The Future IMS Learning Design

More blog coverage from IMS in Montreal.

The Future IMS Learning Design

What is the future of learning design? It has been around seven years. But still it disseminates very slowly. Is it time to revise or reinvigorate the specification? Should it be combined with packaging, such as common cartridge? What kind of support is needed for its dissemination?

Joel Greenberg

My own personal view: maybe the world is moving on, and maybe the idea of 'design' is a bit old fashioned. I have been looking at the world of social networking, and my mind is moving away from the model where we are the experts and package the knowledge and sequence it.

Academics love book, and they're very into narrative, heavy narritive, even light-weight narrative is a challenge to them. It's interesting, because they all use techn ology, but there has been a lack of interest in using technology in their teaching. A lot of them ask, what's in it for me? They get more credit for publishing a book, a LAMS sequence wouldn't even be on their c.v.

The overall project: a generic description of services, a methodology for adding those services, and software that suppots this approach. Which led to SLeD protypes, etc. You can Google it.

Conclusions: issues about efficiency, limitations of the generic approach (it gets very complex), complexity of generic serving descriptions (Symbian. They asked, "What would make it work?" I said, if it's no mor difficult to use than PowerPoint. Which it isn't.) Does it scale? Also (Martin Weller) the content gets bound up withn its description. And finally, the difficulty of integrating tools.

Instead of trying to systematize, to sequence, a more appropriate approach is based around patterns and connections. I would question the LAMS / Moodle thing; it looks like overkill. It looks like a huge overhead just to sequence the use of tools within it. It just packs everything up and tells peopl when to use it.

We had a problem. We couldn't get academics to use any of this stuff. But they all have Word, so we looked at that. They were working to schemas. We found that over 80 percent of OU material could be done with one schema; they were changing it (the schema) to change it, not adding value. So now we have this whole system all based around Word - structured interface, graphics from repository. The idea is to lead academics to a lightweight narrative, and having them design around it. Then (what they liked) they can render it in PDF, Moodle, etc.

It took five years to get academics to use this, and it took an edict, "you will use this." The same academics that use templates produced by publishers.

This approach has been pretty effective. Something like 6000 hours of work produced using this stuff. So I'm less interested in the pedagogy, I'm more interested in the process. My experience is that your standard academic simply won't touch this stuff (learning design).

Mile Halm

I can tell you that my feeling is very similar. I recall being in the room when we decided we would begin working on a learning design spec. It was started in 2001, and the main spec came out in 2003. It has produced some research projects, but not mainstream implementations. It's very powerful, but also very complex. We won't find any faculty willing to adopt these tools.

This idea of a community: it needs a community to be successful. A more open, collaborative support environment has potentioal to lead wider adoption. But we (IMS) need to be more open. We write these specs, but there's no adoption.

So, what is the future of learning design? Similer and lighter, things that kids can do, like YouTube. Interoperable - the widgets are a good example. It needs to be more mobile, interactive (service-orientation), flexible and extensible.

Wilbert demonstrated Recourse. Very visual, contains a lot of information, hides a lot of complexity behind, so the user doesn't have to worry about it. The LAMS tool, very similar. But again, I don't know whether faculty will use this in the higher ed stange, maybe the K-12 stage, where there's a much greater emphasis on outcomes.

Prolix: from .lrn this project again focuses on simplicity. That is again what we must concentrate on if we want people to use these tools. Anyone should be able to create these things. That's what we see in web 2.0 - everyone is a creator. And we need to design learniong content in order to make that happen.

Support needs: this is our second project. The first was open source in name, but not in how it operated. We didn't have contributions, we had a grant and we went out and did all the work. This time, we looked at the process, how to engage the community. This time we're getting a lot more activity out of the partners. This idea of creating learning design communities, interacting in real-time -- getting an email here, aftr a month answering a question: that's not what I'm talking about. I mean, IRC, someon has a question, someone from the community answers (rather than us providing the answer for every sinnle person).

On of the things missing from a lot of the open source projects is documentation. If we have to do something in open source, we'll write it down. This has created a lot of trafic on our site. Documentation has a lot to do with creating a useful tool and supporting implementation. Pretty soon you get a rich resource for the entire community. Not just technical: eg. how do we get reuse out of the design itself. Also: how do we create plugins and extensions?

Take the learning designs pec: how can we help people create a very quick implementation that other people can use.

- demo - Weblion Wiki - user created content and support

Somthing like this for learning design would be really useful, and if we made it open, then anyone could contribute content. A user community would be a tremendous was to think about how to support a learning design implementation. As IMS we need to think - not just about learning design but other stuff too - about this, so you don't have to come to a conference to find that there's stuff out there.

Gilbert Paquette

My view is a bit more optiminstic, I don't know why, because we've been working on this for fifteen years, but we're working on hard problems. Learning design has had more of an impact on workplace learning. In corporations, you have to distribute lots of information, if you want to keep your people, and so you have to prepare thinsg in packages, so we're inclined to try to prepare things.

The web it our learning platform, we have to remember that. It's not to prevent the social web activities. But the instructional design is the most improtant part of learning. The insructional designer sees many paths that students could use. It's more than a qustion of interoperability, it's a question of activities, and interaction between actors.

We all agree, I think, it's vry slow adoption. The tools were form-based, not very user-friendly, not many methodological aids, and still no LD repository. And since the specification we've had web 2.0 and web 3.0 - my group has written IMS about the weakness of collaborative activities. And we're seeing competencies, and the specification is weak on that.

Four lines to extend LD access:

1. Simplify the authoring process: simpler visual modeling, visual pattern repositories, design scenarios and aids. Once question: should the teachers be their own designers? Should the larners be their own dsigners? We can see, in some scenarios at least, designers using the tools to create the designs.

2. Provide a learning design run-time engine for interoperability.

3. Profile or simplify the IMS-LD spec

4. Extend the web to social network and web 2.0 applications and contexts.

Simpler Visual Athoring - We saw examples of that today, eg. editing simplification in TELOS

LD Executable pattern rpository - if we had a good repository, with the best instructionaldesign minds around the world providing things, then we wouldn't hav dificulty covincing the teachers to use them.

Run-Time engine: the idea was to delegate CC sequencing to an extrnal tool that plays an IMS-LD file if it's present in the cartridge.

Simplifications: we need new levels in the specification - a new Level A could integrate the most useful Simple Sequencing components. Also we need a way to bring in services. and collaboprative dsign.

Conclusion: it has been six years now we have been working on the LD specification. We need a new group, not only to simplify the specification, but also to integrate the specification into actual practice. If you are interested, send m your email.


Wilbert: (very soft, something about authoring a spec and then just running it) The question is whether we can ever build tools good enough to run them.

Phil: all the people in this field are not in the main-line course production.

Gilbert: still, you ahve some LO repository, if you have incentive, if people contribute, it is recognized. The same could be for larning design. But I agree, it's more instructional designers than are inclined to do this. That's why OUNL, or us, are more involved in this.

Mike: what we've done at Penn State, we have developers at every college now, because if we put our courses online, we get back some of the tuition. But the stuff the;re developing is still old school, essentially a page-turner.

Motts (?): I remember 2003 or 04 sitting in a room with everyone in LD at the time, coming up with the view that the spec had split personalities, with no agreement on what the spec is for. I'm seeing the same split today. Is it a way to create courses? Is it a means of exchanging complete courses complete with thir pedagogy? Is it a way to exchange pedagogical practices? Probably not - we know how hard it is to ven get colleagues to reuse their own work. We need to home in on what is really important about it. Bucause otherwise I am inclined to agree with Joel, that Google Wave will do away with it.

(Comment): what keeps crossing my mind, what is the mcehanism to test the effectiveness of this pedagogy after you've pushed the learning through? How do you know your learning design is not flawed? Does it integrate with something else?

Phil: in the world of social networking, students would rate it badly and it would disappear without a trace.

Guillaume: I am not sure that we have drawn all the needed conclusion from IMS-LD. I'm not sure we are clear what it was - a modeling language, an interoperable spec? It is probably time top steer the committee. We need to have all the lajor actors taling about how to implement the lowest-cost learning deisgn. But I think we also need a pedagogvical language to communicate between teachers - but maybe that's another part.

Gilbert: it's a question fo the egg and the hen. If you don;t ahve a platform, people won't start doing it. And if people don't start doing it, then commercial vendors don't extend their platforms. I agree, setting the goals - maybe the spec does too much.

Guillaume: maybe we should abandon the completeness of IMS-LD, to model all pedagogies.

Chick: it seems a comflict between nstructional designers and teachrs. The spec is designed for instructional designers. But as a faculty member I will never use LD. I watch. As soon as you say "and then you draw a line" I know I will not use a specification. Because then you're not programming any more. I would fiddle with things. I want to be able do my could without drawing lines - I will not draw a line because I'm not a professional learning designer, and that's not how I think. It needs to be designed more along the lines of the way - we create content and then add commentary. We may be elegant at the end, but we're not elegant at the beginning.

Mike: the faculty create online lctures,m not courses

Chuck: my view is, I own my course, I want it to be as messy as posible

Gilbrt: perhaps the spec was too revolutionary. Your approach is very conservative. Perhaps they should learn some pedagogy. They teach in classes, they think they know pedagogy. But all they do is giv information, telling telling tlling. This is zero for learning.

Chuck: this is a language problem.

(Comment): It's a language problem, and it's a results problem.

Mike: in k-12 they're driven by outcomes - proof of learning - perhaps if that were more the csase in higher ed we would see more of the samne.

(Comment): learning design csn be interpreted in so many ways, and that's what makes it difficult. One view - LD is simply to describe pedagogy. But it was viewed as doing to much. And as technology, it's crap. It doesn't do what it was designed to do.

(Comment): I want to respond to the idea that l;earning outcomes are not important in higher ed. In the U.S., that is their mantra these days. I think the time is ripe to focus on outcomes.


  1. Some missed opportunities here - stemming from a naive assumption about "design" that this must mean that the teacher does all the designing up front and then runs (delivers) the design for a cohort.

    That may be "instructional design" but that isn't all of learning design. By exposing the "design" you get to have the conversation about "how do we want to learn together?" and to provide tools to help turn that conversation into action.

    For example, I know of some cases where teachers have put the LAMS authoring interface in front of the students and got them to figure out how they want the session to be organised (they were probably better at it than most HE lecturers...)

    The iCamp project folks worked a lot on the problem of how people figure out how to work together and select appropriate tools (e.g. use a wiki or a google doc? video chat, text chat, forum or twitter?) This is something the process of design can expose; the "choices" offered by an un-designed course in a VLE are often spurious and driven by defaults.

    When we did the Wookie integration, we did it in such a way that you can say "in here we'll need some sort of quiz widget" but only put the questions in when you're using it - late binding tools. That's essential flexibility. LD (or ld) needs that too, it really needs better support for vague activities ("after lunch we'll just leave a gap and figure out what we're doing when we get there"). I think Chuck's point about "mess" is insightful - technologies supporting design have to support mess and vagueness or they aren't really supporting "design" as its really practised.

    I can see the LD negativity - the spec is old and needs rethinking; adoption has been patchy and experimental - but its also the only thing IMS has done that has created a rich dialogue about learning and technology, especially in Europe, and that has driven people to work hard on trying out new things and going in different directions (e.g. the developments of LD in France, Spain and Quebec have been very, very different from those in the Netherlands, UK and US - I think thats a good thing). For that alone its been worth it.

    A problem is that most of the discussion on LD takes place outside IMS, and the IMS processes aren't really designed to work with that kind of open community. Maybe a successor to LD will have to happen outside IMS. But I'm sure we'll see one.

  2. Very nice report on the session and great capturing of the issues.

    I have always thought that to implement any type of learning design effectively and meaningfully you need to have educational designers involved. That is not the way Higher Ed works. It might provide value in corporate spaces but even then, it is difficult. There does seem to be some interest in compulsory education (K-12). So maybe there is an important relationship between compulsory (education or enterprise or gov or military) and learning design?

    One of the main values of specs like IMS SS and LD was the ability to 'port' these designs and the content/courses to other locations, primarily LMSs. It was an approach to interoperability. LD clearly goes further than that. LAMS took the approach of providing portable designs without being content dependent.

    No matter which way you approach it, it seems that there is always effort involved in using a design somewhere else. More importantly, it seems that the effort involved is more than users are prepared to accept. The simple authoring tools is only part of the solution. An 'adopter' or if simple enough tools are present 'adapter' of such content including a design still needs to understand the design before it is selected for reuse. If that is not easy, reuse will not occur.

    At this stage I agree almost completely with Joel and Mike. As Joel says, "the world is moving on" or maybe already has for the most part.

  3. First of all, thanks for posting this. It's very useful.

    I agree with Scott. The assumptions about what LD can be in practice resulted in tremendous opportunity loss. Frankly, I worry that this is a classic example of a spec that isn't ready to be written yet and certainly wasn't ready to be written six years ago. The time to write a standard is when there are several good implementations of an idea that could benefit from harmonization and interoperability. This idea of building a spec first and then trying to figure out what it's good for in practice is a terrible idea.

  4. Thank you Stephen for sharing.

  5. Interesting discussion.

    I disagree with what I take as a general mood music that "the world has moved on". We never even arrived. Surely everyone needs to be very careful about dismissing something that might be good in principle but failed in the past due to poor implementation or other circumstances. Its like saying "there's no point in going to London because last time we tried we missed the train".

    From a K-12 rather than an HE perspective I think that the instructional role of the teacher constantly needs to be restated. There seems to me to be a real danger that people in HE may muddle (or allow their audience to muddle) their roles as theorists and implementers of Learning Technology. "Instruction" may almost be a dirty word in HE, but in schools/K-12 it is absolutely critical.

    I think that "design" and "instruction" are pretty closely related (and I suspect that people who don't like the one probably don't like the other). Assuming that you accept the need for both at least in some contexts, much real-world learning design is done in a fairly seat-of-the-pants, informal way. Whether you approve of this or not, I doubt there will ever be much demand for specs which try to impose on teachers a degree of formality which is alien to their practice.

    And this is what LD appears to me to do. LD grew up in distance learning and I cannot myself see its regular application in the face-to-face classroom.

    The problems with SS seem to me to be different:

    1. excessive complexity in the specs;

    2. consequent lack of tool support;

    3. and perhaps most important, lack of compelling, robust learning objects to sequence in the first place. There is no call for brick-layers if there aren't any decent bricks around to work with. Sequencing is about managing instructional processes. But the only process encapsulated by most current learning content is "read" and "look". No-one is going to get very excited by sequences which say "read A then read B then read C". The requirement for sequencing therefore follows from the availability of more interactive (as well as encapsulated and reusable) learning content.

    People might be interested in the current discussions on this topic on the LETSI Orchestration Working Group. Details (and participation) via I have also posted a summary and a couple of key documents at

    Incidentally, I also agree with Michael Feldstein that spec development has in the past got ahead of implementation (and conformance testing) and this balance needs to be redressed.


Post a Comment

Your comments will be moderated. Sorry, but it's not a nice world out there.

Popular Posts