The Devil’s Bargain of Massive Online Learning

Georgia Tech has become the first university to offer a masters degree based on massive open online course (MOOC)-style courses in partnership with MOOC provider Udacity. Costing less than $7,000, the offer of the online course attracted about 2,360 applicants of which, 375 have started their degree with courses that cover machine learning and artificial intelligence for robotics.

Already though, administrators at Georgia Tech have been grappling with the unique set of problems that the new MOOC students have brought to the college. The main issue is one of “calibrating” students to the challenge that they have set themselves, especially if they are working full time and have been out of full-time education for a significant period.

This article by David Glance originally appeared at The Conversation, a Social Science Space partner site, under the title “The devil’s bargain of online learning that technology can’t change”

The challenge of online education is not unique to elite universities like Georgia Tech. A report released recently by the Public Policy Institute of California has detailed an analysis of 10 million students at California’s community colleges, of which 11 percent are now enrolled in online courses. The results were mixed. Short-term outcomes for the online students were worse than those attending courses on campus. Failure rates were about 11 – 14 percent higher for online students. Grades were also lower. These differences also varied depending on the race, gender and age of the students. White or Asian, older and female students did better than younger, male, African-American students.

The paradox was that students who took online courses were more likely to go on to a full degree program. So overall, more students were getting an education through the online offerings but this was coming at a cost for their overall short-term outcomes.

The results of this study illustrate a number of things when looking at the use of technology in education generally, but also online education specifically, whether it is a MOOC or any other form of offering. The first point is that students are not a homogenous group who will respond uniformly to any given approach or application of technology. This makes it very hard, if not impossible, to treat education as a process for which there are undiscovered efficiencies to be exploited, even though this is something that politicians, including Obama in the US, have been extolling universities to do. The simple fact of the matter is that education requires two participants in the process, with the student arguably being the one that at the end of the day is required to put in the most effort. This is not a process that can be sped up or done in the absence of that effort.

Efficiencies within the higher education sector can only be achieved if the expectations of both the suppliers and consumers are fundamentally changed. As it stands, students are paying for education irrespective of the outcomes. They expect in return that universities will do their utmost to ensure success. With a different approach, universities could just offer online courses with no support and simply accept high drop-out rates. This could be coupled with a notion of a pay-if-you-pass business model in which students would be willing to accept little assistance from the education providers if they were not required to pay unless they succeeded in passing the course.

Once the model is changed, technology could very well bring in the required efficiencies by reducing the need for full-time academics to steward the students. Without a change in approach, we are left with a labor-intensive process that requires effort from both the educators and the students. The role of technology in this process is still equivocal. A study of research on the impact of technology on education over the past 40 years has concluded that there is a moderate benefit, but as highlighted in the report on online learning in community colleges, the complexity of environments, student demographics, what is being taught and by whom, makes it very difficult to tell.

There may well come a day when artificial intelligence has developed to a point where personalized, one-on-one tutoring is possible for all students, ensuring the best educational outcome at the lowest cost. Until then however, efficiencies in the process of higher education will come at a cost of placing more of the onus on the student and accepting the outcomes. In this context, technology will indeed support managing larger numbers of students with fewer staff and politicians at least may be happier as a result.