The Death of Textbooks?
Artificially intelligent software is reshaping traditional teaching materials—but it's unclear what the new technology will take away from the learning experience.
At a recent sit-down with executives representing one of the biggest players in the textbook industry, my colleague and I felt surprisingly out of touch.
The executives spent most of the meeting touting the evolving market, namely how their newfound allegiance to digital learning materials—rather than old-school physical textbooks—would place them at the forefront of the new wave of education technology. Rhetoric describing the company’s unmatched innovation pervaded the hour-long meeting; they raved about the company’s across-the-board shift to digital, how its new state-of-the-art materials comprise a "single roadmap" that is expected to make its generic, stodgy textbooks obsolete. They largely dismissed us as we—online journalists and Millennials in our mid-20s—reminisced about physical books that can be held, highlighted, and leafed through. And it quickly became evident that these men expected us to marvel at the company’s developments because, as soon as they noticed our eyes weren’t lighting up, they balked: "I don’t think you understand how groundbreaking this is," one of them said.
These executives certainly seem to have popular opinion on their side. Textbooks, which have long accounted for various subjects’ bulk of in-class learning materials, have garnered much vitriol in recent years. For some, the discontent starts as early as elementary school; heavy books can result in chronic back pain for children. But by the time those students are in college, textbooks are much more than a mere nuisance. What was once heavy burden on the back becomes an even heavier strain on the wallet. According to a recent College Board report, university students typically spend as much as $1,200 a year total on textbooks.
Nostalgia aside, it may come as a relief to many, then, that textbooks are becoming anachronistic. Digital in-class learning materials, like software that adapts to the ways in which individual students acquire information, and other forms of virtual education content are becoming more effective and intelligent. College-affordability advocates and others hope this growth could result in the normalization of less costly or even free materials down the road.
But as the executives failed to acknowledge in that meeting, the shortcomings of this trend—and its prospective impact on how humans learn—are worth keeping in mind. Many digital learning materials completely overhaul how classes, from pre-k to grad school, are conducted; how students are tested on knowledge; and how teachers fit into the picture.
While the rising cost of college has gotten a lot of attention in the media, much of that coverage neglects to detail just how much textbook prices have skyrocketed: 82 percent in the last decade, according to a 2013 study by the Government Accountability Office.
Many critics blame this trend on what they describe as a rigged textbook economy that only profits the publishing industry. Which in many ways makes sense: The textbook world operates differently from most consumer markets because the consumers—the students and teachers—rarely get to choose what they’re buying. They typically just receive a syllabus and purchase as told. As a result, the price point is largely exempt from popular demand, giving publishers supreme control.
One of the college textbook industry’s most egregious practices, it seems, is that it constantly republishes new editions, rendering even last season’s versions outdated. One professor told me this tendency is necessary because of perennial advances in the given field. However, although that may be true on occasion, even people who were formally employed in the industry have admitted that these "new" editions typically include little more than glorified cosmetic updates—a new picture, for example, or a heading adjustment.
Advocates have called for more open-source textbooks as alternative—resources available online at little to no cost. Ethan Senack, who advocates on federal higher-education issues for U.S. Public Interest Research Group, is particularly vocal in his support for this alternative, recently authoring a report, "Open Textbooks: The Billion Dollar Solution," that details how much money students can save using open-source materials. "We see price tags as high as $280 to $300 [each], and it’s just too much," he told me. "We have the technology now where we can deliver learning materials in an affordable way."
Senack is, of course, right to say that technology has changed how the textbook market operates. But that evolution, at least initially, hasn’t happened in the way cash-strapped students may have hoped. Many college courses still use physical textbooks and mandate digital content as an add-on—complete with end-of-semester expiration dates, which undermine resale value, for example. These new strategies are exactly what irk Senack: "They [the publishers] have continued the same practices they’ve used to limit the spread of knowledge … It’s just a continuation of the bad practices."
These new materials do come with pedagogical advances. With traditional textbooks, each student is presented with the material in the same way regardless of his or her proficiency in that particular subject. Now, artificial intelligence means that the material itself can actually study a student’s learning habits and adjust in real time.
To find out what’s on the horizon for K-12 textbooks, I recently spoke with representatives (including the ones with whom I met at the sit-down) of two of the world’s largest textbook publishing: McGraw Hill and Houghton Mifflin Harcourt. Both companies, as well as the industry’s other titans, are heavily focused on bolstering their digital arms in an apparent effort to dominate the booming virtual-content industry. Even Pearson, another behemoth in the textbook industry, now brands itself as a "digital learning and services company."
Houghton gave me a live demo of its language-arts program known as "Collections," an English course. The software essentially streamlines the process of analyzing and reviewing a text—think a teacher reading an excerpt, asking questions, and assigning an essay— on a simple interface. This interface allows kids to highlight material, take notes, and ask their classmates questions virtually. It even includes a "raise hand" button that a student can click to alert a teacher if he or she has a question. The program, moreover, pre-selects what it considers difficult or particularly wordy paragraphs, flagging them with a "Close Reading" tag; clicking this button opens up an brief explanatory video. Students can follow along with the audio version as they read the text—perhaps negating the need for the teacher to call on a student to recite the text out loud.
I also tried McGraw Hill’s math software: ALEKS, which stands for Assessment and Learning in Knowledge Spaces, and is a web-based tool that assess students in mathematics, accounting, statistics, and chemistry. (The name alone—which, phonetically, sounds like a human—got me worrying about the prospect of a robot takeover, but I digress.) ALEKS launched in 1999 and operated for more than decade on its own before it McGraw Hill purchased it in 2013 and is predicated on "completely individualized learning"; the program adapts based on an individual’s knowledge and skill set. Behind the scenes, the software builds a database detailing the proficiency of each student, information that is then used to formulate questions tailored to kids based on what they find most challenging. Essentially, the program—which is based on 20 years of research by cognitive scientists, mathematicians and engineers—can instantly assess the individual abilities of an entire class of students at a rate that would be impossible for most teachers.
Many educators, including Idaho’s 2015 teacher of the year, Kim Zeydel, swear by ALEKS. "I work with at-risk kids; most of the kids are at a fourth- to sixth-grade math level, and we’re trying to teach them at the high school level," she told me. "The kids have holes in so many different areas. You have 20 kids, you have 20 different needs."
For Zeydel, a veteran educator who started using the program four years ago, ALEKS has been able to fill in the gaps where she falls short. "ALEKS can diagnose where the kids are. If they don’t have the foundation on the other skills they aren’t allowed to move forward" she said. "They build the foundation really nicely, which I as a teacher will not be able to do."
Zeydel says she uses ALEKS sparingly: typically two days weekly and as combined with her traditional teaching methods. This approach is generally referred to in education circles as "blended-learning," which is described as a middle ground in which teachers and digital learning materials can coexist. Still, despite the seemingly ideal compromise of blended learning, the dynamism of new strategies begs the question: How long before artificial intelligence outweighs what a human teacher is able to do?
Experts insist that won’t happen. John Rice, who oversees blending-learning initiatives for Washington, D.C.’s public schools, is adamant in denying this is the case: "It’s called blending for a reason," he said. "It’s the best of what teachers can do that will never be replaced by technology. And the best of what technology can do that is just way more efficient and advanced that what a teacher can."
But instruction aside, other challenges remain—including, for example, how the software lands at the school. A recent report by Digital Promise, a nonprofit that aims to "help bridge the digital divide in education," found that huge discrepancies exist in who decides what materials a school ends up using: The people with the most decision-making power, it found, tend to be district technology directors (93 percent)—not teachers (50 percent) or even principals (61 percent). But a lack of information is in part to blame for the disconnect. In the study an unnamed source in the industry said that it is "hard to identify which schools/districts are a good fit for us," while another noted, "It’s not as if the districts are really broadcasting what they are looking for, and sometimes they don’t know what they are looking for until they see it."
Ultimately, these digitized materials are somewhat of a paradox. They are standardized at the top—the programs are aligned with the Common Core and rely on big data—but personalized underneath, customized around each student according to what the software gleans from assessments.
This shift also means that kids are spending more time than ever looking at screens, which could be physically and cognitively detrimental in the long run. The American Academy of Pediatrics, at least for now, recommends that kids spend no longer than two hours a day looking at digital devices. The shift is also taking a toll on the frequency that children engage in handwritten work, which reports have shown is far more beneficial than taking notes on a laptop. And these changes could be disregarding how kids want to learn. Recent studies suggest that "digital natives" still prefer reading in print. One University of Washington pilot study of digital textbooks found that a quarter of students surveyed bought the print versions of e-textbooks that they were given for free, according to a recent Washington Post report.
Bill Buxton, the founder of the open-source publisher Textbook Equity, is skeptical of technology as a substitute for traditional learning materials. "I haven’t seen really strong evidence that people are doing a lot better with the online stuff than textbooks," he said. "Where’s the evidence? ... It’s coming from the biased companies; they want to make sure people buy it."
Others are wary of technology’s impact on learning, including Nancie Atwell, who founded the Maine-based Center for Teaching and Learning and is a top-ten finalist for a forthcoming global-teacher contest that will award the winner $1 Million. "I think they are a disaster for teachers. We don’t know anything about the value of eBook. They’ve been foisted on teachers because they are the latest technological advance so they must be good," she recently explained to my colleague. "The problem with eBooks is that the kids remember much less than what they read on the screen compared to the book."
And the current climate is nothing compared to the technology that’s likely being developed behind the scenes in the corporate textbook sector. Indeed, students are learning in a constantly evolving educational space. Rice, of D.C. public schools, pointed, for example, to "educational playlists"—think Spotify for math. Instead of one course from one provider, a student’s math program could be split up into 10 topics that are each designed by a different company. The student can then mix and match these units as long as his or her final "playlist" corresponds with the curriculum.
"I sat down with some students, and I thought this would be boring [to them], but I asked the kids and the kids said ‘no I really like that. I like working by myself and having these adaptive lessons and getting some cool feedback,’" Rice said. "They were really happy with it, which the textbook can never do for you."
But that’s precisely what worries some educators the most: What was once a teaching model is transforming into a learning model. As Atwell put it: "Technology is a means; it’s not an end. And it’s become an end within this country."