- Gates will fund $1.4 million research project to study MOOC-powered courses at U. of Maryland
- Report: Barriers to the rise of artificially intelligent tutors at traditional universities
- Candace Thille talks MOOCs and machine learning
- Massive open courses aren't answer to reducing higher ed inequality (essay)
- Liberal arts college explore uses of 'blended' online learning
Score One for the Robo-Tutors
Without diminishing learning outcomes, automated teaching software can reduce the amount of time professors spend with students and could substantially reduce the cost of instruction, according to new research.
In experiments at six public universities, students assigned randomly to statistics courses that relied heavily on “machine-guided learning” software -- with reduced face time with instructors -- did just as well, in less time, as their counterparts in traditional, instructor-centric versions of the courses. This largely held true regardless of the race, gender, age, enrollment status and family background of the students.
The study comes at a time when “smart” teaching software is being increasingly included in conversations about redrawing the economics of higher education. Recent investments by high-profile universities in “massively open online courses,” or MOOCs, has elevated the notion that technology has reached a tipping point: with the right design, an online education platform, under the direction of a single professor, might be capable of delivering meaningful education to hundreds of thousands of students at once.
The new research from the nonprofit organization Ithaka was seeking to prove the viability of a less expansive application of “machine-guided learning” than the new MOOCs are attempting -- though one that nevertheless could have real implications for the costs of higher education.
The study, called “Interactive Learning Online at Public Universities,” involved students taking introductory statistics courses at six (unnamed) public universities. A total of 605 students were randomly assigned to take the course in a “hybrid” format: they met in person with their instructors for one hour a week; otherwise, they worked through lessons and exercises using an artificially intelligent learning platform developed by learning scientists at Carnegie Mellon University’s Open Learning Initiative.
Researchers compared these students against their peers in the traditional-format courses, for which students met with a live instructor for three hours per week, using several measuring sticks: whether they passed the course, their performance on a standardized test (the Comprehensive Assessment of Statistics), and the final exam for the course, which was the same for both sections of the course at each of the universities.
The results will provoke science-fiction doomsayers, and perhaps some higher-ed traditionalists. “Our results indicate that hybrid-format students took about one-quarter less time to achieve essentially the same learning outcomes as traditional-format students,” report the Ithaka researchers.
The robotic software did have disadvantages, the researchers found. For one, students found it duller than listening to a live instructor. Some felt as though they had learned less, even if they scored just as well on tests. Engaging students, such as professors might by sprinkling their lectures with personal anecdotes and entertaining asides, remains one area where humans have the upper hand.
But on straight teaching the machines were judged to be as effective, and more efficient, than their personality-having counterparts.
It is not the first time the software used in the experiment, developed over the last five years or so by Carnegie Mellon’s Open Learning Initiative, has been proven capable of teaching students statistics in less time than a traditional course while maintaining learning outcomes. So far that research has failed to persuade many traditional institutions to deploy the software -- ostensibly for fear of shortchanging students and alienating faculty with what is liable to be seen as an attempt to use technology as a smokescreen for draconian personnel cuts.
But the authors of the new report, led by William G. Bowen, the former president of Princeton University, hope their study -- which is the largest and perhaps the most rigorous to date on the effectiveness of machine-guided learning -- will change minds.
“As several leaders of higher education made clear to us in preliminary conversations, absent real evidence about learning outcomes there is no possibility of persuading most traditional colleges and universities, and especially those regarded as thought leaders, to push hard for the introduction of [machine-guided] instruction” on their campuses.
“This study,” they later add, “supports a ‘no-harm-done’ conclusion regarding at least one current prototype” -- that prototype being the Carnegie Mellon software.
With no apparent difference in how well students master the material, the main significance of the no-harm-done thesis is how machine-guided software could help campuses curb the untenable costs of a college education. Quantifying these potential savings is an extremely slippery task, the researchers warned. But Bowen and company nevertheless conducted a “careful” inquiry of the relative costs of machine-guided education with a focus on long-term savings on personnel.
The Ithaka researchers decided the results of their cost inquiry were “too speculative and subject to considerable variation” even to include in the body of the study. But they did offer a taste.
In terms of instructor compensation, the researchers estimated, a machine-guided course featuring weekly face-to-face sessions with part-time instructors would cost between 36 and 57 percent less than a traditional course in which a full professor presides over each 40-student section; and it would cost 19 percent less than if a single full professor gave one lecture to all sections before breaking them into smaller discussion groups led by teaching assistants.
The perennial fear among faculty is that the growing credibility of automated teaching software could tempt administrators to replace instructors with robots. But Bowen and company make the case that automated teaching software could enable colleges to save money without firing tenured professors.
A hybrid teaching model could shift a great to deal of the teaching burden from tenured professors to teaching assistants and support staff, they explain. That could allow institutions to enroll more students without hiring an equal proportion of expensive tenured faculty. “Recruitment costs may thereby be reduced along with compensation costs per student, and debates over maintaining commitments to existing faculty are avoided,” the authors write.
Of course, assuring professors that automated teaching software will not take their jobs might not be enough to persuade them to deploy it in their classrooms. Another recent Ithaka study, led by Bowen and Lawrence Bacow, the former president of Tufts University, found various other barriers to persuading instructors to adopt courses designed around “Interactive Learning Online,” or ILO (the researchers’ preferred term) -- not least of all, the resistance of many professors to prefabricated courses that they did not invent and cannot modify to their own needs and tastes.
“The findings in this study warn against ‘too much hype,’ ” caution the authors of the new study. “To the best of our knowledge, there is no compelling evidence that online learning systems available today … can in fact deliver improved educational outcomes across the board, at scale, on campuses other than the one where the system was born, and on a sustainable basis.”
For the latest technology news and opinion from Inside Higher Ed, follow @IHEtech on Twitter.
Search for Jobs