Emerging patterns in MOOCs: Learners, course designs and directions By Hamish Macleod, Jeff Haywood, Amy Woodgate, University of Edinburgh, Mubarak Alkhatnai, King Saud University ©Association for Educational Communications and Technology 2015
Abstract Engagement with Massive Open Online Courses (MOOCs) at the University of Edinburgh has emerged from its strategic priorities to explore and innovate in the area of online and technologically supported approaches to teaching and learning. This paper provides an account of analysis aimed at understanding who Edinburgh MOOC learners are, who elects to participate and the aspirations of that population, and the place that the MOOC will occupy in the University’s online learning ecology. The analysis addresses a number of predictions that have been made about MOOCs since 2012, including their use for providing educational opportunities to the disadvantaged; global uptake of online learning; growth of an ‘educational imperialism’; and the claim that ‘MOOCs are for male geeks’, and concludes with some observations about the University of Edinburgh’s future plans in this space. Keywords: digital education, higher education, MOOCs, widening access
Introduction Technology in education has been high on our agenda at the University of Edinburgh for more than twenty years. We have been early adopters; beginning in the early 90’s with emailfor-all; university-wide VLEs in the late 90’s; online distance education in the mid 2000’s, and most recently MOOCs in 2012 (MacLeod, Haywood, Haywood & Anderson, 2002; MOOC data report #1). Working with others, including our graduate students, we have extended our 56
scope to comparative studies (Haywood, Haywood, MacLeod & Tenhonen, 2003; Hu, 2007; Alkhatnai, 2013). In this article, we shall present some of our most recent findings from our Massive Open Online Courses (MOOCs), and reflect on the insights that they offer into this rapidly changing area of digital higher education. We shall close with some predictions for the next few years. Our reasons for offering MOOCs are described in detail in our recent MOOC Report #1 (MOOCs @ Edinburgh Group, 2013) and they will not be re-stated in full here, but briefly they include: outreach; research for our on-campus blended courses and our expanding range of taught online degree programmes; and injecting fun into teaching with a sandpit for faculty to try out and debate new ideas in online learning. In 2012, when the hype about MOOCs was most intense, many assumptions and predictions were offered about them and what they might achieve. Alongside the bold ‘end of higher education as we know it’ predictions (Kolowich, 2013), were some more testable statements, including outreach of digital education to the disadvantaged (Coughlan, 2014); global uptake of online learning (Martin & Walter, 2013); growth of an ‘educational imperialism’ (MacGregor, 2013); ‘MOOCs are for male geeks’ (Straumsheim, 2013), and ‘a transformation of traditional ways of teaching and learning’ (Ebben & Murphy, 2014). We, and a small number of other education researchers, began to gather data to test these predictions. At the outset it was clear to us that designing a MOOC is quite different to designing a
TechTrends • January/February 2015
Volume 59, Number 1
standard university course; it is a case of ‘designing for the unknown learner’ (MacLeod, Haywood, Woodgate & Sinclair, 2014). Clearly, however, when one designs any course, one has to have some learner cohort in mind, and so five of our first six MOOCs were designed at entry to university Bachelor level and one was at Master level. When we created our first MOOCs in 2012, little was known about those enrolling, and so a first question we asked ourselves as we offered our first MOOCs was: who are the tens of thousands of individuals who sign up to learn on short, free, online courses that offer no qualification or credits, and what are they hoping to achieve? An immediate second question was: as these are the first MOOCs to exist, are they attracting an ‘unusual audience’, and if so, will a stable audience arise and if so, when? To answer these questions we have surveyed our MOOC participants at the start and end of every MOOC with a standardised set of questions. To date we have at least two offerings of each of our first 6 MOOCs, and some patterns are emerging from the approximately 150k responses we have gathered from the 600k people who have enrolled. The data and analysis from the first offerings of these six MOOCs can be found at https://www.era.lib.ed.ac.uk/handle/1842/6683.
enrolments has been broadly stable between the two iterations of the six MOOCs (around 12 months apart) but some interesting ‘spikes’ are visible that may indicate a trend (Figure 1). Enrolments from People’s Republic of China (PRC) rose sharply, most likely due to the release of a Coursera server behind the PRC firewall, as did those from India, where we know publicity for MOOCs has been strong, especially amongst Indian students. We might expect these changes to accelerate, particularly as crowd-sourced translation of particular MOOCs into local languages continues, and non-English platforms emerge. This will result in an even more international MOOC learner population than at present. The data that we will present here are drawn from anonymous surveys of our MOOC participants (response rates approx. 20% on average). The international distribution of learners has however been independently verified by an analysis of data gathered from the Coursera MOOC platform, showing essentially the same pattern of international participation based on the IP addresses from which the participants come. This would seem to suggest that our survey respondents are representative of the wider population of MOOC participants.
MOOC learners are worldwide
MOOCs and widening access to education
Our learners are indeed global, and are very widely distributed across the world, in c200 countries, and although the US is the largest recruitment source on average (originally 26%, although declining and now 20%), it isn’t the majority source, even though Coursera, the MOOC platform, is a US company; and nor is the UK, even though Edinburgh is a UK university. The geographical pattern of
Despite the initial rhetoric that MOOCs would offer universal access to higher education courses for the disadvantaged, this has not been evident in the data. The great majority of learners are well-educated (c70% with a first or second degree) and in employment. This has changed little between offerings of our
Figure 1. Main countries of residence of MOOC learners on 6 Edinburgh MOOCs in the 1st and 2nd iterations
Volume 59, Number 1
TechTrends • January/February 2015
57
MOOCs, and has been seen in the majority of MOOCs from other universities. However, these averages mask some differences between MOOCs. For example, in its second iteration, Equine Nutrition had far fewer learners with a first (24%) or second degree (20%), and more at college (24%) and post-school training (11%) than did Education & Digital Cultures, where 60% of learners had achieved post-graduate level (Figure 2). The latter is perhaps not surprising in retrospect, as it was very innovative and designed around digital education theory and practice. The patterns of prior educational experience were very stable between offerings of each MOOC, perhaps suggesting segmentation in publicity routes as much as the suitability of particular MOOCs for specific groups of learners.
MOOCs for career enhancement The averages of prior educational experiences across all learners on each MOOC may also mask variations in need for study. Assuming that educated learners are only learning ‘for fun’ may be mainly true, but for some learners there may be more at stake. For example, learners in developing countries, or countries in particular economic difficulty (eg Spain, Greece), have more interest in using the MOOC to gain a certificate and/or using it to enhance their careers. Not surprisingly we see that the desire to obtain certification for study, or participating with a view to improve career prospects, were reasons offered increasingly less often as participants aged (Figure 3). Strikingly, and rather reassuringly, the primary reason for taking a MOOC, and one that remained remarkably stable as a reason offered
right across the age range and in all our MOOCs, was the desire to “learn new things”. Despite the high education levels of learners, these learners are mostly taking MOOCs in a subject area other than that of their original studies (70% or more do this; Figure 4). An exception was AI Planning, where almost 60% were educated in computer sciences, which is perhaps not surprising as this MOOC is offered at postgraduate level. It also attracted more participants from India (10.5%) and Russia (3.8%). Young learners are also a rising proportion of the MOOC learner population; under 18’s rose 50% between the first and second offerings, albeit they are still only ~5% of learners on average (Figure 5 on the following page). This increase might be due to school teachers proposing MOOCs to their pupils, independently or with the support of their local university. We promoted our Introduction to Philosophy MOOC to local schools and we note that they have the largest recruitment from this age group. If this trend continued, young learners would become an important part of the MOOC cohorts, and schools and universities will have increasing numbers of students who have learned to learn online to some extent, demonstrating increased educational digital literacy.
MOOCs for male geeks? The first MOOCs to catch the media and public attention in the autumn of 2011, socalled xMOOCs, were in very technical subjects (Artificial Intelligence, Machine Learning, Databases) and they attracted a predominantly male audience, leading to some suggestions that MOOCs would continue to attract a male,
Figure 2. Highest educational level attained by learners on 6 Edinburgh MOOCs, 2013-14
58
TechTrends • January/February 2015
Volume 59, Number 1
Figure 3. Reasons Edinburgh MOOC learners offered as to why they studied, by age band
‘nerdy’, population (Straumsheim, 2013). This has failed to materialise, and we have found that gender participation rates have been largely a function of the subject matter that the MOOCs address, with AI Planning having 15% females and Equine Nutrition 90% females. Overall, Edinburgh’s first MOOCs have recruited approximately the same number of men as women. Clearly, as in degree programmes in higher education, subject strongly influences gender proportions in enrolments (Jordan, 2013). This has been consistent between iterations of the same MOOCs.
Serial MOOC learners One feature of our MOOC learners that changed dramatically from the first to second
offerings was our learners’ prior experience of study on MOOCs, where we noted a rise from 22% to 58% in learners who had previously taken a MOOC. Repeat enrolments might be due either to taking a variety of MOOCs from across platforms or universities, or retaking some MOOCs, and there is evidence of the latter, at least on a small scale. The latter is a feature that is not really encouraged or enabled in conventional higher education courses, so this finding is potentially important. Conventional “for credit” academic courses generally set a limit on the number of times that a student may take a course, the implication being that if an assessment has been failed for the second, or third, time, it is unlikely that the student will ever be able to master the course content. The
Figure 4. Relationship between subject of MOOC studied and previous academic study
Volume 59, Number 1
TechTrends • January/February 2015
59
Figure 5. Ages of Edinburgh MOOC learners on 1st and 2nd iterations
idea that one might continue to take a course or test until one reaches a desired level is not new, and some theorists, notably Benjamin Bloom (Bloom, 1974) have proposed the model of “Mastery Learning”, suggesting that all learners can potentially reach a criterion level of understanding and skills given sufficient time and support.
MOOC platform variations The data we have explored to this point have all been gathered from MOOCs we have offered on the US-located Coursera MOOC platform, so a legitimate question is whether these are universal features of MOOCs and their learners, or whether they are specific to the University of Edinburgh MOOCs on the Coursera platform. One way to begin to answer this question is to compare these data with those from our Edinburgh MOOCs on the UK-located Futurelearn platform, which we joined in July 2013. Futurelearn staff have analysed the entry survey data from 37 MOOC offerings (first and subsequent pooled) and made them available to partners. Summarising the similarities to and differences from our Coursera MOOCs, we observe that the gender, age, education and employment profiles are similar on both platforms, although with a tendency for Futurelearn to have more learners over 45 years of age and less gender bias towards male learners in science and technology MOOCs. Futurelearn didn’t allow respondents to offer multiple reasons for study, but the highest single response was ‘to learn new things’ (25%), with two questions related to trying online learning adding up to a further 25%. Careerrelated reasons given for study were similar for Futurelearn and our Coursera MOOC 60
learners, with 16% giving this response. The big difference between the two platforms was the country of residence: for our Coursera MOOCs, in their second iterations, this was the US (21%) followed the UK (9%), whereas for Futurelearn MOOCs the UK was very prominent (67%), followed by a much smaller fraction from the US (5%), and with Australia in third position (2.1%). We deduce that the effect of differences in publicity and partnerships is very strong.
MOOCs beyond English Another approach to understanding the generalizability of data from Edinburgh’s MOOC learners is to examine data from MOOCs from other universities, and ideally in situations well-distanced from our own. If the charge of ‘educational imperialism’ of the MOOC concept in general is to be countered (Haggard, 2013), these developments will be in the forefront. A MOOC platform recently established in Saudi Arabia offers some insights into the nature of MOOC learning in language other than English. Rwaq [http://www.rwaq. org/] is a MOOC platform that was established and launched by two local Saudi entrepreneurs to offer courses exclusively in Arabic. It is the first platform in the Arab world and has attracted attention in the last few months. Very recently another portal, Edraak [https://edraak. org/], was introduced by the Queen Rania Foundation for Education and Development in Jordan. While Edraak can be categorized as the Arabized version of the open source edX portal, Rwaq took a different approach, and rather than employing open-source platforms, the Rwaq initiators built the platform from scratch. They said that this approach helped them to create something particularly tailored
TechTrends • January/February 2015
Volume 59, Number 1
Figure 6. Countries of residence of learners on the Rwaq MOOC platform
to their targeted audience. They claim it is the only fully-native and fully-Arabic learning hub for Arabic-speakers worldwide, with content from Arab professors across the Arab World (Curley, 2013). The analytics of this platform reveal some interesting trends (data received via personal correspondence, 2014). Rwaq presents a variety of courses from medicine, engineering, and sciences to social sciences, religion, and arts. Currently there are 35 courses on the portal and so far, more than 100,000 participants have enrolled on them, with the largest number from Saudi Arabia (35.01%) and Egypt (20.08%) as shown in Figure 6 below. However, and in relation to the targeted language, one notices a reasonable percentage of participants from non-Arabic speaking countries like the USA (3.17%), which is probably due to Arab speakers residing in these countries and also the increasing popularity of Arabic language over the internet. Some reports indicate a 2500% increase of the use of Arabic over the internet in the last ten years or so, although Arabs still only represent 3.8% of all internet users (Internet World Stats, 2012). In common with the aspirations of other MOOC platforms, Rwaq aims to provide learners, and especially those in deprived areas, with opportunities of learn and advance their existing knowledge. Illiteracy rates within the Middle East are high, especially in low-income countries where tertiary education is rare, and the current education systems are old and failing. MOOCs, and other free educational resources, presented from qualified people and institutions are seen as rare opportunities that Volume 59, Number 1
should be invested in and used. According to Al Farhan, one of Rwaq’s initiators, “it is our hope that MOOCs will reform current educational systems in the area and set high standards for higher education” (Curley, 2013). It is reported that 46.5% of women in the region are illiterate, and women in education in Saudi Arabia are segregated from male students and teachers. Rwaq has addressed this by offering, for example, courses about visual arts and business branding and e-commerce presented by Saudi female professors. Unfortunately, Rwaq did not collect data from its learners about their gender, which would have given some insight into male and female participation. Rwaq like other MOOC platforms has challenges to overcome: developing a sustainable business model, accreditation of learners’ achievements, and increasing the number of learners who complete the MOOCs. Rwaq is one amongst a small but expanding number of non-English MOOC platforms that are appearing around the world. Although it is conceivable that the early US-based platforms will retain a degree of ‘domination’ of open online education, especially as they ramp up their non-English language capabilities, the presence of these new regional platforms, some with very large potential audiences, will ensure that at least some counterpoise to the charge of educational imperialism. We in the University of Edinburgh will seek to collaborate with such platforms, at least in sharing research findings, and perhaps, where appropriate, also developing joint MOOCs with partner universities local to the platforms.
TechTrends • January/February 2015
61
Future trends and possibilities How do these data help us decide on a MOOC strategy to help us plan for the future? In many respects the MOOC learners are quite like those who come to our campus for our open studies (continuing education) short courses. Our 17,000 registered open studies learners tend to be older and well educated, mostly study for interest, pay very low fees and do not seek formal university credits. Thus MOOCs could be seen as a global extension of this existing outreach activity, and we could continue them in that way given a modest financial return from paid-for certificates. However, we also have a strong interest in reaching less advantaged individuals, and especially younger ones, to encourage them into higher education (eg our Football MOOC). We would like to help provide learning opportunities to developing countries and to those facing serious challenges in educating their citizens, and to support people of all kinds to enhance their careers and lives. Clearly, our data show that individuals with these characteristics are not the most prominent participants in our MOOCs, although their numbers are not insignificant given the large numbers in total enrolled on MOOCs. So what should we do? One strategy is to partner with others to reach potential learners we otherwise would probably not reach, and another is to design MOOCs likely to attract the attention of target groups. In the former category we have worked on pilots including a project using the Critical Thinking in Global Challenges MOOC in a supported and tutored way with secondary school pupils in Rwanda (Bartholet, 2013). Colleagues in our Widening Participation Office at Edinburgh have been promoting the use of the Introduction to Philosophy MOOC among teachers of pupils in schools without any tradition of sending their pupils to study subjects like philosophy. The pupils can engage with the MOOC resources, supported by their own teachers, and thus gain a small taste of what studying this subject in a higher education setting might entail. In the latter category (MOOCs designed for particular audiences), we have developed the first of our MOOCs based around sports (in this case, football) and jointly with colleagues in Uruguay are creating MOOCs on app development and computer coding designed to appeal to teenagers. The University of Edinburgh has strong existing relationships with professional bodies which accredited their members (eg engineering, law, medicine) and 62
so we will increase our partnership with these bodies to develop MOOCs for career entry for those wishing to enter the professions and career enhancement for those within them. This parallels the joint online degree programmes we already offer with some of these bodies. We expect our joint working and partnership to grow, both in the UK and worldwide.
Reflections on the impact of MOOCs within Edinburgh Finally, let us return to the topics with which we began, namely our reasons for offering MOOCs, which include injecting some fun into teaching, a sandpit for educational research, and outreach to new audiences for our education. What conclusions can we draw and what actions follow from them? MOOCs did indeed inject a very strong sense of fun into teaching, and they stimulated a vigorous debate in many academic departments about online learning, with some subsequently coming forward to propose taught online degree programmes. Beyond online learning, discussions about pedagogy more generally have been evident amongst the subject areas offering MOOCs. We have seen very little of the negativity that some universities have experienced (Kolowich, 2014; Straumsheim, 2014). In general within the University of Edinburgh, we have gained greater confidence in our abilities to design and deliver large online courses, and to support faculty in those activities. Lessons have been learned about short video lecture sequences structured activities how online course design approaches can be formalised and can be faculty-led with subject and discipline areas. We are increasingly designing MOOCs for outreach, rather than just making MOOCs that individual faculty might like to provide, and also promoting them in schools through teachers, and using our students as volunteers to support the local community to access this form of education. We are exploring with Coursera how to ‘float our MOOCs free’ so that they are can be used independently of our faculty by those who wish to take a course alone, or in self-formed groups, or with a local tutor. How well this will work is not clear, but experiments of this kind will help us to think about how to reach potential learners at the time they wish to learn rather than when we wish to teach them. MOOCs are not just to teach, but also to engage: in some subjects (eg. ecology, astronomy, behavioural economics) the option to engage thousands of learners in research is a powerful incentive to design courses where the data are
TechTrends • January/February 2015
Volume 59, Number 1
gathered by the learners, and analysed and reflected on by them. MOOCs sit very well with the spirit of citizen science, as seen for example in Galaxy Zoo [http://www.galaxyzoo.org/]. In the same vein, we have begun to reflect on what the ‘massive’ in MOOCs means, partly because numbers are falling from the giddy heights of the early days, but also because the M can also mean ‘multitude’ to reflect the complexity of the audiences that are attracted to them, with their varying reasons for study (Knox, 2013; Knox, 2014). Offering MOOCs has been a valuable experience, and continues to stimulate discussion within our academic community, and as MOOCs evolve we expect this to continue. We do not yet know exactly how many MOOCs we shall eventually offer, nor how long some of them will continue to be offered before they reach end-of-life, but we do expect to offer open online education of some kind into the foreseeable future. Address correspondence regarding this article to Hamish Macleod; Digital Education; University of Edinburgh; St John’s Land; Holyrood Road; Edinburgh EH8 8AQ United Kingdom; email:
[email protected]; phone: +44(0)131 651 6665
References Alkhatnai, M. (2013) Strategic use of ICT in the Saudi system of higher education: King Saud University. Ph.D Thesis. The University of Edinburgh. Retrieved from http://hdl.handle.net/1842/8869 Bartholet, J. (2013) Free Online Courses Bring “Magic” to Rwanda. Retrieved from http://www.scientificamerican. com/article/free-online-classes-bring-magic-rwanda/ Bloom, B. S. (1974) Time and learning. American Psychologist, 29, 682-688. Haywood, D., Haywood, J. and MacLeod, H. (2006) Graduate e-Literacies and Employability in Literacies for Learning in the Digital Age, Ed. Martin, A. & Madigan, D., London: Facet. Coughlan, S. (2014) Online Mooc courses deliver Ebola health advice. Retrieved from http://www.bbc.co.uk/ news/education-29521360 Curley, N. (2013) Saudi Arabia’s Rwaq builds an online courseware platform for the Middle East. Retrieved from http://www.wamda.com/2013/12/saudi-arabia-rwaqonline-courseware-mooc-middle-east Ebben, M., and Murphy, J. (2014) Unpacking MOOC scholarly discourse: a review of nascent MOOC scholarship. Learning Media and Technology, 39(3), 328-345. Haggard, S. (2013) The Maturing of the MOOC: Literature Review of Massive Open Online Courses and Other Forms of Online Learning. Department for Business In-
Volume 59, Number 1
novation & Skills. BIS Research Paper (130). Retrieved from https://www.gov.uk/government/uploads/system/ uploads/attachment_data/file/240193/13-1173-maturing-of-the-mooc.pdf Haywood, D., Haywood, J., MacLeod, H. and Tenhonen, P. (2003) Surveys of European Universities Skills for ICT for Staff and Students (SEUSSIS Project Report). The University of Edinburgh. Retrieved from www.intermedia.uib. no/seusiss/results.html Hu, Y. (2007) Chinese Learners and Computer Assisted Language Learning, PhD Thesis, University of Edinburgh Internet World Stats (2012) Arabic Speaking Internet Users Statistics. Retrieved on 23 March 2012 from http:// www.internetworldstats.com/stats19.htm Jordan, K. (2013) A quick look at gender differences in MOOCs. Retrieved from http://moocmoocher.wordpress.com/2013/08/06/a-quick-look-at-gender-differences-in-moocs/ Knox, J. (2013) eLearning and Digital Cultures: a multitudinous open online course. Retrieved from http://elearnmag.acm.org/archive.cfm?aid=2525967 Knox, J. (2014) Digital culture clash: “massive” education in the E-Learning and Digital Cultures MOOC. Distance Education, 35(2), 164-177. Kolowich, S. (2013) Obama Is Advised to let Market Forces Decide Fate of MOOCs. Retrieved from http://chronicle.com/blogs/wiredcampus/presidential-advisersbrief-obama-on-moocs-and-offer-advice/49159 Kolowich, S. (2014) Doubts About MOOCs Continue to Rise, Survey Finds. Retrieved from http://chronicle.com/ article/Doubts-About-MOOCs-Continue-to/144007/ MacLeod, H., Haywood, D., Haywood, J. and Anderson, C. (2002) Gender and Information and Communications Technology - a 10-year study of new undergraduates. Techtrends, 46(6), 11-15. MacLeod, H., Haywood, J., Woodgate, A. and Sinclair, C. (2014) Designing for the Unknown Learner. Under review, retrieved from European MOOCs Stakeholders Summit papers, 245-248 http://emoocs2014.eu/sites/ default/files/Proceedings-Moocs-Sumit-2014.pdf MacGregor, K. (2013) MOOCs make waves in high education worldwide. Retrieved from http://www.universityworldnews.com/article.php?story=20130920142318192 Martin, L., and Walter, B. (2013) Setting an online example in educating women. Retrieved from http://articles.latimes.com/2013/jan/25/opinion/la-oe-martin-onlinegender-equality-20130125 MOOCs @ Edinburgh Group (2013) MOOCs @ Edinburgh 2013 – Report #1. Retrieved from http://hdl. handle.net/1842/6683 Straumsheim, C. (2013) Masculine Open Online Courses. Retrieved from https://www.insidehighered.com/ news/2013/09/03/more-female-professors-experimentmoocs-men-still-dominate Straumsheim, C. (2014) Rhetoric Check. Retrieved from https://www.insidehighered.com/news/2014/05/14/faculty-group-continues-anti-mooc-offensive
TechTrends • January/February 2015
63
