Professor Keith Devlin is a Stanford University mathematics professor who recently finished teaching his very first MOOC, or massively open online course. For those who have been living under a rock, a MOOC is course that anyone with an internet connection can attend for free, which in effect expands a university’s reach from their own tuition-paying students to millions of students around the world. I asked Professor Devlin how he thought his first experience went.
“Compared with the launch of Apple Maps, I think we did really well,” he replied.
Professor Devlin, otherwise known as “The Math Guy” on NPR, has written 32 books and over 80 articles on different aspects of mathematics, contributes articles about mathematics, education, and technology on Huffington Post, is the co-founder of Stanford’s H-STAR institute, and has been teaching the course “Introduction to Mathematical Thinking” — his MOOC — for over 40 years.
Professor Devlin taught the course via video lectures and weekly, live tutorial sessions using the platform Coursera. He went into the class knowing that his first time teaching a MOOC would be a beta experiment. “We had 64,000 students sign up initially, which is large for a beta, but there you go,” he said.
Knowing the high rate at which students who initially sign up for MOOCs later drop, Professor Devlin designed the course so that students could “get something of value” within the first three weeks of the course by focusing on logical thinking and real-life examples. Once he began to dive deeper into problems that required more challenging mathematics in the fourth week, students started to drop out. By the fifth week he was down to 10,000 students. Fifteen hundred students took the final exam in the sixth week — which he expected, he admitted.
It seems like quite a high dropout rate. You could look at that data and think, yikes, a one in six retention rate of students? But when courses offer a certificate of completion instead of college credit, and there’s no financial investment in the course, it’s a lot easier to drop when the course work gets tough. But Professor Devlin still reached over 1,500 students in his course — three to five times the amount he’d reach in a large university lecture hall. The expectation, once these students had completed Introduction to Mathematical Thinking, was that if they took the final exam, they’d be able to take a first year university math course.
One thing Professor Devlin knew early on was that teaching a MOOC was less about him acting as professor, and more about building a community of learners who worked together in order to succeed. Early on in the course, he told students to form groups among themselves to work on the problem sets together. Because Coursera is still working to build out social networking platforms for students to collaborate, Professor Devlin encouraged his students to use any method possible in order to communicate from different locations, including Skype, Google, Facebook, or phone and email — some were even able to meet in person. Professor Devlin recognized the value in students working together, and acknowledged that this element is vital in making MOOCs successful. He would have preferred that his students could have done their group work on the Coursera platform, however, so that he could have tracked their progress better and collected data.
“The opportunity is immense if we can start to crack that data,” he said. “To have it all on Skype or Facebook is a bit frustrating because I don’t have access to it. Education is on the cusp of being data driven for the first time in history. When you have 64,000 students, that’s real data. It turns education research into a hard, data-driven science. Making MOOCs work is hard, but in our favor, we have massive data collections…If you can contour the data, you can do wonderful things. That’s the key to this.”
Professor Devlin found the trickiest part of teaching a MOOC to be the peer evaluations that came after the final exam was finished. The final exam asked the students to create mathematical proofs. Unlike the coursework the students had completed throughout the semester, which could be graded by machine, the proofs the students completed for the final were more like essays, in terms of grading, than calculations. “Success of MOOCs will be determined by whether or not we can make peer reviews work,” he said. This is certainly true for MOOCs in the humanities, arts, and even sciences, as evaluation within those courses generally require more than just multiple choice questions.
So he gave out a rubric to students, and had them do practice peer reviews prior to the final, in which they graded, based on the rubric, sample proofs Professor Devlin wrote. The students were given three exams to grade, and also graded their own. Their grade on the final exam was determined by averaging those four grades. The final exam made up less than half of their grade in the course, the other part based on the coursework they completed throughout the semester.
While Professor Devlin encountered some bumps along the road, he was ultimately satisfied with his first experience in teaching a MOOC. “With peer evaluations working tolerably well, I think we’re going to have something of value,” he said of teaching the course. “I had the belief that it was doable; otherwise I wouldn’t have spent so much time on it…I went into the course with the suspicion that creating a MOOC was possible. I ended the course with the belief that it is possible. If I can make it work much better next time, I’ll upgrade that belief to a claim.”
And while he was pleased with the success of his first experiment Professor Devlin acknowledged that MOOCs can in no way come close to courses taught in physical classrooms, as he believes those are where the real learning takes place. “Is a MOOC as good as a physical class taught by me? No. Is it as good as being taught by a real person in a classroom? Not at all, no,” he said. “Nothing’s going to come close to talking about your mathematical attempts with someone who knows more about it than you do. But very few students — maybe only a class of thirty — experience that.”
Now that a few colleges have begun to give course credit for MOOCs, it will be interesting to see how MOOCs factor in to a typical college experience. Professor Devlin believes it’s inevitable that smaller universities will start providing credit for MOOCs.
“Fiscally responsible school boards are going to push universities to do that,” he said. “Smaller universities are not saying, ‘We’re going away.’ They’re saying, ‘We’re not going to pay premium dollar for lecturers, because we can get them from Stanford and MIT and places for free. What we’ll do is get really talented people who can work with the students. Because no matter how well I can make peer evaluation work in my MOOC, it’s not going to come close to being in a classroom with someone who knows mathematics who can look at your individual work.”
Interested in learning more about Professor Devlin’s experience? Check out his blog, MOOCtalk, here.
What do you think of the MOOC experience? Have you ever taken one yourself? Sound off in the comments below, or tweet your thoughts @Technapex or @ce_doyle.