Mooc completion rates ‘below 7%’

Open online courses’ cohort much less massive at finish line

May 9, 2013

Source: Alamy

To what end? Those who do not take assessment may still benefit, but Mooc completion rates are ‘not entirely meaningless’

The average completion rate for massive open online courses is less than 7 per cent, according to data compiled by an Open University doctoral student as part of her own Mooc studies.

Katy Jordan, whose PhD research focuses on online academic social networks, took time out from her doctorate to gather information on the number of people completing a range of free web-based courses. So far, she has tracked down information on the percentage of students completing 29 Moocs.

“It’s something I’ve been doing as a hobby,” Ms Jordan explained. “I started taking Moocs as a student in around March last year, and I’ve been a bit hooked ever since.”

Her latest Mooc - Introduction to Infographics and Data Visualization, run by the University of Texas at Austin - challenged students to produce a visual representation of some datasets, which is where she got the idea to look into Mooc completion.

According to her findings, which are based on local news articles, university documents, presentations and other information sources (including Times Higher Education), the average Mooc completion rate across the 29 courses was just 6.8 per cent.

The course with the highest rate of completion was Functional Programming Principles in Scala, from Switzerland’s École Polytechnique Fédérale de Lausanne and offered on the US Mooc platform Coursera. According to Ms Jordan’s research, some 19.2 per cent of the 50,000 students who enrolled completed the course.

At the other end of the spectrum was A History of the World since 1300 by Princeton University, also hosted by Coursera, which reportedly recruited 83,000 students with just 0.8 per cent reaching the end.

Five of the top six most-completed Moocs relied on automatic marking alone, meaning that no peer assessment was required. Courses that relied purely on peer grading generally fared far worse in terms of the percentage of students reaching the end.

The average completion rate for the 17 automatically marked courses was 7.7 per cent, while for the 11 that involved some degree of peer assessment, the figure was 4.8 per cent. For one of the Moocs Ms Jordan examined, the assessment method was unknown.

“Generally speaking, for most of the courses, completion rate is defined as people who earned a certificate - the people you could deem to have ‘passed’ the course,” Ms Jordan said.

Although she acknowledged that many people would benefit from taking a course even if they did not reach the end, she said completion rates were indicative of how successful a course had been.

“People might have no intention of completing assessment when they register, but I don’t agree that completion rates are entirely meaningless.”

Register to continue

Why register?

  • Registration is free and only takes a moment
  • Once registered, you can read 3 articles a month
  • Sign up for our newsletter
Please Login or Register to read this article.

Reader's comments (4)

7% of a lot of people is still quite a lot of people, so completion rates aren't really an indicator of anything other than completion rates. I've studied a couple of MOOCs (statistics and programming) and didn't "complete" either of them, simply because I'd got what I wanted out of them and the last bits were unnecessary. So is my failure to complete actually a failure?
As the article says, a failure to finish is not necessarily a complete failure. However, if so many people drop out, at what point does it stop being massive? And if it's not completed, then it's not a complete "course" is it?
But that's applying a traditional definition of a "course" to a new format. A course, as far as a MOOC is concerned, is what the person taking it deems it to be. Maybe you sign up for a Chinese course because you want to update your knowledge of, say, asking for directions. So you do that bit. Have you completed the course? Yes. Or no. Depending on whether you're the student or the researcher looking at completion rates. The Massive bit of MOOCs doesn't (always) refer to the number of people actually taking a course, it refers to the number of people who *could* take the course. I think that mis-definition is what contributes both to much of the criticism of MOOCs, and also to the failure of some of them. Let me use an analogy to answer your points. The M1 is a long motorway. But you can get on or off it anytime depending on what you want to achieve. It's still a motorway whether you travel one junction or do the whole thing. 60 million people living in the UK can use it if they want. So can 7 billion others if they want to and can get to it. It is potentially a "massive, open" motorway. But if that many people did use it at the same time, it wouldn't work. The fact is only a small fraction of the people who *could* use it, do actually use it at any one time. That doesn't stop it being "massive" or "open" or a "motorway". So I could do research like that discussed in the article and claim that "only 7% of motorway journeys are completed" - it may be true, but it's meaningless. MOOCs are the same as motorways. They are likely to work best if people can drop in and out at will (so "dropout rates" is an irrelevant concept, as is the traditional concept of "course") and if they are taken asynchronously where possible. A MOOC which demands synchronous attendance needs to be managed differently, probably needs a limit on the number of people - it's still "massive" no matter how low that number is - and one that offers credit needs to require completion - at which point it's appropriate to comment on completion rates. MOOC is a term that covers many different models - we need to make sure we don't make the mistake of thinking it is one thing.
I'm not sure your analogy quite stacks up. You could of course use the M1 to go from junction 1 to junction 15 (or London to Northampton, if you like) and your own journey would be complete. However, if you then claimed to have completed the whole motorway, you'd be wrong. Also, I'd say your point on how many people *could* take the course if they wanted to relates more to openness, as opposed to massiveness. For example, continuing your analogy, a tiny C road is as massive as a motorway if it's only measured by how many people could use it if they wanted to. Moocs allow many people to take their own journeys, but it seems very few are completing the courses in the traditional "start to finish" way. Which is fine. I agree with your final point. Moocs take many forms, and are used in many ways.