How Coursera Could Walk the Talk About MOOC-wrapping

Daphne Koller, co-founder of Coursera, wrote an article in Forbes this week about the possibility of MOOC-wrapping and mixed models of online/traditional delivery that incorporate free globally offered online courses. I’m glad they are looking at this. I’ve been talking about this option quite a bit on this blog for quite a while, and I’m currently engaged in trying to put a wrapping project together here at Keene State around the Coursera-offered Passion-Driven Statistics course.

There are, however, a lot of obstacles to the MOOC-wrapping paradigm, most of them created by the manner in which Coursera chooses to offer courses. I will list some of the challenges we have faced in trying to do this here in the hope that if Coursera is truly serious about pursuing this they can address them.

  • No class preview. I want to pitch a class to a faculty member as a potential for a wrapped experience. The faculty member, quite rightly, wants to review the structure and the content of the course before committing, both so that they can adequately structure the wrapper and so that they feel comfortable with the content and approach. But there is no way to do that six months out from a class. We get maybe a video or a blurb. It’s very hard to sell faculty on blindly trusting the course. The solution would be to post OCW style materials well before the course — syllabus, reading list, a sample lecture.
  • Semester-unaware scheduling. For a direct to consumer product, the staggered start times are probably a plus — there’s always a new class starting at Coursera, right? For people trying to integrate these into a semester it is a slice of hell. About 90% of the catalog is unusable in a traditional semester course due to start time or end time issues. And yes, I know — the world has different start times, there are schools that are on the quarter system or trimesters. But a good portion of the world starts first semester courses around September and ends around December, and starts second semester courses in January and ends beginning of May. There could be more accomodation of that.
  • No regular offerings, few hard dates in advance. I’m taking an Obesity Economics Coursera course. While there have been some glitches, the content has been great, and I can see this course being embedded in several other courses we offer at Keene State. But when is it being offered again? I don’t have a clue. I understand Coursera wants the flexibilty to not re-run low-enrolled courses. But we have to finalize our semester schedule as much as eight months before running a course, and have to start planning well before that.

That said, there are some things Coursera is doing that help, and could be further advanced:

  • One to two credit courses are wonderful. Keep it up. The Obesity Economics course, a one credit equivalent experience, is a wonderful example of how many ways a low credit experience could be utilized in a class. We have four credit courses here, and they are supposed to be integrative. So you can imagine taking a standard 3 credit economics course and combining it with the 1 credit equivalent Obesity Econ course. Instant integration of econ with policy. Likewise, the Passion-Driven Statistics course we are working on is a “quarter” course, starting mid-semester. Here the small one credit nature of it allows us to run it as a two-credit course where one credit comes from the MOOC, and one credit comes from the wrapper.
  • Courses with a project also work well. Passion-Driven Statistics has students work on a project, and this has turned out to be a great help to our design of the wrapper. We assume that students are coming out of the MOOC with some sort of project, and we can use the wrapper to engage in student-teacher and student-student evaluation of that project in ways that are difficult in the MOOC. In coding terms, I tend to think of the project as the “return variable” from the MOOC subroutine. The MOOC kicks that back up to us, and we dig in to the result.
  • Drilling on Content is OK. Coursera has gotten some flack for the content-focus in some MOOCs. One interesting piece of MOOC-wrapping is that since a lot of the higher-order skills can be exercised in the wrapper portion of the class, the use of MOOC time to reinforce more lower-order stuff is less problematic. In fact, it’s a bit of a blessing if the MOOC is dealing with some of the entry-level stuff that the F2F instructor doesn’t want to spend class time on.

That’s probably enough now, although as we work on our project over here, I’m sure many other things will come to mind. So brace yourself for Part II…

Who is accountable at Coursera?

Coursera wants to be the Google of the education world. You can’t complain about your email if the email is free, right? And the same thing holds true with their courses.

So when things like this happen in the course I am taking, where the exact answers to pass the final are revealed by mistake during the process of taking the final, it’s just a bug. These things happen, right?

But think this through in light of the “let’s transfer MOOC credit in” model that schools like Antioch are looking at.

Think about Antioch’s options. They could transfer in this course, only to find out that passing it was trivially easy, and demonstrated no real aptitude in economics.

In that case, the Antioch brand is damaged. A credit transfer system is only as strong as its weakest link.

Alternatively, Antioch could hear about the snafu, and refuse to transfer this credit in. But in that case a bunch of students took a class for four weeks expecting to get credit only to find out that they are not going to get any credit because Coursera made a technical error. Four weeks of work down the drain.

The only thing that Coursera is offering that extends much beyond straight up OCW is an assessment framework. If they can’t guarantee that  — or if they don’t at least freak out about its failures — why would you ever transfer in MOOC credit without additional assessment?

And yes, this is partially an argument for why all xMOOC credit should be wrapped in a layer of authentic institutional assessment, if only to protect the value of your degree.

But it’s also a straight up question — who at Coursera is accountable? And to whom?

Unbundling vs. Embedding: Approaches to reuse of integrated course objects

For a long time we have talked about the great unbundling. Roughly stated, a college course consists of content, some activity around that content, and some credit/assessment. Using terms from Matheos and Siemens we can talk about Content, Interaction, and Accreditation (I understand that the terms there are applied at the broader institutional level, but they work on the micro-level too):


Many people talking about the future of education (myself included) assumed through much of the aughts the future of education would be more or less “unbundled”. Your content might be OCW. Your accreditation would be done by WGU or Straighter Line. Your interaction might be a cMOOC.

With these multiple providers you (or an education integrator) could assemble the content, interaction, and accreditation parts separately, from separate vendors.  Efficiencies of specialization and scale would kick in. And conflicts of interest would be reduced: you didn’t have to get your car inspected at the place that did your engine work, and so on. Freedom!

Of course, the future didn’t quite co-operate. As Martin Weller expressed it a while back:

[A]fter a decade of OERs, it’s interesting that we’re coming back to educator constructed courses. The vision might have been of learners constructing their own personalised courses from the vast array of content out there. And while this does happen to an extent, and social tools will help it happen more, it’s also the case that one of the core functions the educator provides is to structure content into a sequence that learners can follow and have trust in. The bargain they make is this – if I do the course you have constructed then I will come out with a certain understanding of the topic.

Initially I felt a little down about this. But thinking it through over the past couple years, I’ve realized that there are major problems with severing assessment from content, content from interaction, and so on. If you don’t believe me, hop on down to your kid’s school on standardized assessment day. The problem is that the core offering of any educational institution is integration and coherence. To some extent it’s just Wiley’s Reusability Paradox writ large: modularity requires decontextualization for reuse in multiple contexts, but education is a process of contextualization. The perfectly decontextualized objects you would need to accomplish unbundling do not serve the integrative needs of the classroom.

So the future is increasingly looking somewhat more bundled than it was three years ago.

But then how does reuse and repurposing happen? I’d like to propose an alternative model which I’m calling embedding (and which I’m sure a commenter will tell me already exists and goes by name X, but that’s why I blog things like this — to find out that stuff).

In an embedded model, there is no unbundling. Instead there is “wrapping”. So for instance, a reuse case might look like this:

  • Students take an xMOOC on nutrition, which contains its own internal content, accreditation, and assessment. It is the credit equivalent of 1 to 2 credits.
  • Around that experience we wrap a 1 to 2 credit face-to-face class that recontextualizes the xMOOC. For instance, one F2F wrapper of the nutrition class might be  12th grade class engaged in a project-based assessment of the healthiness of their cafeteria food.  A college economics seminar might wrap the xMOOC with the intention of publishing research on the effect of economics on healthy eating, and so on. A dietetics class might wrap the course minimally, but use the face-to-face time to build group cohesion, provide better quality feedback than the MOOC, and assign additional topics. Again, the second layer of wrapping is not unbundled from the first, but provides a layer of local contextualization for the xMOOC.
  • Finally, there is engagement of all these separate classes in a Community of Inquiry around the subject of dietetics, but unlike the internal xMOOC conversations that deal with knowledge duplication, the CoI is involved with the multiple repurposings of the class — the college class is seeing what the 12th graders are doing and so on. Just as the face-to-face class contextualizes the xMOOC locally, the CoI places the F2F in the broader context of the multiple imaginings of that class.

In other words, rather than “unbundling”, it looks a little like this:

Thoughts? What is this? A dsMOOC? A cMOOC where individuals are replaced with classes? A Community of Inquiry that takes an xMOOC as its content instead of a text? (It also reminds me of the days at CogArts where we’d wrap an entire class in a SCORM wrapper with a single complete/non-complete flag, and call it a BFLO. I’ll let you older folks figure that one out…)

Whatever it is, I like it, and I think it’s the future of place-based education.