Justin Reich has been studying the effects of technology on teaching and learning for many years, and is an advocate for continually tinkering with technology to improve learning opportunities. After an early interest in wilderness education, teaching mountain rescue and first aid, he found himself exploring the wilderness of MOOCs, the massive open online courses that were the subject of much hype in 2012.
Though MOOCs were widely predicted to disrupt and democratize post-secondary education, they proved to do neither. Traditional forms of higher education, while costly, stubbornly continue to provide the most common path toward degrees, and those who participate in MOOCs most successfully tend to be already privileged, well-educated, and comfortable with technology.
Justin’s work as an Assistant Professor of Comparative Media Studies/Writing and Executive Director of the Teaching Systems Lab at MIT, is devoted to helping educators and those who train teachers to connect their practices to our increasingly networked world. In addition to publishing numerous peer-reviewed educational research articles and coauthoring books on using technology in the classroom, Justin has written essays for a wider audience, appearing in publications such as The New Yorker, The Atlantic, and the Washington Post as well as The Chronicle of Higher Education, Inside Higher Ed, and Education Week.
Justin has also used a variety of tech platforms to share his knowledge. After concluding a six-year term as a blogger for Education Week, he launched the TeachLab postcast. In the fall of 2020 he used Zoom to convene a book club for readers of his new and most timely book, Failure to Disrupt: Why Technology Alone Can’t Transform Education.
We caught up with Justin in November 2020, just as the affordances and limits of educational technology were coming into focus during a global pandemic. (Interview posted: December 1, 2020).
PIL: As you put the finishing touches on Failure to Disrupt, the COVID-19 pandemic drove schools to rely on technology to an extent that had never been attempted. What are two or three of the most relevant findings for this moment from your in-depth analysis of research into educational technology? Is there anything that has surprised you?
Justin: In the history of education technology, two findings come up over and over again. First, when teachers get access to new technologies, they use them to extend existing practices. Teachers almost always use technology to digitize what they were already doing in classrooms; it’s much more rare for new technologies to spark truly novel pedagogies and routines. The second key finding is that when learning technologies are used in powerful, interesting ways, they tend to disproportionately benefit the affluent. Even when technologies are free, they rarely “democratize education” or close gaps between more and less affluent learners; instead, people with social, financial, and technical capital are best positioned to take advantage of new innovations.
Very early on in the pandemic, I predicted that we would see new “yawning gaps” between more and less affluent learners in terms of their experiences in remote learning. That has proven to be tragically true.
For some reason, I was more optimistic that some schools and colleges might choose to be somewhat more ambitious in using learning technologies to rethink remote learning. For instance, I predicted that there would be some more-than-trivial adoptions of open courseware and massive open online courses in higher education. Several universities pour substantial resources into creating online courses in typical classes like microeconomics and introduction to physics. I figured that many university faculty would tell their students “Go take those free classes online, we’ll meet once a week to discuss, and I’ll write and grade your mid-term and final. These online providers can make better courses than what I can cobble together from my home office with two kids crawling under my feet.”
As far as I can tell, that mostly has not happened. The two dominant technologies of the pandemic have been two of our oldest learning technologies. The first is the learning management system—things like Canvas and Google Classroom—that were theorized in the 60s and 70s, commercialized in the 90s, and made open source in the 00s. The second dominant technology, when it was introduced in the 1930s, was called video telephony; now we call it video conferencing. It lets individuals broadcast messages to groups, and allows for discussion and peer interaction in limited ways. Together, these technologies allow for an approximation of teacher-led, whole-class direct instruction to happen synchronously. Then, in secondary schools and colleges, teachers mostly walked away from their lecterns, sat down in front of their home office cameras, and kept teaching largely as they had been doing before the pandemic. A brief summary of primary school remote learning might be “unless parents are right there with their kids, nothing works.”
In our lifetimes, we will never see a more powerful demonstration of the conservatism of educational systems. In March, I predicted that most things would stay as consistent as people could make them, but I think even I underestimated just how committed faculty, students, and other stakeholders would be to trying to maintain the traditional structures, routines, and pedagogies of schools. I don’t necessarily mean this as a criticism, that’s a more subjective judgment. But just descriptively, most educational systems did everything they could to select and use technologies that allowed them to reproduce classroom routines.
PIL: In your book, you point out “schools are complex places” and “the material conditions of learning environments” have a strong influence on how students learn. Since last spring, we’ve had a crash course on how much material conditions, such as having a quiet place to study and internet connectivity, affect learning in a highly unequal society. Recently you led a project to interview teachers about their experiences during the pandemic. What are some non-technological factors that could lead to a healthier learning environment for both students and teachers using technology—now, during an emergency, and down the road?
Justin: We need to raise children in well-supported families and households, and this goes far beyond the role of schools. In the wealthiest nation in the history of humanity, we should be able to make sure that every young person has a roof over their head, three square meals a day, guaranteed health care (include mental and dental health), and access to high quality schools, libraries, and cultural institutions. There is no education technology that can teach a hungry child. And we can’t ask schools to solve all of these problems alone; there have to be a variety of social institutions in municipalities and states that address these issues.
One related issue is that in the century ahead, which will be defined by our shared climate emergency, schools will be regularly interrupted like they are now—by new pandemics, by extreme weather events, and by political conflict from migration, famine, and related issues. So we need to develop systems of schooling that are more resilient to interruption, and one way we can do that is through universal broadband access.
PIL: In Failure to Disrupt, you delve into factors that complicate the promises made about technology’s disruptive promise, including the persistence of the familiar and the “EdTech Matthew effect.” How do you define the EdTech Matthew effect, and how does this concept help explain who benefits most from the technological delivery of education? Which of the three approaches to learning at scale that you explore in your book—instructor-guided designs (such as most MOOCs), algorithm-guided personalized learning, and peer-guided networked learning communities—do you think holds the most promise for reducing inequality in education? What is an example or two of how educators “tinkered” with technology in ways that are beneficial?
Justin: The EdTech Matthew Effect is a variation of a phenomenon found widely in sociology, which is that many social forces lead the rich to become richer, and for the poor to have a tenuous hold on their wealth. EdTech evangelists often claim that new technologies will democratize education and close these kinds of gaps, but in reality, they usually do the opposite. Affluent people have more resources to make technology, even free technology, work well for them. I argue that this kind of as-yet intractable dilemma is a problem for all kinds of education technologies. It is a challenge that spans across the three genres of learning and scale technologies that I define.
The evidence for the EdTech Matthew Effect is substantial and in my view very persuasive.* Evidence about possible solutions to this dilemma is much more sparse and less clear. I point to a handful of promising potential directions for a more equitable EdTech. We can reduce the social distance between EdTech developers and the communities who use these tools in schools. For example, we can imagine an EdTech development ecosystem where tools aren’t just made by Ivy League grads in Silicon Valley, but by much more diverse teams. We can study EdTech implementation and outcomes by subgroups, so we make a habit of evaluating how new technologies affect different kinds of opportunity and achievement gaps. We can think beyond students to broader communities: if it takes a village to raise a child, then our education technology should empower everyone in the village. EdTech tools shouldn’t just benefit students directly, but they should help parents, aunts, uncles, and other care providers build capacity, too. I’m not certain that these are exactly the right pathways to a more equitable future for EdTech, but these are the kinds of issues that need to be much more central to the study and development of education technology.
PIL: Two chapters of your book focus on data, both gathering it for routine assessment (which often emphasizes the skills that are most likely to fall to automation rather than the more complex skills needed for the future of work), and the “toxic power” of holding massive amounts of data and using it to conduct experiments on students. Yet data can inform evidence-based practice. What is the role of students in these practices? Are there ways students could be involved in evaluating the technologies they are asked to use? To what extent can they give informed consent for valuable educational research?
Justin: Every teacher varies their instruction to improve student learners. Teachers try this book instead of that book; this technique instead of that technique; this problem instead of that problem. Typically they evaluate the efficacy of these alternatives through a combination of evidence and intuition, and usually these informal methods don’t do a great job of assessing with confidence which alternative is really better. Social scientists have a specialized tool kit for comparing instructional approaches in ways that give us greater confidence in our evaluation of what worked better for whom under what conditions. When those studies give us insights into how to design technologies better for learning, we can instantiate new research insights directly into platforms that can incrementally, steadily improve over time.
While I think this kind of systematic experimentation is great, it turns out that many parents are not so keen on the phrase “experiment with children.” There are also real problems with generating research infrastructures that collect vast amounts of data about children, and run the risk of indoctrinating them into a culture of normalized surveillance.
Kids are busy flirting, texting, playing video games, and punching each other, so I’m not sure that I’d sign them up to be the arbiters of their involvement in these debates. They should certainly be invited into public conversations about the appropriateness of research in education technology, but probably adults will need to do the bulk of this work.
For instance, asking kids to grant consent presupposes that they will actually take the time to become informed, and all evidence suggests that very few people read the fine print that comes with education technology or research. Moreover, and this gets into the weeds a little, but Congress has specifically designated educational research as an area where research requirements should be relaxed relative to other domains, like medicine. We already compel children to go to school; we already compel children to do activities and take assessment; it doesn’t make a lot of sense to cordon off as optional a set of activities that systematically evaluate how we can do those things better, when every day students are subjected to unsystematic variation in instruction.
These aren’t easy issues, but we need a way to use social science to continually improve instruction while simultaneously maintaining the trust of students, educators, and families by holding ourselves to high ethical standards and including all of these stakeholders in public discussions about research practices.
PIL: We were intrigued by your discussion of the dark side of peer learning online—the ways that social media platforms have enabled recruitment into dark and hate-filled parts of the open web. This kind of peer-guided learning is destructive and dangerous, but effective. Given your long-term involvement with collaborative learning tools and your current experience teaching an edX course on sorting truth from fiction co-taught with Sam Wineburg, what are some ways educators could help students become information literate in a world where “research it yourself” is a mantra for conspiracy theorists? How can learning opportunities be curiosity-driven when social media is designed for increasing engagement rather than thoughtful critical appraisal?
Justin: Sam Wineburg’s work on Civic Online Reasoning is definitely the place for teachers to start. Sam studied middle schoolers, Stanford freshman, and tenured and award-winning historians, and found that on average all three groups were not very effective at sorting truth from fiction online. Then, he studied another group: fact-checkers at some of the leading journalism outlets in the country. These fact-checkers passed Sam’s online search tests quickly and effectively, and most importantly they searched differently than everyone else.
Students and historians evaluated websites by “reading vertically,” they looked for clues and markers of authority from within a single web page. This perhaps isn’t surprising, it’s basically how we have taught a generation of people to evaluate quality on the web. Take a long checklist of questions, subject the text on the website to those questions, and come up with an assessment of quality.
Factcheckers, by contrast, read “laterally.” They stayed just long enough on a page to get the name of an author or organization, then they left and searched the rest of the web for information about those authors or sponsors. In one memorable example, a factchecker had seven new tabs up about an organization in less than a minute! This technique of lateral reading is teachable, is realistic to do in most search settings, and it works. It’s the most important first step we can take towards online media literacy, and we need to help everyone in America learn this skill as soon as possible.
*See p. 125, Failure to Disrupt, where Justin writes: “From MOOCs to adaptive tutors to Scratch, evidence suggests that an edtech Matthew effect is quite common: that new technologies disproportionately benefit learners with financial, social, and technical capital to take advantage of new innovations.”
Justin Reich’s teaching experience is varied, ranging from serving as a camp counselor, a search and rescue instructor with the Blue Ridge Mountain Rescue Group, a wilderness medicine instructor with Stonehearth Open Learning Opportunities, and an international expedition leader with World Challenge Expeditions. He also taught at the Shackleton school, an expedition-based school, and at the Noble and Greenough School, where he taught freshman world history, electives for seniors, and co-led the outdoor activities group.
Justin earned his doctorate from Harvard Graduate School of Education and is an alumnus of the Fellows program at the Berkman Klein Center for Internet and Society. Currently, he is an Assistant Professor in the Comparative Media Studies department at the Massachusetts Institute of Technology and an instructor in the Scheller Teacher Education Program. He also is the director of the MIT Teaching Systems Lab.
Smart Talks are informal conversations with leading thinkers about new media, information-seeking behavior, and the use of technology for teaching and learning in the digital age. The interviews are an occasional series produced by Project Information Literacy (PIL). PIL is an ongoing and national research study about how students find, use, and create information for academic courses and solving information problems in their everyday lives and as lifelong learners. Smart Talk interviews are open access and licensed by Creative Commons.
Suggested citation format: “Justin Reich: Tinkering Toward Networked Learning: What Tech Can and Can’t Do for Education” (email interview) by Barbara Fister, Project Information Literacy, Smart Talk Interview, no. 33 (1 December 2020). This interview is licensed under a Creative Commons Attribution-Non-commercial 3.0 Unported License.