With more and more students studying online, what actually helps students?

The spread of coronavirus throughout the U.S. has led to an unprecedented experiment in virtual learning education. With schools across the country moving to an online learning model for the rest of the semester, we decided to take a look at the research on virtual learning and the potential disadvantages and advantages of virtual learning.

“Virtual learning” typically refers to a course that is completely virtual. Students receive instructional content, submit assignments, take tests, and interact entirely online, or virtual. This can happen “synchronously” (e.g., all students joining a live hangouts session at the same time) or “asynchronously” (e.g., students logging in to do homework whenever they want). “Blended learning” refers to a course that mixes online learning and face-to-face learning together.

Many researchers divide online interaction into three types: student-content interaction, teacher-student interaction, and student-student interaction.

A substantial amount of the research in virtual learning has been in the context of higher education. Online learning programs for undergraduates and professionals have expanded dramatically over the past decade, and many researchers have taken advantage of such programs to research instructional design choices and to compare online courses to face-to-face courses directly.

Research on K-12 programs, however, is sparse, despite the increasing popularity of K-12 virtual learning programs. Most of this research focuses on state policies, professional learning programs, and expert recommendations. There’s been little empirical research on instructional design choices.

Yes, it can be. Meta-analyses suggest that online courses are about as effective as face-to-face courses. Blended learning courses, however, tend to be the best of all, with the important caveat that students also tend to perform more work in blended learning courses.

Virtual learning courses, for instance, can potentially improve nationwide community college retention rates. In a large, nationally representative sample, students who took an online course in their first year of college were more likely to have completed their degree after six years than those who hadn’t, in spite of being slightly less prepared for school.

Of course, whether any given online learning experience is equivalent to a given face-to-face experience depends on several factors. There’s huge variation in online programs.

One of the major risks of virtual learning is for the lowest performing students. For instance, one study explored whether an online course could help students recover credit after failing an algebra course by randomly assigning 1000 students to online learning and face-to-face conditions. Unfortunately, students in the online course reported the class as more difficult than those in the face-to-face course, were less likely to recover credit, and performed worse on a final algebra test.

Another study explored a large dataset from DeVry University, which uses a large number of online classes that have the exact same structure as their face-to-face classes (same instructors, same textbooks, same homework materials, same exams, etc.). Students in the virtual learning conditions had lower GPAs for the class and lower subsequent GPAs in future classes. It’s the lowest performing students, however, who drive this trend. In this study, the lowest performing students are the ones most hurt by the online format.

Most researchers agree that effective virtual learning requires more student initiative and effort than comparable face-to-face instruction. This may be why lower performing students struggle to do well in them; if students lack self-regulation skills to monitor their progress and discipline themselves.
​

Teachers face many challenges in the transition to online learning, but here are three notable ones especially salient to online learning:

Social isolation is one of the main challenges of online learning, especially for asynchronous courses. Students can feel like they’re not really connected to a learning community, they’re just performing some exercises alone. Students need to have the sense of “being there” and “being together with others”. This might be achieved through streaming software, where students virtually “attend” the class at a set time. But it can also be achieved through various techniques when course participation is asynchronous.

Participation in many virtual programs dips over time. MOOC dropouts are a widely recognized phenomenon, but participation dips even in courses that students are paying (or otherwise incentivized) to take. For instance, one study noted a drop in student interaction of more than 50% between weeks one and three, even though students chose their own platforms and instructors encouraged them to participate.

It’s also easy for teachers to feel overwhelmed in a virtual course. Not only does the change require learning how to use new technologies and grapple with various SNAFUs, it also changes teacher accessibility. In asynchronous courses, class is “in session” 24 hours a day, and students may be doing different parts of the course at different times. Misinterpretations of text-based explanations are also common.

Experienced online teachers and researchers suggest something simple: get the basics right first. Organize the material, provide a calendar, minimize tech problems, set clear expectations, provide frequent check-ins. In many ways, this is no different from a face-to-face course.

To make the virtual learning environment feel more like a community, it helps to use “social presence cues”: call people by name, use photos or avatars, provide personal anecdotes and encouragement. Although there is not sufficient evidence to link these techniques to learning outcomes, they do seem to encourage student participation and result in higher student satisfaction.

Recent research has focused on two major components of most online courses: multimedia content (e.g., video, audio, animations) and online discussions.

Short, high-quality pre-recorded videos have dominated many college-level online courses and virtual learning platforms. But most research suggests that video doesn’t have much of an impact on learning outcomes—especially when the resources aren’t linked directly to those outcomes. Teachers tend to use video as a generic way to engage students without considering how the video is going to influence learning outcomes. Students also tend to mistake the feeling of being engaged with actual improved understanding of the target concept.

There’s even less evidence that higher production values result in greater learning outcomes. What’s important is that any visual you use serves some purpose: illustrating a process or concept; shows something the student couldn’t otherwise experience, or nurtures social ties through face-to-face contact.

You also might consider alternative modes of online learning​ to recorded video: audio, podcasts, live recordings, short animations. These are less expensive to make and can be equally effective. Overall, few studies show that adding multimedia elements of any sort really improves learning outcomes.

One way of using video or audio to your advantage, however, is as a way to provide feedback to students. For instance, use software like Audacity or Voicethread to record feedback to your students. Teachers report that providing feedback this way is easier than providing text feedback to students. Consequently, when teachers provide audio feedback, they tend to provide more feedback than they do through text. Providing feedback this way also increases a sense of teacher presence for the students. Other studies suggest that students prefer this kind of feedback.
​

Asynchronous virtual discussions are another class feature common to online courses and absent from face-to-face courses. Some consider these discussions to be the heart of an online course; it is usually where a lot of the student interaction occurs. But how to set up and manage productive discussion threads?

Posing a rich question is the first step. The discussion might be focused on resolving a problem, on completing a project, or debating an issue. Ideally, students should move beyond just discussing a problem, project, or issue, but move forward to synthesizing multiple ideas and applying ideas to new situations.

It’s tempting to think that the more involved teachers are on discussion boards, the more involved students will be. But that’s not the case. In studies exploring instructor involvement, higher levels of critical thinking were associated with fewer, more meaningful comments by instructions during discussions (rather than more frequent comments by instructors). It’s important, however, that teachers respond quickly to address inappropriate or off-topic comments.

Another promising approach is to use peer facilitators. In contrast to instructors, more participation by peer facilitators tends to result in more participation from other students. You might consider using one peer facilitator to start the conversation by posing questions and another peer facilitator to summarize the main take-aways from the discussion.

Research at the college level suggests that giving students course credit for participation in online discussions promotes participation. In one study, giving students a 10-20% participation grade based on their discussion participation resulted in significant increases in weekly student messages over groups where discussion was purely voluntary. Giving students more credit (25-35%), however, didn’t provide any additional benefit.
​

Teachers can help their virtual students by adding opportunities for student self-reflection. This helps students to manage their own learning experience.
​

​Several different approaches to self-reflection have research support. For instance: provide formative assessments to your students. If they miss a question, don’t provide them the answer—just provide them with links to resources where they can figure out what the answer is. Or ask them self-monitoring questions. How much time have they spent studying? How well do they think they will do on the test? How deeply do they know the material?

Of course, all of the science of learning principles still apply in the online context. If you can find a way to incorporate visualizations, retrieval practice, elaboration, and other science-of-learning techniques, all the better.
​
It’s always a challenging time to be a teacher. But this is a particularly challenging time. We welcome your thoughts and experiences.

-Ulrich Boser