How can online college be both “promising” and “predatory”?

As with health news about chocolate or wine, higher ed news about online college can sometimes seem contradictory.

One recent study, for example, found that online students who pursue undergraduate degrees at nonprofit and for-profit colleges graduate at lower rates than their in-person counterparts. The fact that students self-select into online courses was deemed “unlikely” to explain the outcomes, which included less desirable student loan repayment rates. As a result, the study from the U.K.-based Centre for Global Higher Education asserted that online college is a form of “predatory inclusion” in which access is paired with increased risks for students.

But in another recent study published in the Journal of Higher Education, students who voluntarily enrolled in at least one fully online college course were more likely to complete their courses, especially when controlling for time poverty. Paid work, parental status and age all contributed to the students’ time constraints. In this study, taking online courses did not negatively impact the students’ persistence to graduation.

To confuse matters more, both studies were published in peer-reviewed journals.

Part of the problem is that educational researchers who seek to test the efficacy of online learning in real-world settings face challenges that physical scientists working in controlled laboratories do not.

As a result, many educational studies have caveats, including an inability to consider all populations, control for all variables or depict causation rather than correlation. At the same time, some in the field suggest that questions about modality should not drive educational research. Instead, researchers should ask questions that center the problems students themselves are trying to solve.

“Educational research is really murky and messy and can be very complicated,” said MJ Bishop, vice president of integrative learning at the University of Maryland Global Campus, adding that the public sometimes understands the messiness better than some academic colleagues. “Why would we do anything online if it weren’t for the fact that we’ve got a whole boatload of people who need access to postsecondary education?”

Caveat: Not All Studies Consider All Online Students

When Christina Barlas, who grew up with a single mom outside of Baltimore and never aspired to attend college, felt unsatisfied in her post–high school career as a cosmetologist, she reconsidered college attendance. But Barlas, who had witnessed domestic violence as a child and survived a traumatic home invasion during high school, was uninterested in spending time on a college campus.

“The number of shootings going on now concerns me,” said Barlas, who enrolled at the University of Maryland Global Campus. “I feel safer and more secure taking online classes instead of going to a school.” Barlas, who radiates optimism and has worked full-time as a marketing assistant to pay for college, expects to earn a degree in digital media and web technology this year.

“Online is not predatory at all,” said Barlas. “I’ve found it to be very helpful, effective, efficient and affordable. Also, it’s fast.”

But Barlas’s positive experience does not contradict the study from the Centre for Global Higher Education that deemed online college “predatory.” That’s because she chose between attending online college or not attending college at all, whereas the study considered students who chose between online college or in-person college.

“Our data is not equipped to answer the question of whether students fare better in online education versus in no higher education at all,” said Christian Michael Smith, postdoctoral fellow at the Higher Education Race and the Economy Lab at the University of California, Merced, and co-author of the U.K. center’s study. “What we are able to say with our data is that, on average, online programs are associated with worse outcomes” than in-person programs, Smith said.

But an “average” human experience is an abstract concept that may not apply to individuals working to solve unique problems. On this note, Barlas’s story is a lot like the parable of the starfish, in which a boy standing on a beach littered with thousands of starfish is seen throwing one back into the ocean. When an old man tells the boy that his effort won’t make much of a difference, the boy nonetheless throws the starfish in the sea.

“It made a difference to that one,” the boy says.

Caveat: Not All Studies Can Control for All Variables

In isolation, Bunker Hill Community College’s six-year graduation rate (approximately 15 percent) compares unfavorably with Harvard University’s six-year graduation rate (approximately 97 percent). But many would cry foul on drawing conclusions from such a narrow comparison, especially given the two institutions’ vastly different admissions criteria.

That’s why researchers seeking insight into the differences between online and in-person outcomes often control for a range of variables.

“An important feature of our analysis is comparing online students to in-person students who are equivalent on a rich set of characteristics that we can measure,” Smith said of the research team’s effort to isolate modality as a variable. The team controlled for age, high school academic performance, parents’ highest level of education, whether the student was Pell Grant eligible and other characteristics. “We still see very wide disparities in outcomes we care about.”

But educational researchers may control only for a finite number of measured variables. As a result, the possibility exists that an unmeasured—or unmeasurable—student characteristic could alter a study’s finding.

Besides, variables beyond student characteristics may influence the outcome of an educational study considering online learning outcomes. For example, a course that is taught both in-person and online is often designed differently for each modality. That is, an in-person course that moves online may be tweaked to maximize the online modality’s potential. Also, the online and in-person versions of the course may have different instructors, which can impact students’ experiences.

“We design a controlled study and think it’s going to be a slam dunk in our ability to understand whether the online version did better or did worse than the face-to-face version,” Bishop said. “Quite frankly, the instructor that has been open to moving to online is probably more innovative, and that would make that an uneven comparison. Or vice versa. Someone might be really comfortable in the physical classroom … There’s an awful lot to unpack when we try to do these kinds of studies.”

Caveat: Many Studies Depict Correlation, Not Causation

Education researchers seeking to understand the differences between online and in-person learning outcomes are often unable to design randomized controlled studies, according to Claire Wladis, professor of mathematics at the Borough of Manhattan Community College, part of the City University of New York. Wladis’s research is focused on determining the extent to which online learning broadens access to higher education, and she is a co-author of the Journal of Higher Education study.

Without randomized controlled trials, students decide on their own whether to enroll in the online or face-to-face versions of the courses. As a result, many studies comparing modalities are observational, which, at best, suggests correlations rather than causation.

Observational studies on course modality often proceed in one of two ways, Wladis said. They either focus on students with similar profiles who pursue a range of different courses, or they focus on similar courses that enroll a range of students.

In the first option, researchers might consider, for example, time-poor students who self-select into different online and in-person courses. But in this case, the students chose the modality, and a growing body of research suggests that students often pursue courses they consider to be less important online, Wladis said. As a result, conclusions drawn from this kind of study are limited.

In the second option, researchers might consider, for example, a single course that is delivered both online and in person. But the populations of students who self-select into each modality of the same course are known to differ. Many opting into online are older, work for pay or have significant family responsibilities that make them time poor, as Wladis’s work has shown. As a result, conclusions drawn from this kind of observational study are also limited.

“It can be very difficult to actually isolate the effects of the online medium itself,” Wladis said.

Others underscored the point that creating a lab-like environment in educational research may be more complicated than doing the same in the physical sciences.

“Creating a very large data set makes it possible for us to run all our statistical analysis,” Bishop said. “But the limitations sections of those papers go on for pages and pages.”

Caveat: The Past May Not Inform the Future

In the investment world, a common disclaimer is that past performance is not indicative of future results. The statement may also apply to educational studies, especially those conducted before periods of great disruption.

“Crucially, our data only cover years before the onset of the COVID-19 pandemic,” Smith said. “So, we are cautious about extrapolating our findings fully to the present post–COVID-19 moment.”

To be sure, some institutions today perpetuate outdated claims about online education. But others assert that COVID spurred online teaching innovations, many of which are worth preserving and enhancing. Also, online student demographics are changing, including a surge of traditional-aged undergraduates opting to pursue college fully online. Also, colleges have raised their awareness that, while students are mostly self-reliant when it comes to troubleshooting tech problems, they still have a role to play in easing students’ technology burdens.

“Some features of online education will persist because of the simple fact that you are not in person,” Smith said. “It is not a completely different animal pre- and post-pandemic.”

Caveat: Online Modalities May Receive More Scrutiny

Face-to-face college outcomes may not receive the same level of scrutiny as online college outcomes, according to several researchers consulted for this story.

In the brick-and-mortar world, large, highly selective universities with endowments that rival the GDPs of small countries; broad-access institutions; and small colleges on the verge of closure are rarely lumped together to make a broad statement about the “average” in-person college experience, these researchers say.

“I’ve been screaming from the mountaintops [about this] since I was a graduate student,” Justin Ortagus, associate professor of higher education administration and policy and director of the Institute of Higher Education at the University of Florida, said. When he began his research program, some online program administrators shared their frustration about the unequal inquiry, especially given low completion rates in some in-person programs. They asked him, “Why not ask those questions about face-to-face offerings? Why only ask whether online education is doing poorly? Why not consider how to improve student outcomes in face-to-face options?” Many were motivated by genuine concern about retention and graduation rates in their in-person courses and programs, Ortagus said.

But many campus leaders may continue to view online courses with more skepticism than in-person offerings. For example, in an Inside Higher Ed survey released this month, college presidents reported viewing their in-person courses much more favorably than their technology-enabled ones. More than two-thirds said their in-person courses were of excellent quality, compared with fewer than one-third who said the same about their hybrid courses and one in five who rated their fully online courses as excellent.

A traditional, on-campus, four-year university may be thought of as a “paragon of higher education pedagogy,” according to Ortagus, but high-enrollment classes for entry-level students may make more financial sense for the college than pedagogical sense for the student. Also, some professors who are subject-matter experts may not possess an understanding of best practices for face-to-face teaching.

Other researchers echoed this sentiment.

“Face-to-face courses have become the standard by which the online courses are judged,” Wladis said, adding that, in her study, students who took online courses had slightly better outcomes. “Students make rational decisions on their own about what they need based on their resources.” In Wladis’s study, time was considered a precious resource.

Focus on the Student, Not the Technology

Online learning critics can rightfully point to examples of online learning that have not led to strong student outcomes, Scott Pulsipher, president of Western Governors University, wrote in an email. Yet many institutions today are investing in online learning options to enhance access, stay current, help their bottom lines and for other reasons.

“A moral hazard exists for many providers of online education where the benefits of tech-enabled learning—namely the ability to deliver it at scale more affordably than in person—are flowing more to the institutions than to the students,” Pulsipher wrote of the prospect of “predatory” online institutions. “If these institutions can’t deliver strong outcomes, then increased access is an empty promise.”

Centering the student, not the technology, may be sound advice for colleges that offer online options. But it may also be sound advice for educational researchers, according to those consulted for this story. The pandemic has raised both institutional and learners’ awareness that online courses and programs may enhance access for some populations, such as working adult learners. For this reason, some researchers say that one-word descriptors such as “predatory” are too simplistic and sweeping to be helpful in making sense of all online learning, especially given that such behavior is not limited by modality.

“It doesn’t serve students to suggest that online education can’t work,” Ortagus said. “It can work for some, and it can harm some. It’s important to parse through that nuance rather than just seeking one word to capture it all.”

Many of those consulted for this article have found themselves gravitating to neutral terms that center learners when attempting to make sense of rigor and quality in education—online or otherwise.

“If you understand the problem you’re trying to solve with the online, at-a-distance telecommunications tools that you’re using to deliver instruction remotely, then you’ve either solved that problem or you haven’t solved that problem,” Bishop said. “That’s the question.”