Completion rates is something that MOOC watchers on either side of the fence have been obsessed with from day one. In this blog post, Martin Weller publishes some data collected by OU researcher Katy Jordan on completion rates, in particular factors that influence them. Although a full paper is promised in due time, the preliminary results are already quite interesting. To mention a few, the median completion percentage is around 12%; the larger the enrollment (or the larger the number of active users), the lower the completion rate; the longer a course takes to complete, the lower the completion rate. Martin Weller then raises a few questions, tongue in cheek as the data are correlational only. However, it would indeed be nice to know if shorter MOOCs fare better in terms of completion, or if people get better at learning with MOOCs over time.
Via Peter B. Sloep
And then there is of course the question of whether completion matters. Martin raises that question too. The fundamental mistake is to compare a MOOC setting with a classroom setting, opening the gates to anybody even with only a remote interest versus a handpicked, in many respects homogeneous group of students who paid good money. Putting the comparison this way points out that the really interesting thing would be _not_ to find significantly lower completion rates in MOOCs.
The second mistake relates to an assumption that underlies many MOOC discussions: MOOCs are cheap, if they are as good as ‘regular’ courses, then they should replace such courses. That argument of course fails if completion rates are abysmal. The mistake here is that costs aren’t the only dimension in which MOOCs and regular courses are to be compared. Costs matter, but effectiveness (do people learn useful things, do they learn what they are supposed to) matters at least as much. I would suggest to contextualise discussions about completion rates by looking at MOOCs as a means to some end. That implies that MOOCs for, say, continuous professional development need to be evaluated differently then MOOCs for Psychology 101. @pbsloep