Study finds some groups fare worse than others in online courses
Online education is often held out as a way to increase access to higher education, especially for those -- adult students, the academically underprepared, members of some minority groups -- who have historically been underrepresented in college. But that access is meaningful only if it leads somewhere, and if the education students get helps them reach their goals.
New data from a long-term study by the Community College Research Center at Columbia University's Teachers College suggest that some of the students most often targeted in online learning's access mission are less likely than their peers to benefit from -- and may in fact be hurt by -- digital as opposed to face-to-face instruction.
The study did not, however, account for the quality of the online courses studied, making it difficult to draw from its findings overly sweeping generalizations about the efficacy of online learning.
The working paper, "Adaptability to Online Learning: Differences Across Types of Students and Academic Subject Areas," by Di Xu and Shanna Smith Jaggars, researchers at the center, expands on work from 2011 that found that students who enrolled in online courses -- controlling for various factors that tend to predict success -- were more likely to fail or drop out of the courses than were those who took the same courses in person. Notably, there was not a gap in completion between those enrolled in hybrid and in-person courses.
The new study is a follow-up prompted by questions from officials at the Washington State Community/Technical College System whose courses were examined. (The study examined the performance of 40,000 students in about 500,000 online courses.) "They asked us, 'So who? Is it all students who fare less well, or certain subgroups?' ” said Jaggars.
The answer is that virtually every group of students fared less well (defined by the number of course credits they completed, and/or by their grades) in online courses than they did in on-ground classes.
But some groups fared worse than others. Men showed a more negative effect from online courses than did women in terms of both course persistence and grades. Black students’ grades fell significantly more in online courses, as did those of Asian students. Students with stronger academic skills saw their course persistence and grades decline less in online courses than did students with weaker academic credentials.
Like other groups, older students were less likely to complete online courses than they were on-ground courses, though their grades were actually slightly higher. But traditional-age students saw their comparative performance decline such that while they outperformed adult students significantly in face-to-face classes, they lagged their older peers in online courses.
To the researchers, the working paper’s findings that “students who are already doing poorly in college do even more poorly when they take online courses” suggest several possible implications, said Jaggars. It may make sense, she said, “to restrict online courses only to students who demonstrate they do well in those courses.”
Other options would include incorporating into the sorts of lower-level courses in which struggling students tend to cluster training in online-learning skills, to help such students adapt better to online environments.
And most of all, the researchers suggest, colleges should focus on improving the quality of all online courses, to “ensure that their learning outcomes are equal to those of face-to-face courses, regardless of the composition of the students enrolled. Such an improvement strategy would require substantial new investments in course design, faculty professional development, learner and instructor support, and systematic course evaluations.”
The Study’s Implications
Jaggars acknowledged that the researchers did not do any analysis of the quality of the Washington State community college courses examined in the working paper. And that led numerous observers to urge caution in applying its results too broadly, as a New York Times editorial about the study arguably did last week.
The editorial focused on the terribly high attrition rates of noncredit massive open online courses and used the Community College Research Center’s study to extrapolate about online learning generally: “The picture the studies offer of the online revolution is distressing.”
Russell Poulin, deputy director for research and analysis at the WICHE Cooperative for Educational Technologies, criticized the New York Times editorial’s “over the top” conclusions but said the CCRC study was an important contribution to research on online learning.
He applauded the study’s focus on the need for students to “adapt” to online learning, and agreed that institutions needed to do a much better job ensuring that students are prepared to take online courses – not by restricting which ones have access to them, though, but by giving them more training in such courses, such as a required short course on online learning.
Poulin also said the study gave short shrift to the importance of building student services into online courses. “For the underprepared students that the study worries about most, student support services (advising, tutoring, library resource materials, study skills assistance, technical assistance) could be the differentiator,” Poulin said in a blog post. “These services may be readily available on campus, but might be available on a limited basis or not at all for online students. Those differences are not measured by the study.”
The Bill & Melinda Gates Foundation has been a major promoter and funder of the use of technology in expanding access to higher education – but it funded the community college center’s research, too.
Josh Jarrett, Gates’s deputy director of education for postsecondary success, said in an e-mail message that the foundation hopes the Columbia study will “add additional research rigor to the ongoing conversation about online learning.”
He said the working paper buttressed Gates’s view that blended courses, by meshing online and face-to-face components, is “the approach we favor for the majority of the low-income, first-generation students for whom we work.”
And Jarrett said the study should “caution us against simplifying the conversation into ‘online learning is good’ or ‘online learning is bad.’ I have seen really bad online courses and quite amazing online courses. We need to continue to understand what pedagogy, features, and supports lead to greater student success and encourage more of those practices – whether online or off.”