Yesterday, I was on WBUR's RadioBoston with Matthew Chingos, discussing his new study about online learning in higher education. Matt's study involved recruiting several hundred Introductory Statistics students at several college campuses who were willing to be randomly assigned to either a regular class or a hybrid online class. In the hybrid online class, students took an online version of Intro Statistics mediated entirely by a computer, with online readings, quizzes, activities and so forth. They also met once a week for a discussion section to answer questions.
Matt and colleagues found no significant differences in academic outcomes between the two conditions: students in both courses did equally well in terms of grades, pass rates, and standardized measures of statistics knowledge. This falls in line with many, many other studies that have showed no difference in student outcomes between online (or distance) and face to face courses, and the website No Significant Differences is a good place to explore this research.
It's not just that the hybrid course is just as good though; it also took less time from teachers and students. The kids in the hybrid condition had 2 fewer hours of class per week (although they apparently spent an extra 30 minutes per week outside of class). Students learned the same amount with less time investment from students and much less investment in faculty. Its exciting to imagine how gaining efficiencies in certain courses could help students finish school faster and learn more and help universities allocate more resources to more complicated learning tasks. In the radio interview, we discuss more about what this means, how it would not apply to all subjects or learning goals, and why the results are promising.
Education policy makes strange bedfellows sometimes. Matt and I probably have some major differences in opinion about the appropriate role of free market inspired reforms in education policy. We do agree that technology is worth studying, and I bet we disagree about exactly what to do with technology that works.
In spite of any differences, I have to commend Matt and his colleagues for a very well conducted study. The authors used a randomized assignment to ensure that we could isolate the causal impact of the online course: we know for certain that the same kinds of students were in both regular and hybrid classes. The authors conducted the study at typical universities so that the findings would be more likely to generalize to broad populations; previous similar studies, for instance, had been conducted at Carnegie Melon University, and there are good reasons to believe that those students would simply be better at taking online courses than typical students.
The authors are very careful about how they present their findings, very upfront about their limitations, and very circumspect in their conclusions, and Matt echoed this careful language in his on air discussion. This study did not prove that all forms of online learning are the savior of education and humans can pack their bags and let computers take over. The study showed that in one subject area, we can prove conclusively that for academic outcomes (statistics learning) in this subject area, a self-paced online course with a 1 hour weekly section is as good as a class of 40 students taught by a professor. This kind of careful research is tremendously important to making good decisions.
And all of the tough decisions are still ahead of us. It will be bad news if we use this insight to de-fund education further and replace humans with computers. It will be good news if we allow online courses to teach what they are good at teaching in order to free up resources to direct human instructors to more high leverage areas: working with struggling students, advanced students, and more complicated topics.
I do have one criticism for Matt, and that has to do with his characterization of this research as "rigorous" and other kinds of research as "not rigorous." I'm going to poke him in the eye for that (online) in a couple of days, but for now I'll just commend he and his colleagues for a very useful contribution to the research on online learning.