Online, Size Doesn't Matter

Conventional wisdom (backed by many research studies) holds that students benefit from smaller classes. They receive more personal attention from instructors, who can spend more time evaluating each assignment turned in and can spend more time with each student. Many rankings systems reward colleges for small class sizes. Many potential undergraduates judge colleges on the availability of small classes.

But a large national study presented this weekend at the annual meeting of the American Economic Association challenges that conventional wisdom. The study finds that increases in online class size have no impact on student grades, student persistence in the course or the likelihood of students enrolling in future courses. The study focuses on courses that are not the size of large lectures or massive open online courses, but courses that are typical of those offered at many colleges and universities (in person and online). And the authors -- a research team from Stanford University -- write that their findings could suggest ways for colleges to save money, by enlarging online sections and cutting the number of instructors employed.

The study was conducted in online sections of courses offered by DeVry University, one of the largest for-profit institutions in the United States. DeVry's approach to course registration -- unlike the system used by most nonprofit colleges -- has students select courses without knowing which instructors will teach which sections. This eliminates the danger for a statistical study if students select popular instructors, who may have more success than others regardless of class size. When DeVry sections reach their enrollment caps, the university simply creates another section. In this experiment, the Stanford researchers were able to track outcomes for more than 100,000 students in nearly 4,000 sections of 102 different courses (undergraduate and graduate).

The control group consisted of courses of 16 to 40 students, with the average number being more than 30. The larger sections had an average of 3 additional students per section, or at least 10 percent larger than normal, with some classes being 25 percent larger. The course sizes examined were intentionally not MOOC-style, as DeVry focuses on classes in which an individual instructor is teaching and grading students. Courses studied included (without notable differences in the impact of the larger classes) a wide range of disciplines and different types of courses (some requiring more student participation, writing or lab work than others).

All of the classes required students to participate in online discussions, and all involved professors interacting directly with students through comment boards.

The paper makes no claims about in-person classes or very large online courses, but says that the study's findings provide "the first evidence that increasing class sizes in the online context may not degrade the quality of the class." And the paper says that "these results could have important policy and financial implications."

Those implications may not be good news for instructors, especially adjuncts.

The Stanford paper says that DeVry would need 4,146 of the standard sections to cover all of the students in this study. But if sections were shifted to the larger size -- going up in enrollment by 10 percent -- DeVry would be able to enroll all of the students in 3,763 sections. "From a financial perspective the cost savings is likely to come from professor compensation, as the marginal cost of adding a section is negligible otherwise. Many professors are adjunct and get paid by the section. Assuming the section wage rate is constant for the class size increases of this range, and assuming that enrollment is not a function of class size, an average 10 percent increase in class size will therefore reduce expenditures by approximately 9.2 percent, a significant savings," the paper says.

Asked if faculty members should fear the results of this study, Eric Bettinger, associate professor of education at Stanford, and co-author of the study, said via email that he wasn't sure. "As higher education becomes more digital and moves toward online programs, the role of professors changes more generally. These findings and others provide feedback about how best to use professors and how traditional roles might change in the digital future," Bettinger said.

The other authors, all at Stanford, are: Christopher Doss, a doctoral student in the economics of education; Susanna Loeb, the Barnett Family Professor of Education; and Eric Taylor, a doctoral student in education.