A college education today is ridiculously expensive, and tomorrow it will be even more so. Can the Internet change that? Some people are hoping it will. They argue that online courses, or more specifically, massively open online courses (MOOCs) will make “the best courses, from the best professors, and the best schools” available to the masses at a fraction of the cost of a brick-and-mortar education. MOOCs would solve the problem of a hefty college price tag while improving everyone’s educational experience to boot. But far from overturning the staid and overpriced traditional lecture model of education, MOOCs reinforce that model and conflict with recent research on how to teach technical subjects like science.
The traditional classroom lecture has been a primary component of a college education for about as long as colleges have existed, but this approach to teaching makes some bad assumptions about the purpose of education, particularly education in science. As Nobelist and physicist Carl Wieman has argued, classroom lectures wrongly assume that the role of the instructor is simply to transfer knowledge, “as if it were bits of information, to the receptive students, much like pouring water from a large jug into a set of small receptive cups.”
Far from overturning the staid and overpriced traditional lecture model of education, MOOCs reinforce that model and conflict with recent research on how to teach technical subjects like science.
If education were primarily about transferring knowledge, then naturally you would want the most knowledgeable sources—big-name professors at Ivy League schools—pouring knowledge into as many small, receptive cups as possible. MOOCs are premised on that idea, offering hundreds of thousands of prospective students online access to the most elite faculty, and in some cases elite degrees, like Georgia Tech’s $7,000 master’s degree in computer science. Via MOOCs, prestigious professors can reach many more students than ever before, including students attending colleges at the lower rungs of the academic hierarchy. Those less prestigious schools can use MOOCs to turn their own classrooms into satellite units of Harvard and MIT.
While the best MOOCs have high production values and are designed using the latest research on how to transfer knowledge online, recent studies on how students learn science show that the key to effective teaching does not require the best professor from the best school, but rather an instructor who understands how to coach students in scientific thinking. Athletes, musicians, and dancers improve their skills with a coach who carefully observes their performance and provides feedback; similarly, the most effective science instructors follow their students’ performance on a problem-solving task and provide immediate, targeted feedback. Carl Wieman, along with two colleagues, recently demonstrated that students in a large-enrollment physics class learned more from an inexperienced instructor who used new teaching methods than they did from an experienced, well-regarded professor who gave traditional lectures.
What are these new teaching methods? They are a nearly complete inversion of the traditional approach to teaching: what used to be homework is now done in class, while lecture material is read outside class. These new methods are grounded in the idea that classroom time in science courses should be primarily spent practicing scientific thinking by “making and testing predictions and arguments about the relevant topics, solving problems, and critiquing their own reasoning and that of others.” Rather than spending class time on lectures that present factual knowledge, students spend nearly all their time on problem-solving tasks. You get your factual knowledge outside of class from textbooks, or, yes, even online videos. A crucial part of this new approach to teaching is immediate, targeted feedback from other students and from the instructor. The instructor gauges the students’ performance on a task using tools like classroom clickers, and then adjusts the classroom discussion accordingly. Wieman and his colleagues found that when they used these methods, student attendance increased by 20 percent and average test scores increased by 33 percent.
Scientists at other universities are finding the same thing, including instructors in Northwestern University’s Gateway Science Workshop and Washington University’s Genomics in Education program. (Full disclosure: I work at Washington University.) The innovative features of these programs will be very difficult to replicate in MOOCs. The large, moderated online forums featured in many MOOCs are no match for the focused, real-time feedback from engaged instructors and classmates. Truly effective science education at the college level can’t be mass produced, but the good news is that innovators in science education are creating tools and resources that can be shared by all colleges, including the low-cost community colleges that provide the lion’s share of college teaching in America.
One final reason to be wary of brand-name MOOCs is the risk that colleges will start to resemble big box retail stores—always the same, no matter where you go. The pre-packaged content of MOOCs requires a lot of up-front work, and that means that mid-course adjustments are not feasible. In February 2001, when the initial results of the Human Genome Project were published, I was enrolled in a class called “Eukaryotic Genomes.” Once the human genome papers came out, our professor quickly modified the lecture schedule, incorporating the new results into the class and pushing us to evaluate the new findings in light of what we had previously learned. It was a chance to engage with up-to-the-minute science, and all over the country similar classes were also taking stock of this research. You can bet that different conclusions were reached in each of hundreds of different classrooms around the country, which is a good thing, because learning how to think independently is more important than learning what the best professors at the best schools think.