They're big: but will they survive? Image: © Wikipedia
They’re big: but will they survive? Image: © Wikipedia

The story so far

For my open textbook Teaching in a Digital Age, I am writing a chapter on different design models for teaching and learning. I have started writing the section on MOOCs, and in my previous post, ‘What is a MOOC?‘, I gave a brief history and described the key common characteristics of all MOOCs.

In this post I examine the differences in philosophy and practice between xMOOCs and cMOOCs.

Design models for MOOCs

MOOCs are a relatively new phenomenon and as a result are still evolving, particularly in terms of their design. However the early MOOC courses had relatively identifiable designs which still permeate most MOOCs. At the same time, there are two quite different philosophical positions underpinning xMOOCs and cMOOCs, so we need to look at each design model separately.


I am starting with xMOOCs because at the time of writing they are by far the most common MOOC. Because instructors have considerable flexibility in the design of the course, there is considerable variation in the details, but in general xMOOCs have the following common design features:

  • specially designed platform software: xMOOCs use specially designed platform software that allows for the registration of very large numbers of participants, provides facilities for the storing and streaming on demand of digital materials, and automates assessment procedures and student performance tracking.
  • video lectures: xMOOCs use the standard lecture mode, but delivered online by participants downloading on demand recorded video lectures. These video lectures are normally available on a weekly basis over a period of 10-13 weeks. Initially these were often 50 minute lectures, but as a result of experience some xMOOCs now are using shorter recordings (sometimes down to 15 minutes in length) and thus there may be more video segments. Over time, xMOOC courses, as well as the videos, are becoming shorter in length, some now lasting only five weeks. Various video production methods have been used, including lecture capture (recording face-to-face on-campus lectures, then storing them and streaming them on demand), full studio production, or desk-top recording by the instructor on their own.
  • computer-marked assignments: students complete an online test and receive immediate computerised feedback. These tests are usually offered throughout the course, and may be used just for participant feedback. Alternatively the tests may be used for determining the award of a certificate. Another option is for an end of course grade or certificate based solely on an end-of-course online test. Most xMOOC assignments are based on multiple-choice, computer-marked questions, but some MOOCs have also used text or formula boxes for participants to enter answers, such as coding in a computer science course, or mathematical formulae, and in one or two cases, short text answers, but in all cases these are computer-marked.
  • peer assessment: some xMOOCs have experimented with assigning students randomly to small groups for peer assessment, especially for more open-ended or more evaluative assignment questions. This has often proved problematic though because of wide variations in expertise between the different members of a group, and because of the different levels of involvement in the course of different participants.
  • supporting materials: sometimes copies of slides, supplementary audio files, urls to other resources, and online articles may be included for downloading by participants.
  • a shared comment/discussion space where participants can post questions, ask for help, or comment on the content of the course.
  • no or very light discussion moderation: the extent to which the discussion or comments are moderated varies probably more than any other feature in xMOOCs, but at its most, moderation is directed at all participants rather than to individuals. Because of the very large numbers participating and commenting, moderation of individual comments by the instructor(s) offering the MOOC is impossible. Some instructors offer no moderation whatsoever, so participants rely on other participants to respond to questions or comments. Some instructors ‘sample’ comments and questions, and post comments in response to these. Some instructors use teaching assistants to comb for or identify common areas of concern shared by a number of participants then the instructor or teaching assistants will respond. However, in most cases, participants moderate each other’s comments or questions.
  • badges or certificates: most xMOOCs award some kind of recognition for successful completion of a course, based on a final computer-marked assessment. However, at the time of writing, MOOC badges or certificates have not been recognised for credit or admission purposes even by the institutions offering a MOOC, or even when the lectures are the same as for on-campus students. No evidence exists to date about employer acceptance of MOOC qualifications.
  • learning analytics: Although to date there has not been a great deal of published information about the use of learning analytics in xMOOCs, the xMOOC platforms have the capacity to collect and analyse ‘big data’ about participants and their performance, enabling, at least in theory, for immediate feedback to instructors about areas where the content or design needs improving and possibly directing automated cues or hints for individuals.

xMOOCs therefore primarily use a teaching model focused on the transmission of information, with high quality content delivery, computer-marked assessment (mainly for student feedback purposes), and automation of all key transactions between participants and the learning platform. There is almost no direct interaction between an individual participant and the instructor responsible for the course.


cMOOCs have a very different educational philosophy from xMOOCs, in that cMOOCs place heavy emphasis on networking and in particular on strong content contributions from the participants themselves.

Key design principles

Downes (2014) has identified four key design principles for cMOOCs:

  • autonomy of the learner: in terms of learners choosing what content or skills they wish to learn, learning is personal, and thus there being no formal curriculum
  • diversity: in terms of the tools used, the range of participants and their knowledge levels, and varied content
  • interactivity: in terms of co-operative learning, communication between participants, resulting in emergent knowledge
  • open-ness: in terms of access, content, activities and assessment

Thus for the proponents of cMOOCs, learning results not from the transmission of information from an expert to novices, as in xMOOCs, but from sharing of knowledge between participants.

From principles to practice

Identifying how these key design features for cMOOCs are turned into practice is somewhat more difficult to pinpoint, because cMOOCs depend on an evolving set of practices. Most cMOOCs to date have in fact made some use of ‘experts’, both in the organization and promotion of the MOOC, and in providing ‘nodes’ of content around which discussion tends to revolve.  In other words, the design practices of cMOOCs are still more a work in progress than those of xMOOCs.

Nevertheless, I see the following as key design practices to date in cMOOCs:

  • use of social media: partly because most cMOOCs are not institutionally based or supported, they do not at present use a shared platform or platforms but are more loosely supported by a range of ‘connected’ tools and media. These may include a simple online registration system, and the use of web conferencing tools such as Blackboard Collaborate or Adobe Connect, streamed video or audio files, blogs, wikis, ‘open’ learning management systems such as Moodle or Canvas, Twitter, LinkedIn or Facebook, all enabling participants to share their contributions. Indeed, as new apps and social media tools develop, they too are likely to be incorporated into cMOOCs. All these tools are connected through web-based hashtags or other web-based linking mechanisms, enabling participants to identify social media contributions from other participants. Downes (2014) is working on a Learning and Performance Support System that could be used to help both participants and cMOOC organisers to communicate more easily across the whole MOOC and to organise their personal learning. Thus the use of loosely linked/connected social media is a key design practice in cMOOCs
  • participant-driven content: in principle, other than a common topic that may be decided by someone wanting to organise a cMOOC, content is decided upon and contributed by the participants themselves, in this sense very much like any other community of practice. In practice though cMOOC organisers (who themselves tend to have some expertise in the topic of the cMOOC) are likely to invite potential participants who have expertise or are known already to have a well articulated approach to a topic to make contributions around which participants can discuss and debate. Other participants choose their own ways to contribute or communicate, the most common being through blog posts, tweets, or comments on other participants’ blog posts, although some cMOOCs use wikis or open source online discussion forums. The key design practice with regard to content is that all participants contribute to and share content.
  • distributed communication: this is probably the most difficult design practice to understand for those not familiar with cMOOCs – and even for those who have participated. With participants numbering in the hundreds or even thousands, each contributing individually through a variety of social media, there are a myriad different inter-connections between participants that are impossible to track (in total) for any single participant. This results in many sub-conversations, more commonly at a binary level of two people communicating with each other than an integrated group discussion, although all conversations are ‘open’ and all other participants are able to contribute to a conversation if they know it exists. The key design practice then with regard to communication is a self-organising network with many sub-components.
  • assessment: there is no formal assessment, although participants may seek feedback from other, more knowledgeable participants, on an informal basis. Basically participants decide for themselves whether what they have learned is appropriate to them.

cMOOCs therefore primarily use a networked approach to learning based on autonomous learners connecting with each other across open and connected social media and sharing knowledge through their own personal contributions. There is no pre-set curriculum and no formal teacher-student relationship, either for delivery of content or for learner support. Participants learn from the contributions of others, from the meta-level knowledge generated through the community, and from self-reflection on their own contributions.

This is very much a personal interpretation of how cMOOCs work in practice, based largely on my own experience as a participant, but much more has been written and spoken about the philosophy of cMOOCs, and much less about the implementation of that philosophy, presumably because cMOOC proponents want to leave it open to practitioners to decide how best to put that philosophy into practice.

What is clear though is that Downes was correct in clearly distinguishing cMOOCs from xMOOCs – they are very different beasts.

Coming next to a web page near you

Now for the fun part. Over the next few days I will be writing about the strengths and weaknesses of MOOCs, focusing particularly on the following question:

Can or do MOOCs provide the learning and skills that students will need in the future? 

I can in fact provide you with the short answer now: a resounding NO, for both kinds of MOOC, although one is a bit better than the other! Tune in later for the full details.

Feedback, please

In the meantime, I need to know whether I have got it right in describing the two kinds of MOOCs. Does my description – because that is all it’s meant to be at this stage – match your experience of MOOCs? Have I missed important characteristics? Do I have my facts wrong? Is this useful or is there a better way to approach this topic?


  1. “no or very light discussion moderation: the extent to which the discussion or comments are moderated varies probably more than any other feature in xMOOCs, but at its most, moderation is directed at all participants rather than to individuals. Because of the very large numbers participating and commenting, moderation of individual comments by the instructor(s) offering the MOOC is impossible. Some instructors offer no moderation whatsoever, so participants rely on other participants to respond to questions or comments. Some instructors ‘sample’ comments and questions, and post comments in response to these. Some instructors use teaching assistants to comb for or identify common areas of concern shared by a number of participants then the instructor or teaching assistants will respond. However, in most cases, participants moderate each other’s comments or questions”

    I do wish people would not post stuff like this as a standard. It isn’t impossible, as I know of lecturers (including two I have worked with, and one who I know) who have happily moderated forums on a MOOC.

    You should also add Ask Me Anything style threads to the above – so structuring the forum to focus forum activity into areas on which an instructor can focus.

  2. Hi Tony

    x or c, what a mooc is depends on context to shape it. MOOC’s are described much as if they were a substitute for a course that delivers a “package” of content. The “x” version seems closest as a modified form of the traditional large lecture. But “c” has such an idea buried in it. Yet the MOOC’s, unlike the simplistic bricks to clicks of e-learning, are mostly “events” and not part of a MOOC “U”. In fact, it has been noted that many who participate have degrees and thus selectively pick what they find useful from MOOC’s. Thus design, purpose and value need to be recontextualized.

    The problem goes back 1000 years to the idea of a university. That “idea” is changing/has changed both for those who attend and the institutions. MOOC’s need to be considered in the original and the shifting context or they need to be judged in form/function differently. Thus, the descriptions above of both the “x” and “c” MOOC’s and evaluations are incomplete.

  3. I disagree with your resounding no. MOOCs are useful for anyone who already has a foundation, to stay current in their field. Computer science is an easy example here. If you did your degree 10 or 20 years ago, you have the foundations – but may not have the skills in the latest computer programming languages. MOOCs provide a great way for you to develop those skills. You can do it, because you have the foundations already.
    So, MOOCs are really useful for professional development. I think this is to true for the field of education as well. As long as you have the foundations, MOOCs introduce to new ideas and concepts.
    The issue is that they are not a good tool for learning the foundations – they are great for professional development but not so great for developing the professional.

  4. Tony: Pls. clarify where the distinction between cMOOCs and xMOOCs was articulated? By whom? Who came up with the concept of cMOOC and xMOOC? If cMOOC means “connectivist MOOC”, then all of the definitions of connectivist learning theory apply. And it makes cMOOCs MORE rather than less, part of the MOOC family.

    Both Siemens and Downes have acknowledged that MOOCs represent the ‘modus operandi’ of connectivist learning theory.

    Moreover, the original MOOC ( created and implemented by Siemens/Downes) was the for-credit University of Manitoba course: CCK08. This course, ‘Connectivism and Connective Knowledge’ enrolled 25 for-credit paying students,and was open to public participation (which generate 2,500 enrollees, of which approximately 1% completed the course).

    Tony: would you not agree that the original ‘MOOC’ by Siemens and Downes was an xMOOC?

    So where does cMOOC come in, either in theory or practice? And more specifically, by whom?? Is it real, and distinct from the connectivist MOOC and if so, how? And why should we care? What is wrong with the human-organized Community of Practice?????

    • Hi, Linda
      Stephen Downes made the distinction shortly after Coursera, Udacity and edX MOOCs appeared, in a post on his blog July 20, 2012 ( and endorsed at the same time by George Siemens and Dave Cormier
      CCK08 was offered by the Extension (Continuing Ed) division of the University of Manitoba, so my understanding is that although there were tuition-paying students as well as the open and free students, this was NOT part of a degree course.
      No, I would not agree that CCK08 was an xMOOC. It had both a different design and different philosophy to those that came later from Stanford, et al. I think therefore the distinction is useful, although given the way MOOCs are evolving,the distinction is increasing becoming blurred.
      The main difference I see between cMOOCs and communities of practice is that the latter is a general approach to lifelong learning, while cMOOCs are a particular application using the web and social media. That’s my view though and it may not be shared by cMOOC proponents.
      There’s nothing ‘wrong’ with human-organized communities of practice, or even cMOOCs, but as with all approaches to education, their success or otherwise depends on a wide range of factors which are often not always present, and proponents have to be careful not to oversell them.

  5. I visualize xMOOCS in the deductive model of learning and cMOOCS in an inductive model. I’ve participated in both and on a personal level find cMOOCS much more engaging due to the “your experience is valid truth” philosophy over the xMOOCS constraint of conventional core content, due to its need to narrow the content in some manageable way.

    In a cMOOC I could author learning artifacts, which were accepted as equally valid sources of knowledge as the conventional learning resources. However, in an xMOOC my contributions of self-authored content was (rightly) viewed as interpretation, and assigned a status falling below that of the selected content set by the xMOOCS course authors.

    In the xMOOCS, the need for the community to begin with a fact based, common core of knowledge is critical if the goal of building upon or debunking the content is achieved. The important distinction in an xMOOC is the presentation, exploration, debate and primacy of explicit knowledge, which is embedded in the community, from that of a cMOOC with an emphasis on tacit knowledge, acquired through unique perspectives that serves as the point of departure.

    In a cMOOC the common core of knowledge is experience and perspective, which is often not yet recognized even by the owner until the act of creating it while connecting with others presents itself. The robust dialog between the participants serves as the fund of primary, common core of knowledge, and learning is more a product of argumentation than a outcome of discussion of the empirical evidence so essential to an xMOOC-learning mode.

    Bottom line is that both learning models are legitimate ways of learning and serve the participants in different and important ways. Perhaps they are better approached as a learning model continuum made possible by social media and the net. They share more than they differ in terms of learning potential, community building, knowledge processing, reflective learning and design architecture. They differ most in definition of knowledge authority.

    The most intractable problem for both learning models however remains assessment. So far there has been no agreed upon way to assess a participant’s progress- either by a decision of peers or via an mentor/agent assigned such powers of subjective evaluation, and still keep the principles of a MOOC in tact. Without some sort of conventional award system of academic currency, the learning benefits remain intangible, personal, often highly valued and useful, learning experiences.

  6. Tony – I wrote a similar but much less detailed post about a year ago.

    I think what you have written matches my experience in many respects. The only difference would be that I see cMOOCs and xMOOCs as being on a spectrum from one philosophical/pedagogical approach to another. Even cMOOCs can occupy different positions on this spectrum, but in my experience tend to be more similar to each other than xMOOCs.

    It’s interesting that the paper that has just come out which reviews MOOC research, makes a point of mentioning bMOOCs (blended) and smOOCs (small open online courses).

    Mohamed, A., Yousef, F., Chatti, M. A., & Schroeder, U. (2014). MOOCs: A Review of the State-of-the-Art. In CSEDU2014-6thInternationalConferenceonComputerSupportedEducation 10 (pp. 9–20).

    • Many thanks for this, Jenny. I really admire your blog, and it’s great to have your endorsement for what I’m writing in my book.
      I completely agree that any application of teaching or theory is likely to be somewhere on a continuum than at an absolute point. My experience of #Change11 was that it had elements of xMOOCs in it, in the form of weekly ‘lectures’ for instance, while many xMOOCs have networks of connections in the discussions or comments, and of course we are likely to see more convergence over time between cMOOCs and xMOOCs.
      I do worry though when I begin to see descriptions such as bMOOCs and SPOCs. There seems to be an attempt, particularly by the Ivy League universities, to try to re-define all online learning as a variant of MOOCs, when in fact we have had SPOCS – small, private online courses – since 1995, in the form of online distance credit courses. Open universities also have been offering sMOOCs since the late 1990s, as have many continuing education departments. However their designs are very different – and I would argue, much more effective. It’s about time MOOC providers read and applied the research literature on, and best practices of, for-credit online learning, instead of trying to force an ineffective MOOC design on all online learning.

      • Tony – thanks for your reply. I agree that there’s a lot of literature on effective and best online teaching and learning practices, but until recently all this has been related to small closed online courses. In the past, I have done a lot of teaching and learning in such courses and so I’m familiar with what is widely considered to be best practices in relation to these.

        But the ‘massive’ of MOOCs means that not all these practices will transfer although a very good attempt has been made by some xMOOCs, e.g. Coursera’s Modern and Contemporary American poetry course which is hugely successful and which I have taken, and Edinburgh University’s EDCMOOC which I have not taken, but I have read the related research papers.

        It seems to me that both these MOOCs (and there must be more that I am not aware of) have considered carefully how best practices will scale. However, whilst of course as educators we should always strive for best practice (however this may be individually interpreted), I think the original cMOOCs were looking for *new* practices. They were experimenting, which necessarily, I think, means that not everything in the first instance could always be good or best practice, simply because it’s an untested practice.

        Given that these original cMOOCs were looking for new practices, I’m not sure how helpful it is to judge them against old practices. On the other hand, as has been noted by some researchers recently, educators need to be ethical, meaning they need to be aware of their duty of care and be committed to ‘doing no harm’. It seems to me that perhaps what we need to be considering is not whether the “new” practices measure up to past best practices, but whether in looking for new practices, any harm has been done, and if so, what are the implications of this harm.

        What I am seeing is not that MOOC designers are forcing ineffective design on all online learning, but rather that they are either looking for ways to scale best practices (in the case of xMOOCs) or to find ‘new’ practices in the case of cMOOCs (disrupt existing systems).

        • Many thanks, again, Jenny, for great comments, both to me and Linda.
          With respect to innovation vs best practice, I completely agree. Innovation by definition means doing something different, but here again I think we have to make some distinctions between Coursera/Udacity, edX and cMOOCs, and also understand how effective innovation works.
          What I really object to is the hubris of the Coursera people, and from MIT the deliberate avoidance or acknowledgement of previous, relevant research. There is nothing innovative pedagogically about either Coursera or edX. What they are both trying to do though is to prove you can teach thousands of people effectively with one teacher using a transmissive pedagogy. That’s fine as a statement of intent, but good innovators build on previous research, not ignore it, and they wait until they have some evidence to support their claims before rushing to the media. edX people such as Agarwal claimed they had found that immediate feedback helps comprehension, among many other claims, a fact that has been known since Skinner and teaching machines in the 1950s. WebCT, the very first LMS, built multiple-choice CMAs with immediate feedback into its platform from the start in 1995. So with regard to xMOOCs, as was said about Freud, what is true is not new and what is new is not true.
          cMOOCs are different. This is a new pedagogical development, or rather a bolder step forward but nevertheless there are still relevant previous practices, such as social constructivism and online collaborative learning, which show up in cMOOCs. Again though from my perspective the claims for cMOOCs to date stretch far beyond the evidence, and the proponents have again failed to take account of relevant previous research, such as how to make collaboration lead to deep learning (Linda Harasim’s work) and how to make online groups collaborate effectively (Brindley and Walti, 2009). These practices may not scale up, but they should be considered and tested or adapted for large scale applications, not just ignored.
          But then most MOOC innovators are computer scientists and are either ignorant of or willfully ignore educational research. If educators treated computer scientists the way they treat educators, e.g. ‘AI doesn’t work, because I believe it doesn’t work and/or is too expensive, so we will just ignore it’, they would – rightly – be up in arms at educators.
          Having said that, I really do appreciate your comments, Jenny. My argument/anger is not with you but with the arrogance behind the main Ivy League MOOC proponents.
          This is an important debate and it’s great that you, Linda and Stephen are open and willing to participate.

  7. Tracking definitions of xMOOCs and cMOOCs is frustrating and almost futile because:
    1. those who coined the various terms tend to be all over the map, divergent and not on the same page when it comes to defining what a MOOC is:
    2. Siemens coined the term “Connectivist Learning Theory” (2004), and Downes went along (2006) although they each use the connectivist term very loosely and without specification in subsequent years, especially when referring to cMOOCs;
    3. Dave Cormier, manager of web communications and innovations at the University of Prince Edward Island, dubbed the CCK08 course a “massive open online course.” (Visit for more on the Canadian connection.)” (Tamburri, 2012, para. 6)..but altho he coined the term MOOC, he has been left of subsequent discussion of what a MOOC stands for or whether there is a cMOOC or xMOOC;
    4. Tony: you cite Downes as claiming that he coined the terms xMOOC and cMOOC to distinguish different pedagogies. I understand that cMOOC refers to connectivist MOOC but I have found no evidence that CCK08 was not a formal course: it was. Even DOwnes agrees that it was a course.
    5. Tony: You argue that the 2008 CCK08 course by Siemens and Downes was non-credit and therefore not an xMOOC. But xMOOCs are typically non-credit. What is the difference between the UofManitoba course and the Stanford course, except that the UManitoba course eschewed much instructional organization or curriculum (altho a wiki was eventually created by the instructors (the students chose not to contribute at all to this wiki).
    “Dr. Siemens, in fact, and a small cohort of Canadians were the pioneers of the model that has taken the higher-ed world by storm. In 2008, he and Stephen Downes, senior researcher at the National Research Council of Canada, launched a course on learning theory through the University of Manitoba. About 25 paying students enrolled in the course, along with some 2,300 online students who took it for free. “ There was 1% completion by the participants.
    7. Sounds like CCK08 was a wonky formal course that did not succeed, much like most xMOOCs.
    8. Downes argues that cMOOCs are COURSES, NOT COMMUNITIES. (The Rise of MOOCs: Past Successes, Future Challenges Mar 24, 2014) SLIDE 12
    9. Siemens writes that the differences between cMOOC and xMOOC are two:
    a. Ideology (???): what does this mean?
    b. Level of funding: “Largely lost in the conversation around MOOCs is the different ideology that drives what are currently two broad MOOC offerings: the connectivist MOOCs (cMOOCs?) that I have been involved with since 2008 (with people like Stephen Downes, Jim Groom, Dave Cormier, Alan Levine, Wendy Drexler, Inge de Waard, Ray Schroeder, David Wiley, Alec Couros, and others) and the well-financed MOOCs.”

    10. The definitions are bewildering, because they are very subjective and not empirically not theoretically defined. The main difference, Siemens notes, is the level of funding.

    11. You describe cMOOCs as ‘online’ communities of practice, while Downes is emphatic that “cMOOCs are NOT communities, websites or collections….”

    12. Connectivist ‘theory’ links both cMOOCs and xMOOCs; AI or some form of machine learning is acknowledged as essential to coordinate the tsunami of communication bites that result if participants respond to a flurry or notes, tweets, messages with little or no context. It is to me totally unclear and wrong that humans are not responsible for organizing the learning experiences.
    13. Downes is very much against human teachers or moderators or facilitators. Why is he so adamantly against human educators and/or group learning that is facilitated by human beings? He in effect replaces human organizing by technical (AI) intervention. And now he favors” PLEs (Personal Learning Environments) over collaboration.

    14. Siemens wishes to get rid of the concept of MOOCs:

    In the interview with Tamburri (2014, bottom of page), Siemens claims to want to get rid of the acronym “MOOC”:
    “University Affairs: I read an interview in which you said “If 2012 was the year of the MOOC, 2013 is shaping up to be the year of the anti-MOOC.” Will we still be talking about them in 2014?
    Dr. Siemens: I think we still will be. As the year progresses, though, I’m hoping that we will start to talk more about digital and online learning and blended learning, because I think that’s what MOOCs really reflect. MOOC is still a term that allows us to encapsulate a movement. I’m reminded of what happened with Web 2.0. In 2004 and 2005 we were inundated with Web 2.0 nonsense. Web 2.0 was supposed to do everything, make our coffee, raise our kids, change society. We all hated the term by the time it was done. MOOCs, I think, will do exactly the same thing. Going forward, I hope we will be able to do away with the MOOC acronym.
    University Affairs: Do you really want to do away with the MOOC acronym that you helped popularize?
    Dr. Siemens: If you overhype something, you eventually learn to hate it. The legacy of MOOCs in the higher education system, I think, will be valuable. The acronym, though, not so much.”
    What has been lost (among so many important things) is the crucial role of Learning Theory. The connectivist theory of learning, whatever it may or may not mean, is not being referenced by Siemens, DOwnes, or most of those who are writing about MOOCs.

    Why have we avoided any examination or discussion of ‘connectivist learning theory’? Let’s look at and discussion what Siemens and Downes meant by connectivism, and how MOOCs (whether x or c) reflect that theory?
    Is the theory: Weak? Wrong? Wanting? or Right? And why/how????

    Why divorce discussion of cMOOCs from serious consideration of Connectivist theory?? Too frequently discussions are based on: ‘How I understand cMOOCs’ or what I think they mean’ or ‘my experience of cMOOCs’. NO! The theory provides the analytical framework and let us use the theory to substantiate our ‘experiences’.

    Learning theory is an essential tool: why generate a theory and then ignore it? Let us retain the major distinguishing arguments of Connectivist Learning Theory, by Siemens and Downes, to assit us in considering and analyzing this theory and practice of learning.

    • Linda – I have found your long comment very interesting. Two points in particular have caught my attention. The first is whether or not a cMOOC is a course and the second relates to the status of connectivism as a theory.

      1. Recently Stephen Downes seems to have given in to the idea that a MOOC is a course – and who can blame him given the hype around xMOOCs which definitely are courses in the traditional sense – but even in the link that you provide he is careful to distinguish a cMOOC from a traditional course. However, in the early days some of us, including Stephen, felt it was a pity that the word ‘course’ was used in relation to CCK08, which was intended as an experiment to test out the proposed theory of connectivism in practice and disrupt traditional thinking about education –

      2. The status of connectivism: The subject of CCK08 was connectivism and connective knowledge. It’s purpose was nothing to do with being a MOOC – but rather to discuss these new ideas. In 2009, George Siemens tried to tease out how connectivism related to other learning theories –

      Only a few researchers have tried to challenge connectivism as a learning theory, most recently Clara and Barbera. Stephen Downes made a robust response to their challenge –

      Why have so few researchers discussed connectivism? As you will know, connectivism proposes that learning is the ability to construct and traverse connections and that these connections can be neural-biological, conceptual and external/social. I think its unfortunate that what has captured people’s interest is almost exclusively related to the social aspect rather than to the conceptual and neural-biological, which require an understanding that to ‘learn’ is to ‘acquire certain patterns’. Stephen Downes has been consistent since 2008 (and probably before) in saying this, but few have shown interest or followed it through. Perhaps its too much to take in all at once and so the easier aspect of social connection has been focused on first. Theories take time to be internalized and fully integrated?

      Two more quick points (apologies Tony – for taking over your blog!).

      Yes Stephen Downes has always said that MOOCs were not intended to be communities – but in my experience they can lead to the formation of communities. This has happened in ModPo and in Rhizo14.

      Is Stephen Downes against human educators? He has written extensively about the role of the educator, for example in this post – – which doesn’t quite fit with him being against human educators?

      – and yes, my understanding is that he has always favoured cooperation over collaboration, and given his reasons, but has also acknowledged that each has their place.

      But hopefully he will comment here in his own voice. Mine is just an interpretive voice. Thanks Linda and Tony. Lots to think about here ☺

  8. Dear Sir – I am trying to navigate this crazy world of MOOCS – can you help me out? What is difference between MOOC, cMOOC, xMOOC, SPOOC, SPOC? Thank you

  9. HELP! I have spend more than a month researching cMOOCs versus xMOOCs to find evidence that peer-to-peer interaction in MOOCs makes a difference. This is my only massive term paper and I have to submit my first outline Thursday. Yesterday, I found this recent document co-authored by Siemens: Preparing for the digital university: a review of the history and current state of distance, blended, and online learning.

    Siemens’ group cites Stanford as the first MOOC in 2011. References to connectivism and cMOOCs are hidden predominantly in references, but it is like a censored or self-distancing Siemens. Based on his interviews, I doubt he’d be censored. So when did connectivism and cMOOCs disappear from his writings? Has Siemens parted way with Downes?

  10. Dear Carol,
    I appreciate your frustration and anxiety. But you have focused on a tough topic. First, the big issue is why did you pursue this topic? WHat evidence did you have that this was even do-able? Or that peer-to-peer interaction even occurred ?? I’ve been reading and studying the Siemens and Downes stuff for years and have found absolutely NO Evidence to suggest such a claim as you are trying to prove.

    I don’t mean to belittle you at all—but to point out that research requires prior study of a Research Question, its do-ability and its relevance. Frankly, there is really no demonstrable evidence of any significant difference wrt cMOOCs or xMOOCs In fact, I would argue that cMOOCs and xMOOCs share a common denominator—no teacher.

    I understand that some educators have a warm fuzzy feeling that cMOOCs are a form of Community of Practie. But no one has demonstrated that the current definitions of CoP benefit from or are advanced by cMOOCs. In fact cMOOCs are, imho, pretty much just hanging in space. CoPs have some theoretical and empircal frameworks and bases. cMOOCs are sort of fairy dust. Wannabes. WHo cares? what’s their point? cMOOC folk need to think about the manipulation of ideas, comments, searches, etc by algorithms that are not acting in our best interest. Digital media is not magic. It is algorithms that ‘manipulate’…don’t be like Downes and assume that these AI are your new BFFS.

    From that point, it is 6 of this and a half dozen of that. Whether the ‘MOOC’ is an xMOOC making huge profits for Coursera, or a cMOOC just meandering about. WHat is the difference? Where is the learning? How do they even define learning? I have just written a chapter on Connectivism in my book on Learning Theory, and wish to console you. It is a real mess and there is no evidence, imho, that learning occurs in either x or cMOOCS. In fact, Siemens, Downes or the xMOOC proponents do not even define learning, and hence are unable to address it. Mostly, they focus on content transmission.

    I hope that you are beyond that. And I wish you the best. Do not jump into hype. Seek evidence. We need rsearchers such as you!

    • Thanks for the link, Linda! Gosh, just another 100 research articles down the road and I feel embarrassed that I panicked. When I delved into Andrew Ng’s years of work leading up to the 2 other MOOCs launched with the Stanford AI course, I gained a lot of perspective on the solid pedagogy built into the platforms. I also see connectivism naturally integrating into most MOOCs through learner autonomy to choose how to use them and what to access. I’d hate to take math-based courses in a connectivist MOOC, but PD is workable. I took a short MOOC on accessibility issues in instructional design. I learned a lot, but time prevented my full participation. That would put me in the attrition category, but I met my learning goals, so maybe completion needs to be assessed based on use of the MOOC for those of us in need of quick, cherry-picked learning?


  11. Hi Tony,

    You just directed me here from faceboook to get aquatinted with your research. And as someone who was raised and worked in a number of alternative education communities I am certainly intrigued by the nature of your work. Although this article was written a while ago and you have probably answered this question already, I just wanted to build on the question you closed the article with: “Can or do MOOCs provide the learning and skills that students will need in the future?” I am interested to learn more about what you define as the skills and learning necessary in the future. When I approach a question like this I lay it over the nexus of the future I want and the future most likely to occur. For me also its important to move away from a behaviorist epistemology in which learning and skill are treated as commodities that are exchanged from the teacher to the student. I think humanistic approaches to education make more sense in which learning and skill are based on “tri-ciprocal” ways of being in which symbiosis is achieved in relation to the self, the other, and the environment. Thus through maintaining meaningful relationships and through tending to the learning environment with care I see learning as a process of developmental growth in which the sense of self, ethics, empathy and so on are expanded. This can be quantified (or perhaps qualified) through measuring student worldview progression to one of the many maps developed in developmental psychology.

    Anyway thanks for sharing the article. I thoroughly enjoyed it and have wishes to have the opportunity to jump into an xMOOC and see what its like.

    • Hi, Robin

      Thanks for your comment – much appreciated.

      You write: ‘I am interested to learn more about what you define as the skills and learning necessary in the future.’ The original blog post was a draft of what eventually I wrote in my (free) online, open textbook, ‘Teaching in a Digital Age.’ In this book, I have outlined some of the skills and learning I believe are necessary in a digital age. See specifically Chapter 1.2. However this is not a comprehensive or definitive list and I think it’s a topic that needs a lot more discussion and definition. And I totally agree with you that such skills are not really appropriate to behaviourist approaches – see Chapters 3 and 4 for a range of alternatives. And yes, do try an xMOOC – if you can find one these days


Please enter your comment!
Please enter your name here