Last week, during my visit to the University of Maryland for my presentation on “The Academic Life,” I met Ryan Sochol, Assistant Professor of Mechanical Engineering, who has interesting data regarding the merits of online versus in-class instruction. Here’s the story:
Ryan teaches an upper level undergraduate subject on Additive Manufacturing. Because of room-size limits, the class enrollment is limited to 50, and attendance is expected (and checked, via clickers). There are also graduate programs that depend on this subject, so an additional 20 graduate students are allowed to take the course online, some in the vicinity of the campus, others at remote locations. Ryan’s lectures are visible to the online students in real time; they are also recorded and posted for viewing at times of the student’s choice.
Now for the point: the online students typically do not log in with their clickers in real time, and there is no way they can ask questions or participate in the in-class discussions, even when viewing in real time. They are completely passive participants.
In other words, the online students get no live conversation.
And guess what. On the midterm exam, the online graduate students performed less well than the in-class undergraduates, especially on those questions that, during class, were similar to what had been discussed among the students. I wanted to probe this further, so Ryan and I shared a Skype session to dig into details. I also was given the link to the online lectures, so I could witness for myself how the class runs.
Ryan’s instincts about getting in-class engagement are terrific. He uses clickers not just for attendance but also for polling on various questions, just like the example I presented some months ago from the University of Arizona. What he does after a poll is to ask students who voted one way or another to justify their choice. He feels that the visibility of the polling data, where a student can see that other students shared their particular viewpoint, plays a significant role in lubricating class discussion. And in the sessions I viewed, there was plenty of discussion. His students are not shy, a result of his encouragement and his welcoming style. But there’s more.
Toward the end of class, Ryan sometimes poses a question for the students based on the material presented in class. He pauses his lecturing and encourages the students to speak with their neighbors, come up with ideas, and he invites students to volunteer to come to the front and sketch their ideas on the slide panel for display. In this Mazur-like encouragement of conversation among peers, Ryan not only specifically encourages that peer conversation, he finds that the conversation itself emboldens students to come forward. Ryan feels that if he were to ask for volunteers without the “talk to your neighbors” pause, he would not get such good responses.
The online students are shut out of this exercise. And the online students do significantly less well on the exam questions derived from this kind of exercise, in spite of the expectation that as graduate students, they might be more mature and stronger students.
From my viewpoint on the role of live conversation in education, I would translate this to say that the cognitive systems of the online graduate students are not as actively engaged (through only passive observation) compared to the in-class undergraduates, who have three types of opportunities for active engagement: polling, explanations supporting polling choices, and peer discussions in response to lecture-based questions. This is a clear win for in-class instruction, and for live conversation in those classes.