For a year or two there, free online classes seemed like they just might be the future of higher education. Why, some influential computer scientists wondered, should there be thousands of colleges and universities around the country all teaching the same classes to small groups of students, when you could get one brilliant professor to teach the material to the whole world at once via the Internet? In a March 2012 Wired cover story about the phenomenon, Udacity founder and Stanford artificial-intelligence whiz Sebastian Thrun predicted that within 50 years there would be only 10 institutions of higher learning left in the world. Udacity, he reckoned, might be one of them.
As of this month, that prediction is looking overblown. After a year in which almost every big-name university in the United States rushed to get in on massive open online courses, or MOOCs, the backlash is in full force. And no wonder: The idea of free online video lectures replacing traditional classrooms not only offends many educators' core values, but it threatens their jobs. Worse, the early evidence suggests the model may not work very well: A partnership between San Jose State and Udacity this spring ended with more than half the students failing. In the same spaces where advocates not long ago trumpeted the MOOC revolution, critics now warn of the "MOOC delusion."
As much as everyone wants to see college costs reined in, replacing thousands of professors and classrooms with a handful of websites populated by remote talking heads cannot be the answer. But before we throw the whole idea out the window, it's worth asking: Might there be a way that online lectures could complement the traditional higher-education experience rather than replace it?
Anant Agarwal, president of EdX, believes there is. Like Coursera and Udacity, EdX began by offering full-service online classes for free, taught by professors at Harvard and MIT, the initial partners in the venture. Unlike Coursera and Udacity, though, EdX is a nonprofit, which frees it from the expectations of venture capitalists bent on reaping millions from the concept. As a result, EdX has appeared less focused on getting big quickly and more open to experimentation in terms of how it can best serve professors and students. One of those experiments is what UC-Berkeley professor Armando Fox calls SPOCs - "small private online classes," as opposed to massive open ones. The approach is also known, less acronymically, as "hybrid" or "blended learning."
The basic idea is to use MOOC-style video lectures and other online features as course materials in actual, normal-size college classes. By assigning the lectures as homework, the instructors are free to spend the actual class period answering students' questions, gauging what they have and haven't absorbed, and then working with them on projects and assignments. In some cases the instructors also use some MOOC-style online assessments or even automated grading features. But in general they're free to tailor the curriculum, pace and grading system to their own liking and their own students' needs.
The notion isn't entirely novel. A similar approach has been popularized at the high-school level in recent years by Salman Khan, who encourages teachers to use his free online lessons to "flip the classroom": Students watch lectures at home and then do their "homework" in class. Freed from the need to prepare a lecture for each class session, instructors can focus their time on the rest of the educational experience - the individualized, hands-on instruction and collaboration that no MOOC can provide. In this model the online lecture starts to look less like a poor substitute for traditional classes and more like a 21st-century twist on the traditional textbook.
The early results are promising. At San Jose State - the same college where so many students failed the Udacity course taught entirely online - a SPOC partnership with EdX has gone much better. There, professor Khosrow Ghadiri used an online circuits and electronics course taught by Agarwal, the EdX president, as part of a flipped-classroom model for two of the three sections of his required engineering class. At home, students would watch Agarwal's lecture, then fill out a survey designed to gauge which parts they understood and which gave them trouble. Ghadiri spent the first part of each class reviewing the parts that proved most problematic. Then they'd break into groups of three and work on solving problems together, after which each student would be quizzed individually on the day's material. Ghadiri told me the students were skeptical at first. But as the semester progressed, they consistently outperformed their peers in the nonflipped classroom on the quizzes. And in the end, 91 percent passed Ghadiri's course - a huge improvement, Ghadiri says, over the 65 percent average pass rate over the past seven years.
In other cases the approach has allowed instructors and students to tackle high-level material that they might not have attempted otherwise. Jaime L'Heureux, an information technology professor at Bunker Hill Community College near Boston, told me that she signed on to co-teach an experimental SPOC using online materials from an EdX class on the Python computer programming language. It was a daunting assignment: Neither she nor her co-professor were fluent in Python. And although the online course was billed as introductory, it was taught by an MIT professor and geared to MIT students. But L'Heureux said EdX worked with her to adapt the material to a slower-paced syllabus, and the flipped-classroom model allowed her to learn along with her students. Even then, half the students ended up dropping out. But those who stuck with it all earned a B-minus or above.
Asked whether she was worried that the MOOC material made her replaceable, L'Heureux laughed. Her students would never have made it through the course, she says, without both the instructors' hands-on help and the persistent motivation of having to come to class and work through the difficulties alongside their classmates and the people who would be assigning their grades.
EdX isn't the only MOOC provider experimenting with hybrid classes. In fact, Coursera co-founder Andrew Ng has been enthusiastic about the model from the outset, and several of Coursera's university partners have adopted it in one form or another, including Duke and Vanderbilt. "We're not interested in replacing professors," Coursera's partnerships manager, Connor Diemand-Yauman, told me. "When it comes down to it we understand the instructors' place in an on-campus educational experience."
Key questions remain, the biggest of which is whether a flipped classroom using video lectures is really any better than one that uses good old textbooks. Ghadiri believes students find the videos more engaging and are more likely to actually watch them than they are to complete their assigned readings. Ian Bogost, a Georgia Tech computer science professor and acute MOOC critic, acknowledges that replacing textbooks with MOOCs might make the material more accessible to some otherwise unmotivated pupils. But in an essay last year, he asked, "If the lecture was such a bad format in the industrial age, why does it suddenly get celebrated once digitized and streamed into a web browser in the information age?"
Whether SPOCs amount to some sort of pedagogical revolution, it seems clear that they hold more promise than pure MOOCs when it comes to delivering students a full educational experience - not to mention saving academics' jobs.