Like so many other universities around the world, the University of Michigan relied on its existing technological resources to design a distanced education for the COVID-19 era. The pandemic certainly came as a surprise, though ultimately a widespread familiarity with the basic tools of email, course websites and video conferencing helped students and teachers adapt to a new normal.
Online education in general, however, is not a new phenomenon, nor is it likely to disappear with the return of in-person instruction.
In fact, the prevailing narrative is that the pandemic has catalyzed an inevitable shift to online learning. The World Economic Forum called the pandemic a “paradigm shift” for colleges and universities that has “accelerated the transformation of higher education.” Similarly, Byeongwoo Kang of the Institute of Innovation Research, Hitotsubashi University, Tokyo, wrote in an article published by the National Center for Biotechnology Innovation that, although the expansion of distance learning was advancing steadily in Japan, “COVID-19 has accelerated digital transformation in the education sector.”
Lots of people are saying that changes to higher education are accelerating, but what is higher education accelerating towards?
Online education is the present reality for most residential students, but for many others, it has represented and continues to represent a viable avenue for expanding access to higher education. While thinking ahead about the future of education in general, I began to wonder: How will the University contribute to the expansion of opportunities to learn and seek a degree in the post-pandemic era?
***
Behind the hype around online education are the researchers, designers and engineers who are making it happen. Here at the University, the Center for Academic Innovation is one organization dedicated to the design of educational technologies and to supporting faculty who want to integrate digital components into their classrooms.
Recently, Sarah Dysart, the center’s director of hybrid and online programs, and Lauren Atkins Budde, director of open learning initiatives, filled me in on the workings of the center and how it has been pursuing its mission of designing the education of the future.
During our Zoom conversation, Dysart explained how the initiative to found a dedicated office for academic innovation formed in response to popular demand for online learning opportunities. In 2012, the online learning platform Coursera began to offer its first batch of massive online open courses, available to anyone with a browser and an internet connection. The University was one of Coursera’s first partner institutions and helped co-create some of its first MOOCs, including “Python for Everybody,” which is consistently one of the most popular courses on the platform. This collaboration served as the impetus for the creation of the Office of Digital Education and Innovation in 2014, the institutional ancestor of what is now CAI.
Today, the center has the dual role of supporting the University’s non-credit and credit-bearing online programs. Some of the not-for-credit learning programs include MOOCs, specializations, Teach-outs, podcasts and other forms of widely available educational content. “We have, I think, almost 15 million enrollments from learners of over 200 different countries that engage with our learning experiences,”’ Budde said.
Dysart, on the other hand, manages the programs that are eligible for U-M credits or have a pathway to earning them. This includes several online master’s degrees programs, as well as programs that learners can begin as open experiences and then convert to credit later on. “Anytime somebody is converting an open learning experience into a for-credit initiative,” Dysart explained. “Like taking content from a MOOC course and then applying it to a residential course, that’s often a time when I’d be pulled in.”
Along with an array of software developed for U-M students, the center has been instrumental in helping different schools at the University develop online courses and master’s programs that can reach a broader audience of worldwide learners.
The School of Information, for example, offers a “Master of Applied Data Science” online on Coursera. The online curriculum is designed to be flexible and accommodate learners who have to balance school with other major responsibilities. While the residential Master of Science in Information program requires students enroll in a minimum of nine credits to be a full-time student, the courses in the MADS program are one-credit each, and students take an average of one asynchronous credit per month. The workload, therefore, is more spread out over time, even though students can still expect to finish the degree in an average of two years.
One of the major challenges of online programs is fostering a sense of community. For MOOCs, online discussion forums provide the bulk of interactions between learners and U-M faculty.
“You’ve probably experienced for large classes on campus how hard it is for a professor to really connect with, say, the 500 students they might have in their CHEM 125 section,” Budde said. “Imagine if that instructor had 10,000 students every semester: That’s the scale that you’re looking at for those sorts of online courses.”
To support online learners, an internal team at CAI helps out with the Coursera courses. Additionally, though, some faculty members have made special efforts to connect with their students; Budde even told me about a few professors who in pre-COVID-19 times held pop-up office hours in local coffee shops while they were traveling around the world.
While online learning at-first-glance seems to be all about the technology, there is clearly a human component that is crucial to the success of any online venture, whether it’s a course or a full degree.
***
Designing the future, then, is not about designing the perfect tools, but about ensuring that lots of different aspects of the learning experience are taken into account. So, along with engineers, it is the job of educational researchers to think about how these digital tools can be applied to academic contexts.
To learn more about the work of an academic researcher, I talked with Juan D. Pinto, an alum of the center’s Learning Experience and Design Graduate Certificate program.
Pinto earned his master’s in education studies from the School of Education in 2020 and completed the certificate program as part of his degree. He is currently pursuing a doctorate in the Digital Environments for Learning, Teaching & Agency program at the University of Illinois Urbana-Champaign, and is especially interested in the applications of artificial intelligence to educational technologies.
As one might expect, Pinto is no pessimist about the possibility for technology to make a positive impact in the lives of learners. However, he admitted to me during our conversation that his field had lost its way when it comes to the successful integration of educational technology.
“I’ll be honest with you,” Pinto said during our phone call. “I think educational technology as a whole has failed. The more I learn about it, the more I study it, the more I feel like we are failing.”
Part of the reason for this failure, Pinto explained, has to do with the overhype and profit motive of the technology industry. However, he also made clear that his own field of educational research needs to make more concerted efforts to reach out to the communities they are trying to support with their work.
“It’s much more than just, ‘I study how people learn in a lab, and then I design a technology that I think will be beneficial,” Pinto clarified about the goal of his research. “There has to be so much more that goes into it and actually implementing it in context, so that it makes sense for a specific school, specific classrooms, and specific students to use. And we have to make sure that we’re not leaving people behind because right now we are, we definitely are.”
Pinto pointed out, for example, that the accelerated adoption of technologies during the pandemic has reinforced existing structural inequalities in access to education. While technological innovation is often framed in terms of broadening access, digital technology just won’t have this effect as long as there are significant deficits in reliable access to the Internet and computers.
Because of the ambiguous relationship between technologies and educational outcomes, Pinto advocates the fostering of a critical approach toward the design and implementation of new educational technologies.
One component of this critical approach is the examination of algorithms and their relationships with human subjects. Most technology companies, however, do not disclose information related to proprietary algorithms, constituting a major obstacle to an open and equitable digital future.
“The problem with proprietary software or proprietary algorithms is that essentially they can do anything,” Pinto said on the issue of algorithmic transparency. “And because the technology has developed so quickly over recent years, the government, legislators, and policymakers just can’t keep up.”
The algorithms used in AI, for example — are susceptible to reproducing biases due to their design or the data on which the algorithms train. For example, a report from The Brookings Institution showed how the use of AI in schools threatens to reproduce racial discrimination associated with standardized testing. Thus, if schools contract with companies that protect their algorithms as trade secrets, then it might be costly to hold these companies accountable in the event that their products cause harm.
The future of online education, then, is ideally rooted in accountability, transparency and active engagement with local communities. However, in the absence of close regulatory scrutiny, it is up to public and private institutions to exercise proper care when considering the introduction of a new dimension of their online educational infrastructure.
***
So, as the University continues to hurtle into the future, its administrators, researchers and faculty must also continue to think critically about the relationship between technological change and equal access to higher education. The future is not set in stone, and what we do now, in large part, forms the kind of people and the kind of university we will be.
The stakes are high, too, because the university is still dealing with an identity crisis that technology alone cannot solve, but in which technology will indeed play an outsized role. This identity crisis stems from the tension between the University’s contemporary self-presentation as an elite university on the international stage, and roots as a public university, for which a major goal is to serve the people of the state that bears its name.
The 17th Edition of the Michigan Almanac, published in March 2021, reports that, based on data from Fall 2018, 53% of incoming first-years were from in-state, 43% were from out-of-state and 4% were international students. The University enrolls a lower percentage of in-state students than most other public schools, but the University provides an explanation: “In large part this is because U-M’s primary competitors for these students are selective private universities.”
To be sure, the University’s international reputation helps attract young professionals and growing businesses to the state. The expansion of U-M-sponsored online education can therefore continue to play a role in brightening the University’s brand in an increasingly competitive landscape.
As an out-of-state student, though, the above statistic and its subsequent justification have never sat well with me. In my view, their combined implication does not reflect well on the University as an institution of public service, even if at-first-blush it seems like a badge of honor to declare yourself a “competitor” of Harvard University, Stanford University, the Massachusetts Institute of Technology, the University of Chicago and the other universities that tend to follow.
At least from a fiscal perspective, the public character of the University has decreased dramatically since the 1960s: in 1960, state funding accounted for 78% of the U-M Ann Arbor General Fund budget, whereas in 2020 state funding accounted for only 14% of the General Fund. (In FY 2020, the General Fund accounted for around 24% of the total budget). All in all, state funding, therefore, accounts for around 3-4% of the total budget.
The trend of declining state funding has been associated with increasing tuition, as well as efforts to find alternative revenue streams such as online learning. Nevertheless, even if the University does not have strong financial ties to the state, it does have strong cultural and historical ones. These are written into the word-order of the University’s mission “to serve the people of Michigan and the world through preeminence in creating, communicating, preserving and applying knowledge, art, and academic values, and in developing leaders and citizens who will challenge the present and enrich the future.”
I think a good example of this public-spiritedness is encoded in “Python for Everybody,” the Coursera course taught by Information professor Dr. Charles Severance.
One of Severance’s goals in developing the course has been to make sure the course materials are accessible to the widest range of learners possible. When he says, “Python for Everybody,” he means it. His courses are accessible to people with vision problems, for example, because he takes care to integrate storytelling as a part of his lectures, turning visual aids such as graphs, charts and diagrams into rich verbal descriptions. His most basic Python course, too, does not have a calculus or trigonometry prerequisite, a common requirement to pursue a computer science education that he says filters out a lot of potential students who would have otherwise learned a valuable skill.
“If you’ve taken calculus and you’ve taken programming, it will be very clear to you, very rapidly, that programming has nothing to do with calculus,” Severance explained. The computer science programs at the University have these requirements, he argued, as a harsh filtering mechanism.
Students in “Python for Everybody,” by contrast, can stop taking a course, leave it alone for a while, and come back with their previous progress saved. The rapid pace of semester-based education in college teaches students how to learn quickly and efficiently, though, in Severance’s experience, the best learning takes place in a nonlinear fashion across long stretches of time. In his view, this is a more forgiving method of learning and actually ends up making knowledge stick a little better than if one were learning the same material while cramming for an exam or burning through a coding project within the span of a few days or less.
Finally, nearly all of the course materials for “Python for Everybody” are available under a Creative Commons license, free to re-use with attribution for educational purposes. According to Severance, this has resulted in people from all across the world drawing on his lectures, worksheets and auto graders available from the course website.
In terms of improving the University’s brand, Severance sees his course as a major contributor to a positive image of the university around the world. On the other hand, he did express some disappointment in the university’s efforts to engage with high school students and local communities within the peninsulas. When I asked him about the relationship between the University’s online programs and its engagement with state residents, he grew sullen.
“I have historically wished for better,” Severance said with a sigh. “I don’t think we have much good to report in that area. And it makes me sad.”
Part of this remorse centered around the university’s lack of recruitment of talented students from two-year community colleges around the state, and lack of interaction with underserved high schools and middle schools. While his outlook on the University’s transition toward the grandeur of an elite international institution was fairly bleak, Severance made clear that he was determined to bring his vision of promoting computer programming education to fruition.
“I have built a high school curriculum,” Severance said, as he started laying out his master plan. “I created it five years ago. And slowly but surely I have been building all the classes and not just Python, but an entire high school technology curriculum. Do you think I’ll ever get it in the high schools in the state of Michigan? Probably never.”
Severance, however, has not given up on his goal of exporting computer programming skills to the largest possible audience. For one, Severance had been approached by an ex-convict to design a curriculum for computer programming to be taught in Michigan prisons. Second, the president of Adrian College, a small liberal-arts college in southeastern Michigan, reached out to him to get some help developing a computer science program for liberal arts colleges. Their idea was to design a liberal arts degree where technological skills are treated as a key component. To Severance, this represented a possible point of departure for his bigger dreams.
“If I can get it into liberal arts colleges, and I can teach them, maybe I can get it into community colleges,” he said. “And if I can get it into the community colleges, maybe I can get it into high schools… I have got a long-term plan to help the state of Michigan. And for every opportunity that comes across, I am ready.”
***
Serving the world is a good goal for the University to pursue, and the internet has made it easier than ever to reach billions of people with quality educational programs. This broad, universal impact suits the University of Michigan as a leading international research university.
But serving the world and supporting the state of Michigan need not be mutually exclusive. Each additional augmentation of the university’s reputation, which will likely accelerate due to its early mastery of delivering the online degree from a top-class university, need not result in proportionate oblivion towards the University’s neighbors.
In my view, it is not a good idea to expect entrepreneurs, engineers and designers to think on their own through the greatest possible number of variations as to how their products will play out in real-world situations. There have to be checks on the process of technological innovation and attention to the limits of online opportunities that do not extend in a meaningful manner to marginalized communities. Along with developing new software, how can an organization such as CAI, for example, support the expansion of high-speed internet access in rural areas, including, but not limited to, underserved areas in the state of Michigan? How can MOOCs and specializations act as points of entry into high-demand careers in science and technology while also providing avenues for literary and artistic expression? How can innovation research at the university incorporate the needs of people with disabilities and chronic illnesses into their design considerations, not as an afterthought but as a structural part of the innovation process?
If this really is the accelerated digital transformation, and great changes really do lie in wait, then let us take the present moment to reflect with care about what kind of university we want to be when the future finally arrives.
Statement Correspondent Alexander Satola can be reached at apsatola@umich.edu.