MOOC companies still face challenges, such as dealing with low course-completion rates and proving that they can make profit. And they have a lot of convincing to do among faculty members, says Uriagereka. “Some salivate and can't wait to be a part of it,” he says, noting that his university had 20 volunteers for its 5 inaugural MOOCs. “Others say, 'Wait a minute. How do we preserve quality? How do we connect with students?'”
MOOCs are largely a product of one corridor in the Stanford computer-science department, where the offices of Andrew Ng, Daphne Koller and Sebastian Thrun are just a few steps apart. But they are also the fruit of research dating back to at least the 1990s, when the explosive worldwide growth of the Internet inspired a multitude of efforts to exploit it for education. Campus administrators tended to regard such projects as a sideshow — the higher-education financial crunch was not quite as serious back then — so most experiments were the work of committed individuals, departments or research centers. But with the relentless advance of technologies such as broadband, social networking and smart phones, researchers' interest continued to grow.
Ng got involved in 2007 because he wanted to bring Stanford-quality teaching to “the people who would never be able to come to Stanford”, he says. Following a path blazed by the open-source software movement, and by earlier open-source education initiatives, he started a project to post online free lecture videos and handouts for ten of Stanford's most popular engineering courses. His approach was fairly crude, he admits: just record the lectures, put them online and hope for the best. But to his astonishment, strangers started coming up to him and saying, “Are you Professor Ng? I've been taking machine learning with you!” He began to grasp how far online courses could reach, and started working on a scaled-up version of his system. “When one professor can teach 50,000 people,” he says, “it alters the economics of education.”
One of the many people he talked to about his work was Koller, who began developing her own online-education system in 2009. Whereas Ng looked outwards, Koller wanted to look inwards and reform Stanford's teaching on-campus. She particularly wanted to promote 'flipping', a decade-old innovation in which students listen to lectures at home and do their 'homework' in class with their teachers, focusing on the most difficult aspects or discussing a concept's wider implications. This lets the instructors concentrate on the parts of teaching most of them enjoy — interacting with the students — and relieves them of the repetitive lecturing that they often dislike.
Koller also wanted to incorporate insights from the many studies showing that passively listening to a lecture is a terrible way to learn (F. I. M. Craik and R. S. Lockhart J. Verb. Learn. Verb. Behav. 11, 671–684; 1972). Following an approach pioneered by other online developers over the previous decade, Koller broke each video into 8–10-minute segments separated by pauses in which students have to answer questions or solve a problem. The idea was to get them to think about what they had learned; the deeper their engagement, studies showed, the better their retention.
Finally, to encourage greater interaction among the students themselves, Koller took a cue from social-networking sites such as Facebook and gave her system an online discussion forum. As Ng explains, the idea was to extend what happens in a face-to-face study group: “Students sit with their best friends, they work on problems together, they critique each others' solutions — lots of pedagogical studies show that these more interactive modes of student engagement result in better student learning.”