How Can We Improve MOOCs?
Photo/ Ilonka Hebels
With help from Google, computer scientists at Carnegie Mellon have launched a new initiative to make massive open online courses (MOOCs) more effective. They’re starting by paying attention to learning styles.
Massive open online courses, or MOOCs, are one of the most-hyped recent developments in education.
As Eric Westervelt reported for NPR’s “All Things Considered,” “In theory, students saddled by rising debt and unable to tap into the best schools would be able to take free classes from rock-star professors at elite schools via Udacity, edX, Coursera, and other MOOC platforms.” These online courses seemed capable of fulfilling the internet’s early promise of making knowledge accessible to all.
But whether MOOCs can make good on that promise remains to be seen.
Despite the high enrollment numbers, according to many estimates approximately 90 percent of people who sign up for most MOOCs fail to complete the course. In addition, a few high-profile flops have made headlines in the past two years. For example, San Jose State University’s highly publicized experiment in offering for-credit MOOCs, in partnership with online course developer Udacity, fell far short of its goals.
“Completion rates and grades were worse than for those who took traditional campus-style classes. And the students who did best weren’t the underserved students San Jose most wanted to reach,” Westervelt said in the NPR segment.
The University of Pennsylvania released a 2013 study of MOOCs that said only approximately one-half of those who registered viewed a lecture and that completion rates for the courses averaged only 4 percent.
Even Sebastian Thrun, cofounder of Udacity, has acknowledged the failure of MOOCs to live up to their early promise: “Online education that leaves almost everybody behind except for highly motivated students, to me, can’t be a viable path to education.”
The major reason for these failures, posits Geoffrey A. Fowler in the Wall Street Journal, is that “for all but the most self-reliant, online learning can be isolating.” Disengaged students are more likely to perform poorly or drop out altogether.
Some instructors mitigate this problem by recording more audio or video segments instead of only providing written lessons. They also update content frequently, send motivational messages to students, or congratulate them for work completed. Online mentors, active discussion boards, quizzes, and other activities to break up lessons also have been proven to increase student retention and engagement.
Another tactic some online course providers have used is charging a nominal fee, from $30 to $90, to confirm a student’s participation in and completion of a course. Coursera found that students who had shelled out cash for a course were more likely to complete it.
Regardless of the delivery method, social interaction is key to effective learning, studies have found. “The most important thing that helps students succeed in an online course is interpersonal interaction and support,” Shanna Smith Jaggars, assistant director of Columbia University’s Community College Research Center, said in the Wall Street Journal article.
A recent study has suggested that blended learning, which combines interactive online components with face-to-face instruction, can be effective as traditional classroom instruction.
Now comes the recent announcement that Carnegie Mellon University has received a Google Focused Research Award to tap the potential of MOOCs. A multidisciplinary team of researchers will use data-driven approaches to develop techniques for “automatically analyzing and providing feedback on student work, for creating social ties between learners, and for designing MOOCs that are effective for students with a variety of cultural backgrounds.”
As part of the research, Emma Brunskill, assistant professor of computer science at CMU, and Kenneth Koedinger, professor at CMU’s Human-Computer Interaction Institute (HCII) and director of the Pittsburgh Science of Learning Center, will leverage machine learning techniques to personalize MOOCs for each user, identifying which subject areas the student has already mastered and which might offer additional learning opportunities.
The second component—led by Carolyn Rose, associate professor in CMU’s Language Technologies Institute, and HCII professor Robert Kraut—will consist of determining ways to improve retention through increased socialization opportunities, such as mentoring and team tasks. The two will also explore how to better identify students at risk of dropping out of the course and more effectively engage these at-risk students in coursework.
Finally, the team will research ways to make courses more fun and engaging through game play and culturally relevant content for users outside the United States. That component will be led by Amy Ogan, assistant professor in HCII, and Jessica Hammer, assistant professor in HCII and the Entertainment Technology Center.
Google will provide $300,000 in annual funding for two years, with the potential to extend the research to a third year.
CMU hopes that the research will help MOOCs live up to that early hype.
“Unless the MOOCs pay attention to how people actually learn, they will not be able to improve effectiveness, and will end up as just a passing fad,” said Justine Cassell, associate vice provost of technology strategy and impact at CMU and codirector of the Simon Initiative, a university-wide program to use science and technology to improve student learning.