MOOC design and retention impact factors

It is widely cited that MOOCs have poor retention rates, anywhere down to 5% of those who start the course may complete it to the end (Allione and Stein, 2016). On any other form of online or distance learning course, those figures would be a sign that the course was ineffective and not meeting the educational needs of participants. However, the rules that apply to non-massive online courses don’t appear to be a good fit for MOOCs. The openness of a MOOC may lead to low retention simply down to factors associated with learners preference for flexibility, but the distinctiveness of MOOC learning design could also play a significant part in attrition. Perhaps, it’s not all the learners’ fault.

Course content influences

Hone and El Said (2016) found that ‘perceived effectiveness’ of the course (stemming from course content) had a positive impact on retention, ‘course content’ itself was not shown to be a statistically significant factor. Further, whilst ‘interaction’ was not shown to contribute towards perceived effectiveness, interaction was a factor that influenced retention. Interestingly, as a result of the pattern of responses to their survey, Hone and El Said were not able to determine the influence of course structure. This, they argued, is largely down to uniformity of MOOC platforms and “strong platform-wide organisational framework[s] for the presentation of course material” which take the course “experience” out of the hands of the course designer (Hone and El Said, 2016, p.166). This raises the question then about the influence not only of course content, but how much course structure as a common element of MOOCs impacts on retention. Both are involved in the designed learning experience, so to exclude one seems erroneous.

Even though the structure of courses may appear similar, typically with a faux weekly timetable, the nuance is in the type of activity that learners are expected to undertake in order to achieve specific learning objectives. The sequence of tasks which build up a learning activity may lead to different learning experiences if the order of those tasks is altered. For example, an activity that requires observation of domain practice, reflection and then reveals a theoretical construct will lead a learner to have a different interpretation of practice at the point of observation, compared to a sequence that begins with theoretical underpinning. Structure of courses should be viewed therefore not just by the number of weeks and expected hours of study, but the requirements of activities within a week, including how ‘synchronous’ those activities are expected to be. This itself introduces the factor of competing priorities and how course design enables, or not, a flexible approach to study which is favoured by distance learners. In particular for continuing professional development (CPD), where learners are fitting studies around existing work, flexibility is one of the strengths of an online course.

MOOC pedagogy

The structure of a course is also dependent upon its pedagogic angle, whether it is a cMOOC (connectivist, collaborative, constructivist, learner-centred) or xMOOC (content delivery, independent, educator-centred) (Seimens, 2012). The influence of pedagogy is perhaps captured through a measure of “perceived effectiveness” (Hone and El Said, 2016), but also requires factoring in preconceived ideas learners might hold about online courses and whether their expectations are met or challenged (positively or negatively). If a learner joins a cMOOC expecting it to be self-paced, require minimal interaction and largely independent learning (i.e. expecting traits of xMOOCs), it is not too much of a stretch to argue that the activity design of a cMOOC requiring contribution, awareness of other learners and knowledge building (rather than receiving) would lead to dropout.

Similarly, informed by pedagogical approach, the interface of a MOOC platform may meet, or not, learners expectations or preferences for online interaction. FutureLearn, for example, promotes loosely-structured commenting across all course content. Interactions are possible at any point in the course content, though may be prioritised and shaped by instructional design from course authors. Constructivist and collaborative behaviours are part of the platform. Allione and Stein (2016) suggested that familiarity with platform has a positive impact on learner retention, “possibly… tailor[ing] their expectations of the course” as a result. Familiarity with the platform, by extension is linked to familiarity of the learning experience. Factoring in expectation management when learners sign up to courses, retention appears to become related to the marketing of the course itself and how much it promises a particular type of learning.

Recruiting the ‘right’ learners

Whilst large enrolment numbers may impress course sponsors, recruiting learners that are not aligned to the learning design (let alone the content) of the MOOC invariably impacts the retention figure. DeBoer et al (2014, p.77) noted that MOOC participants have “no monetary cost to enter and no penalty for leaving”, leading to the whole course run acting as a “shopping period” with no commitment to complete a purchase. This ‘try before you buy’ approach is, strangely enough, how a large number of higher educational institutions perceive MOOCs: fulfilling a marketing role to showcase flagship subjects and leading professors to entice students to join a degree course. However, in online CPD, the purpose of a MOOC requires active engagement throughout the whole course in order to build an individual learner’s capacity for change in their practice. The course content, structure and interactions are designed to meet specific learning outcomes through activity, resulting in measurable professional development. With traditional retention measures, there is an assumption that all those who started the course are intending to complete the course, yet that ignores the ‘openness’ factor of MOOCs. DeBoer et al (2014, p.77) suggested that “MOOC registrants will vary in preparation and goals far more than in traditional residential courses” and as a result, there will be a proportion of learners who never intended to complete.

“Part of the issue here stems from the fact that those who initially sign-up for a MOOC may do so without intending to participate, so including them in the statistics for calculating retention can be seen as misleading.”

Hone and El Said (2016, p.159)

Where a MOOC is aiming to have an impact on a professional sector, the title, language of description and activity design should play to that sector’s vocabulary and knowledge base. One of the fundamentals of MOOC design is to identifying the audience very clearly beforehand, to recruit the learners who intend to complete the course for a personal or professional motivation. Then, to design the learning activities that take advantage of that professional context. There is a tension forming though between this approach to narrowing down the pool of learners and the ethos of ‘openness’ and few prerequisites that MOOCs strive for. However, for online CPD the identification of a target audience and designing learning activities that exploit and build upon prior knowledge through collaboration, sharing and reflection, could be more significant to perceived effectiveness of the course for non-window shopping learners than activities that invite more generic contributions from non-experts.

Professional communities

There appears to be scope to explore the effect of professional communities on retention in online CPD MOOCs. If learners come from similar contexts, through targeted recruitment for example or integration with other programmes, would they buy-into a common way of learning together online in order to achieve similar professional development goals? In doing so, the cohesiveness of the learning group may positively affect the perceived effectiveness of the course, and in turn (drawing upon Hone and El Said, 2016) sustain engagement throughout the whole course.

One observation is that the first week leads to the largest drop off of learners (Ho et al., 2014 indicated on average 50% of learners from courses they studied leave after the first week). Whilst this may support the concept of learners who are non-committed from the outset as the likely cause of attrition rates, there is also the consideration that the activities of the first week must meet the expectations of intended committed learners. If they don’t, even those who the course has been designed specifically for may decide to leave early. In traditional online learning course design, induction and community building is seen as a key component in first-week activities. However, this may need to be reconceptualised for MOOC design.

MOOCs need different design principles

Salmon (2000) proposed a five-step model for induction and socialisation of learners that has been widely cited and continues to be designed into online courses that draw on social constructivism pedagogy. The five-step approach aims to equip learners with the appropriate technical, communication, social and learning skills for the online environment, with facilitation by educators and by participating in tasks. Yet, with MOOCs, technical considerations are largely overlooked with an assumption those who know about MOOCs are savvy enough to work the platform and similarly the principles of online communication are now, rightly or wrongly, seen as second nature. Individuals will join at any point and the traditional socialisation step of cohort induction isn’t as defined within a MOOC space, even if a dedicated ‘welcome, introduce yourself’ activity is included. For learners who join at the end of the first week, other learners may well have moved on to the next thing in the course or even subsequent week’s content. The socialisation occurs at a massive level, significantly different from the intentions of Salmon’s scaffolded and incremental approach. Another consideration, if we refer back to previous points about meeting expectations and judgements of effectiveness of courses, is that induction-type activities are not course content relating to the subject of interest to potential committed learners. Starting courses with housekeeping rather than ‘interesting’ or ‘relevant to practice’ content may also be a source of attrition, possibly more so than any impact caused by lack of socialising learners.


There are many factors that influence retention on MOOCs, but it appears to me that course design is a significant aspect, right from the first click on the course. As more surveys and papers emerge showing case studies of MOOC participation, differences of course pedagogy and intended purpose must be factored in. Retention for online CPD may well differ to retention in lifelong learning MOOCs or ‘marketing loss-leaders’ for higher educational institutions. MOOCs, as a branch of online learning, are still finding their pedagogical feet as they borrow from traditional online course approaches and supporting design theory. I am hopeful that with a little more experimentation, design models will be developed that sustain learner engagement throughout courses to help more participants reach final learning outcomes.


Allione, G. and Stein, R. M. (2016). Mass attrition: an analysis of drop out from principles of microeconomics MOOC, The Journal of Economic Education, 47(2), 174-186.

DeBoer, J., Ho, A. D., Stump, G. S. and Breslow, L. (2014). ‘Changing “Course”: Reconceptualizing Educational Variables for Massive Open Online Courses”, Educational Researcher, 42(2), 74-84.

Ho, A. D., Reich, J., Nesterko, S., Seaton, D. T., Mullaney, T., Waldo, J., & Chuang, I. (2014). HarvardX and MITx: The first year of open online courses (HarvardX and MITx Working Paper No. 1).

Hone, K. S. and El Said, G. R. (2016). ‘Exploring the factors affecting MOOC retention: a survey study’, Computers & Education, 98, 157-168.

Salmon, G. (2000). E‐moderating: The key to teaching and learning online. London: Kogan Page. The Five Stage Model is also available online at

Seimens, G. (2012). ‘MOOCs are really a platform’, Elearnspace.





One response to “MOOC design and retention impact factors”

  1. […] being scaffolded by the educator before knowledge construction and more collaborative interaction. I have previously commented on the limitations of such a model in open online learning that is informal, fluid and more […]

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.