total) since, totaling over three thousand course completers for credit, and has also been launched as a MOOC; over ten thousand students have completed a MOOC version of the course. This scale was possible only because of the favorable learning gains we observed in our experiments over the first couple of years of delivering the course: without evidence of the learning outcomes, we would have been reluctant to expand the course so heavily.

These results run counter to an influential thread of research about online education, where the finding has been that outcomes suffer in online environments compared to traditional environments. In response to this result, some have argued that students in selective and prestigious research institutions like MIT, Georgia Tech, and Carnegie Mellon are themselves better prepared to succeed in online classes; they possess the discipline and self-regulation skills necessary to monitor their own progress with limited external structures ensuring their continued engagement.¹⁰ Much of the research finding poorer outcomes in online classes comes from community colleges and MOOC providers, and so some argue that the achievement difference is due to differences in the students. Online classes, then, could contribute to a widening of the achievement gap as they allow already well-educated students to move forward even faster based on their ability to succeed in more flexibly-available online courses.

Others—ourselves included—pose a different explanation. These large research institutions have and are devoting significant resources to developing online initiatives. When David developed our online Introduction to Computing class, we spent a full year writing the textbook, filming the lectures, and developing the initial assessments. We had a team of nearly a dozen people supporting David, including video producers, textbook copyeditors, project managers, technologists, and teaching assistants; in many ways, we had far more support than even traditional face-to-face classes have, before we even consider David’s own prior experience teaching online. Our online master’s-level courses are similarly developed by teams of professors, teaching assistants, instructional technologists, and project managers using world-class facilities. It is perhaps unsurprising that such a large investment of resources creates an educational experience leading to superior learning outcomes.

Although our offerings are all asynchronous, this increased investment can be seen in synchronous environments as well. In 2014, Harvard University launched HBX, a virtual synchronous classroom that had many features—such as virtual hand-raising and a lecture hall–like visualization—that have become major parts of the use of tools like Zoom and Microsoft Teams in the post-COVID-19 shift to remote learning.¹¹ Minerva Schools at KGI, a joint project between Keck Graduate Institute and the Minerva Project, uses a custom virtual classroom interface called Forum with features for efficient breakout groups, live class polls and discussions, and dynamic collaboration.¹² The same investment we made in high-quality recorded course material and artificially intelligent autograders can instead be made into platforms that support efficient synchronous interactions. Most schools, however, do not typically invest so heavily into individual courses; it is not uncommon for teachers at community colleges to learn only weeks before a semester starts that their class will be online. How can the learning outcomes of these wildly different types of course offerings be lumped together under the catchall category of online?

So, comparing online and traditional education is already an erroneous endeavor: there are so many different ways to conduct both experiences. We are not concerned here with which is considered better or worse; our concern is with the extent to which we consider them mutually exclusive opposites in the first place. Our concern is that the question posed to teachers is often first, Do you want to teach online or in person? rather than deeper questions about who, what, and how to teach.

The Distributed Classroom

Adopting the dichotomy between online and traditional education introduces numerous problems, but here we are most interested in the extent to which adopting that paradigm limits the design decisions that we make. In electing to teach in person, we take for granted several of the constraints that the decision carries: our class size will be limited, our presentations will be live, and the physical classroom serves as the home of all classroom instruction and administration. In deciding to teach online, we may instead assume compromises that need not necessarily be made, such as that instruction will be more one-directional and asynchronous, that interaction among students will be more limited, and that assessments must be more summative and rote. Most important, in teaching online, we risk losing many of the functional and pedagogical roles played by the physical classroom, such as fostering peripheral community, supporting interaction among peers, and facilitating rapid feedback between students and teachers. In effect, we risk teaching without a classroom.

We propose in this book that instead we ought to dispense with the dichotomy between online and traditional instruction. Universities, programs, and classes need not be labeled as specifically online or traditional. The alternative is that we no longer need to view classes as located in specific times and locations at all.

This is the notion of the distributed classroom, the premise of this book. In this context, distributed means that the class is not restricted in time or space, but rather can be distributed across multiple locations and multiple times. It breaks the assumption that a traditional class must meet together in a room—as well as the assumption that an online class should exist entirely without a classroom.

By that definition of distributed alone, distributed education options are already abundant: MOOCs, HyFlex courses, asynchronous online degree programs, informal learning communities, and more already represent distributed learning. However, the second part of the term—distributed classroom—is not merely a throwaway noun to give the adjective something to modify. In the distributed classroom, classroom is as important as distributed. Classroom is not merely a stand-in term for any learning environment; rather, it refers to the specific roles that a classroom plays in traditional education, especially those that rely on live, and ideally in-person, communication, potentially making use of affordances of the shared physical classroom. While some underlying functions of the physical classroom may be offloaded to other interfaces—a course forum for questions and discussion, tools like Peer Feedback¹³ and Peerceptiv¹⁴ for peer review—there remain others that do not translate evenly into an asynchronous remote environment, like group work on a shared physical artifact, creation of a three-dimensional space conducive to learning that course’s content, or the sense of connectedness that comes from simple peripheral awareness of classmates’ presence and participation in the course material.¹⁵ Even among those functions that may be offloaded onto interfaces to facilitate remote asynchronous interaction (such as course discussion forums), a degree of connectedness and empathy may be lost when communication is anchored to static two-dimensional images rather than a live video display; Albert Mehrabian famously claimed that 93 percent of all communication is nonverbal, which may be lost in text-only asynchronous online media.¹⁶

Taken this way, we hope it begins to become apparent how a distributed classroom expands on other innovative learning experiences. A distributed classroom is a class designed such that students can participate in as much of the full learning experience as possible within their individual constraints, especially constraints based on place and time. If they can attend in-person in the live classroom, they may do so; if they can commit to live attendance with a cohort of classmates in another location, they may do that instead; and if they can commit only to remote, asynchronous learning, they may participate that way. No matter their individual constraints, they give up only those parts of the learning experience that are fundamentally tied to that constraint: inability to attend in-person on-campus need not force students into the same remote asynchronous experience as students with additional constraints. Where other initiatives, such as MOOCs, focus on expanding access to a particular