computer or phone providers. With such a broad range of learners, MOOCs find themselves up against a tough comparison with formal education. To use Weinberger's (2007) phrase, higher education 'filters on the way in', whereas MOOCs 'filter on the way out'. The quality measures are therefore very different. Student satisfaction rates for a system that has completely open enrolment and filters on the way out are unlikely to compare favourably with a very different system where there has been a filtering already. Filtering on the way out and operating in the open does, however, allow for new types of quality measures. These could be altmetrics-type measures (what kind of 'buzz' does it create, what is the public reaction of participants) or analytics (how many people come back, what is the dwell time, bounce rate, etc.). But the comparisons should be with other MOOCs, not with formal education. Quality, and what is measured, is therefore just one example of established practices that the attention on MOOCs should make us reconsider.
MOOCs and Cost
A second issue that MOOCs raise for formal education is that they force an examination of the costs associated with teaching. Estimates of how much it takes to produce a MOOC vary, with Udacity budgeting US$200,000, EdX US$250,000 (DeJong 2013) and University of North Carolina estimating US$150,000 for their Coursera MOOC (Goldstein 2013). Once created, the idea is that they can be run at next to no cost, although this will depend on how closely involved the lead academic is in each presentation. Clearly if you are not charging fees for people to study on a course, then its presentation costs need to be low if it is to be a sustainable model. As we saw with OERs, there are different models of
