Title: Demographics and Learning Behavior in MITx and HarvardX MOOCs
Group Meeting Date & Time: Monday, March 10, 2014 @ 2:00 pm
View the slides from Dr. Seaton’s Group Meeting
Daniel Seaton, Ph.D., recently joined the Office of Institutional Research at MIT as a Data Analyst. During his group meeting, titled “Demographics and Learner Behavior in MITx and HarvardX MOOCs”, Dr. Seaton demonstrated the immense data analysis that has been performed on the first 17 MITx and HarvardX courses offered on the edX platform.
All of the work he discussed during his group meeting is the result of a productive collaboration between edX, MITx and HarvardX. While the group meeting focused on the first 17 MITx and HarvardX courses, there have been 26 different MITx courses and 24 different HarvardX courses thus far. Dr. Seaton focused on the demographic and learning behavior of registrants in these first 17 edX courses. Demographic data, including geographic location of registrants, education background, and gender, from the first 17 courses are now publicly-available and the data can be self-explored in depth on the Insights website (see links below). In addition, MIT and Harvard have published working papers, or course reports, on registrant’s demographic data and learning behavior for each of the first 17 MITx and HarvardX courses.
Dr. Seaton highlighted how the data in the working papers demonstrate that course components, which includes lecture videos, textbook material, simulated laboratory components, homework, and exams, vary dramatically between courses. By color-coding each courses’ components, for example exams are in red, homework problems in gray, laboratory components in green, and compiling the data into a graphic, the variation across courses is clearly demonstrated.
In addition to the variation in course components and requirements, there is immense diversity among course registrants. Dr. Seaton illustrated this point by analyzing the amount each registrant viewed and explored a the chapters within both MITx and HarvardX courses, where a chapter represents the highest organizational unit in one of these courses. The number of chapters viewed clearly highlights how differently users are choosing to interact with these courses. Intuitively, in comparison to registrants who did not receive a course certificate, those registrants who received a course certificate were more likely to have explored almost all of the course chapters. While this holds true for most courses, some courses emphasized content exploration, leading to high levels of variation of chapter accesses.
Dr. Seaton emphasized how different the course registrants are from traditional students who may enroll in MIT or another undergraduate curriculum. More than 70% of the registrants are international and have diverse educational backgrounds. Particularly striking is that the majority of the students have a college degree. As a result, the motivation of the registrants to take an edX course varies dramatically and can explain their use of the courses’ materials.
The analysis of clickstream data, or the data collected from each registrants use or clicks on the course website, can be particularly informative. They can track the amount of time a registrant spends on the course website, analyze a registrants’ transitions between course components, and course component consumption. By comparing these data to demographic information, and in the future, registrants’ motivations for taking a course, we will understand more about how and why registrants take MOOCs.
Insights: Interactive Visualizations
As a Data Analyst in the Office of Institutional Research at MIT, Daniel’s current work focuses on the organization and analysis of learner data from MITx courses, applying techniques from the fields of learning analytics and educational data mining to understand student behavior. Previously, Dr. Seaton worked as a Postdoctoral Research Fellow under the guidance of Professor Isaac Chuang in the Office of Digital Learning (ODL). His role at ODL involved collaboration with the HarvardX Research Committee, leading to a number of Working Papers providing early insight into the first 17 open online courses from MITx and HarvardX. Dr. Seaton has also worked as a Postdoctoral Associate in Professor David Pritchard’s RELATE physics education research group at MIT. His work involved learning analytics, content development, and teaching a reformed pedagogy in flipped classrooms. This work culminated in studies of eText use by students in both residential and open online courses, as well as the first detailed analysis of student data from the inaugural MITx (now edX) course, 6.002x: Circuits and Electronics.
Dr. Seaton received his B.Sc. in Physics from Auburn University, and his Ph.D. in Physics from the University of Georgia. Under the supervision of Professor David Landau of the Center for Simulational Physics, his graduate work focused on the application of generalized-ensemble Monte Carlo methods to physical systems; exploring protein-folding with coarse-grained polymer simulations. Dr. Seaton was also a dedicated instructor at the University of Georgia, with experience teaching an introductory physics course, a freshman seminar based around Nintendo’s Wii console, and a number of undergraduate laboratories. His interest and experience in teaching, combined with his training in computational science, provided a bridge to his current work in physics education research and online learning.