What is the optimal size for an online course?

Inside Higher Ed reporter Susan D’Agostino’s recent article about the surge in online classes at Virginia Tech intrigued me for a number of reasons.

Ms. D’Agostino leads off her article with a story about a course, Concussion Perspectives, offered by Stefan Duma, a professor of engineering and director of the Institute for Critical Technology and Applied Science at Virginia Tech. With in-person enrollments doubling every term from 50 to 100 to 250 and then 500, Professor Duma offered the course online using recorded lectures in an asynchronous format. More than 1,000 students enrolled.

Professor Duma’s courses were not the only asynchronous online courses at Virginia Tech with more than a thousand students enrolled. Ms. D’Agostino reports that associate professor Gregory Tew, a faculty member in the College of Architecture and Design, moved two classes online – Design Appreciation and Life in the Built Environment. This semester more than 5,000 students enrolled in the two courses combined. Professor Tew maintains that his courses are better taught online than they were in large lecture halls where students were distracted by their devices and attendance was sketchy. Professor Tew states that students report to him that they feel more connected in an online class with 3,500 students than they do in their live classes. His courses were designed with flexibility in mind, including the ability to watch recorded lectures.

Ms. D’Agostino writes that Virginia Tech students appear to be enthusiastic about these large online courses. Professors Duma’s and Tew’s student course evaluations received above average scores in “overall teaching effectiveness.” Ms. D’Agostino cites a research study led by Justin Ortagus that reported better outcomes toward degree completion when first year students took some online courses in addition to their in-person courses.

After her interviews with the Virginia Tech professors about their experiences teaching online courses, Ms. D’Agostino poses the question: “What About Outcomes?” Mr. Ortagus, the researcher who reported the positive outcomes when first year students take some online courses, said “’When you have over 1,000 students, even if you have the most compelling lecture in the history of civilization, it’s very challenging to continually interact with students and have an engaged course experience.”

According to professor Tew, he has taught 30,000 students in online courses over the past five years without administrative or graduate teaching assistance support. Ms. D’Agostino interviewed Columbia University’s Center for Technology and School Change research and evaluation consultant Fiona Hollands who questioned the impact on students when they have to navigate through thousands of discussion board posts each week. Another interviewee, Luke Dowden, chief online learning officer and associate vice chancellor at the Alamo Colleges District, had similar concerns. Mr. Dowden questioned if appropriate online course design standards had been considered when the class size increased from a live lecture to a larger, asynchronous online course. “Does a discussion board assignment that says ‘read four posts and respond to two’ scale with 1,000 or 1,500 people?” I might also ask if it meets the requirements of the Title IV program for online classes qualifying for Federal Student Aid through interactive discussion board posts versus correspondence courses that do not?

According to Ms. D’Agostino, Virginia Tech does not appear to be interested in finding out more about the outcomes of these online courses versus in-person classes. Her source was an email received from Mr. David Guerin, the associate vice provost for communications at Virginia Tech.

Professor Tew acknowledged that student evaluations are not comparable to studies of learning outcomes. However, he replied “at the moment, it’s all we have. College has basically become – maybe has always been – a way to learn for people that are motivated enough to learn on their own.”

As I reread Ms. D’Agostino’s article, I couldn’t help but reflect on the differences between online courses offered as part of the curriculum of a very large traditional university versus online courses offered by a predominantly online university. In the 16 years that I led the American Public University System, we constantly observed, read, and discussed ways to improve learning outcomes for our students, all of whom enrolled in online courses. Not only was this necessary to improve learning outcomes and student retention, but it was also expected by our institutional accreditor and specialty accreditors that we implement policies and practices for the assessment of learning and demonstrate the implementation and regular review of those practices. I do not believe that those policies and practices were limited to online institutions only.

Almost every college has its large lecture courses. Many of these courses have weekly sessions led by teaching or graduate assistants offering assistance in smaller sections. Professor Tew’s comment “college [is] a way to learn for people that are motivated enough to learn on their own” may apply to a selective university known for its engineering and science programs. However, it’s well known that most colleges admit almost all students who apply. Is it reasonable to expect that all students are motivated or capable of learning on their own? I would expect a less selective institution that expected students to learn on their own to have low student retention rates.

At APUS, we capped our online class size at 25 with a few STEM classes capped at 18. A recent Online Learning Consortium blog post by Rebecca Thomas states that class sizes less than 30 are “optimal for certain pedagogies and course designs.” Much of the asynchronous course designs incorporate best practices from constructivist learning theories such as the Community of Inquiry. Learner-centered engagement is part of constructivist learning theory. Lengthy recorded lectures are difficult to stimulate engagement.

As more and more colleges and universities offer online courses, they should research and implement best practices in online course design. They should also implement better assessments of learning outcomes than just course grades and end of course student evaluations. Unlike two decades ago when online learning was in its infancy, evidence is available regarding best practices for achieving better student learning outcomes as well as student persistence. Students deserve it, employers should demand it, and third-party evaluators should demand it as well.

Subjects of Interest


Higher Education

Independent Schools


Student Persistence