Disaggregation vs. de-coupling of decisions

By Gregor Kiczales posted on May 31st, 2016

A common theme in discussions about MOOCs and other forms of innovation has to do with disaggregation or unbundling. The idea being that as offerings now provided by universities move online, some of them may end up being offered by other providers. Alex Usher touched on this last week. (Alex seems to be warming to the idea that MOOCs and the digital learning thing might have impact after all.)

I’ve written before about teaching assistants being one possible pressure point for unbundling. I still think that’s the case. Companies like Tutor.com and CourseHero.com are already well established. The Jill Watson story shows us another approach providers might soon be able to use to provide online TA support. Once enough students are paying alternative providers for TA support, they may begin to ask the university for a discount in exchange for not using university TA office hours. Or they may ask us to provide TA support on par with the 24/7 support provided commercially. Either way would be painful for universities. The economics of TA support is that the many students who make little use of TA support subsidize the small number of students who make extensive use of the service. In addition we use TA support to cross subsidize research by employing graduate students as TAs. So any significant change in the economics of TA support could have significant ripple effects.

But recently I’ve been thinking about unbundling in a different way, not in terms of unbundling services but instead in terms of decoupling key decisions over the course of a degree program. Consider the new MIT Micromasters in Supply Chain Management. In brief, the idea is to offer the first part of a professional Masters degree online. The Micromasters pulls part of MIT’s existing Supply Chain Management Masters program out into a separate offering; so clearly it is a case of unbundling an existing offering by offering part of it through a digital channel.

But looking at the Micromasters that way doesn’t capture all the value it provides learners.

In a traditional Masters program there are three big decisions. The learner decides to apply, the school decides to accept, and the learner decides to enrol. All the actual learning and engagement with the program follows the last of those decisions.

Deciding to apply is relatively low risk, since very little is committed other than the application fee. But the accept and enrol decisions are much more significant and are based on relatively little information, so they involve much more risk. If it doesn’t work out, the learner may have spent a lot of tuition, given up a job, moved, or declined other opportunities. They may also feel worn down and demoralized. The school will have lost the opportunity to accept a more successful student.

In the Micromasters model, the three big decisions are broken up into a chain of smaller decisions:

  • The learner chooses to take the first course in an MM series.
  • They can then choose to get a verified certificate for that course.
  • They can then choose to take the other courses in the Micromasters, one or more at a time.
  • They can then choose to get the Micromasters certificate.
  • They can then choose to apply to the residential Masters program.
  • The school can make an admissions decision informed by the learner’s performance in the Micromasters courses.
  • The learner can then choose to enrol.

The early decisions involve much less commitment. Without certificates the courses are free, and one course is a light enough load that it can be pursued after work, without moving, etc. So little ventured, for a modest gain.

If the learner sticks with it, then by the time they apply for the residential Masters program, both parties have a great deal more information to go on. By then it is going to be pretty clear to the learner that they like the material and they like the school’s approach to teaching it. Their performance in the online courses will be a good indication of their aptitude for the material. At that point, from the learner’s perspective the decision to stop working, move to Cambridge and enrol in the program will be far less risky.

From the school’s perspective, MIT sees this as “widening the admissions funnel.” Because more people can try to take the online courses, they will be getting a look at far more potential candidates than before. There may also be candidates who would not have “looked good on paper”—perhaps they did poorly as an undergrad—but who do extremely well in the Micromasters courses. This allows the school to accept learners who might have seemed too risky before. It may also allow them to reject students they might have accepted and regretted before.

The classic unbundling perspective is still useful for thinking about a lot of what will happen as the digital education restructuring plays out. It will certainly help understand the TA economy, student services outsourcing, online content provision, practice problem banks, crowdsourced exams, etc. But the unpacking of key high risk decisions lens is useful for understanding the real value to learners and schools associated with the unbundling of programs into online and residential components. Anytime we can delay the apply/admit decisions until after some of the engagement, then we are reducing risk for both the learner and the school. In the undergraduate space AP exams have played a similar role to date, but we may start to see explicit micro-bachelors programs at some point soon.


Gregor Kiczales is a Professor of Computer Science and the Provost’s Fellow for Flexible Learning Strategy.

Leave a Reply