I guess I have a visceral aversion to hype, because my initial reaction to ‘buzz’ is focusing in on the core affordances and disparaging mistaken uses of a new technology.  However, I do eventually open to taking advantage of the affordances in new ways. Case in point: learning styles.  I pointed out the flaws in the thinking several times, and then rethought them (without removing my previous views, I looked for the positive opportunities).  Now, preparing for a presentation, I’m rethinking some of my stances on learning in virtual worlds.

I’ve previously opined that there are two key affordances in virtual worlds: the spatial and the social, and that the technical overheads mean that unless there’s a long term relationship, the associated costs really argue that you should be hitting both.  I’m not changing that, but I was wondering what we might do if we did try to leverage those key affordances deliberately to support learning.

Taking a slightly cheeky approach, and quite willing to discredit presenting powerpoint presentations ‘in world’, I’ve tried to think through some subordinary, ordinary, and potentially extraordinary approaches to learning in a virtual world.  That is, opening learners up both cognitively and emotionally, presenting concepts, having examples available, creating meaningful practice, and scaffolding reflection.  What might we do?

Starting with pedagogy, I think a standard instructional design (read: presentations) is clearly subordinary.  An ordinary pedagogy might be a problem-based approach, but a really extraordinary approach might be to create a full immersive storyline in which the problem is embedded, turning it into a game world: a World of LearnCraft.  The idea is to mimic more closely the urgency typically felt when applying the knowledge in the real world (where it counts) by creating a similarly meaningful storyline to develop the associated motivation.  Then embedding resources in the story would scaffold the learning.  Of course, what I’m really talking about is game design ;).

Working with concepts, just presenting them is subordinary. Ordinary would be having them explorable, mapping them out in space, maybe with a scavenger hunt asking learners to find answers to questions embodied in the model.  A truly extraordinary approach would be to have the learners co-create the concept representation, using the collaborative creation capability available at least in Second Life.

Just having a poster for an example seems subordinary.  Having an example ‘gallery’, where you can examine the problem, the approach, and the results would be an ordinarily good approach. Ideally, the example could have the conceptual model layered on top of the decisions, mapping them to represent how th concept played out in context.  Beyond that, however, having the example be truly exploratory, where you could make certain decisions and see how they play out, and being able to backtrack (particularly with annotation about the mistakes the original team made) would be really extraordinary.

Practice is where we can and should be looking to games.  While having a quiz would be truly subordinary (if not maniacally mistaken), having a problem to solve ‘in world’ would be an ordinary approach. Again, having the problem be situated in a storyline, as the overall pedagogy, would be truly meaningful.  It’s easiest if the task is inherently spatial and social, but we certainly can benefit from the immersion, and building in social learning components can lead to powerful outcomes.

I’m somewhat concerned about trying to make reflection ‘in world’, because it’s inherently an ‘immediate’ environment.  It’s synchronous, and it’s been documented where normally reflective kids can go all ‘twitch’ in a digital environment.  It may be that reflection is ‘best’ when kept out of the world.  But for the sake of argument, let’s consider external reflection to be subordinary, and consider what might be ordinary and extraordinary.  Surely, having an ‘in-world’ but ‘post-experience’ discussion would be the ordinary approach.  Again, co-creating a representation of the underlying model guiding performance would be a really powerful reflective opportunity.

You still want to make some very basic learning decisions about virtual worlds.  If you don’t have an inherent expectation that there’s a long-term relationship with the world, the technical and learning overheads to facility in using the world would clearly suggest that you should seriously ensure that the payoff is worth it (like if the learning outcome is inherently spatial and social) and otherwise consider alternatives.  After that, you want to ensure that you’ve got meaningful practice.  That’s your assessment component, and you do want them applying the knowledge.  I suppose you could have the world be for concepts and examples, and have practice in some other format, but I admit I’m not sure why.  Around the practice, figure out how to embed concept and example resources. Finally, seriously reflect on how you support reflection for your learners.

Serious learning can and does happen in virtual worlds, but to make it happen systematically is a matter of design, not just the platform.  Fair enough?

I was invited (thanks, Eilif!) to attend SRI’s workshop for ADL on SCORM and Virtual Worlds (VW) today.   I furiously tweeted it (check out the #adlvw hashtag), but now it’s time for reflections.   Represented were a number of people from various VW vendors (at least Qwaq, Second Life, Thinking Worlds), as well as SRI and ADL folks, and Avron Barr representing LETSI.

In case you don’t know, SCORM was developed to be a way to support interoperable content for learning.   However, the demands have grown. Beyond interaction, there’s a desire to have assessment reportable back to an LMS, and as our digital content resources grow larger, to address data quantities that go beyond download.   Angelo Panar from ADL   helped us understand that there are myriad ways that SCORM doesn’t scale well to handle things other than stand-alone objects. Peter Smith from ADL emphasized the importance of game-based learning, and the potential of VWs for meaningful learning.

Ron Edmonds from SRI nicely summarized the intersection: SCORM is standardized and interoperable, VWs are in competition and have vastly different models. The question is, what is the relationship between the two? Eilif Trondsen nicely characterized the situation that learning spans a gap from formal to informal.   SCORM’s highly focused (as of now) on asynchronous independent learner experience, but VWs are about social interaction, and are platforms, where learning experiences can be built.

The questions they were trying to answer were how to design learning experiences and measure/assess them, and then to decide what role SCORM plays.   It occurred to me that there are no unique issues to VWs except the social, so one particular solution is that the problems for SCORM and social media need resolving, and then can be ported to VWs without requiring a unique VW solution.

Another issue is the level of granularity.   If you design a collaborative exercise, and the interaction and collaborative response to reflection questions are what is key for the learning, then it’s a very different situation than when the goal is tightly constrained responses to very specific situations, e.g. the difference between training and education.   Back to the continuum Eilif was talking about, it seems to me that we can match the level of definition of the measure to the desired outcome (duh!).   However, SCORM has trouble with free-f0rm responses, so we get into some issues there.

The obvious ‘easy’ answer is to have SCORM just be a mechanism to introduce existing content objects ‘in world’.   That’s what a number of platforms have done, whether having SCORM objects appear as objects, or an embedded browser presents them.   A more complex alternative is to have an instructor or the learner respond via a custom interface with a response that’s relayed to an LMS using SCORM protocols. But can we go further?

I’ve argued in the past that social interactions should be a design feature only if the learning objective includes social components.   However, I also pointed out today that the VW may only be part of the solution, and when we look at the broader picture of the learning experience, we may well wrap reflection outside the world.   So then our learning model needs to include more than just content presentation, and we start veering off to Educational Modeling Language and the IMS Learning Design specification, which really isn’t yet a part of SCORM (but arguably should be).

Really, our learning categorization has to include activities as broad as mentoring, coached real performance, and social interaction, as well as content exposure, and interactive activities. It needs to span VWs, social media, and more.   It’s about developing learners richly, not just presenting a prix fixe menu.

I’m mindful of the conversation I had with Adam Nelson from Linden Labs (one of many fruitful conversations at the breaks that helped frame the thoughts above), and I asked whether his role for enterprise learning applications included my broad view of learning, that it’s not just about formal learning, or, worse, just ‘training’, but includes mentoring, discussions, all the way to expert collaboration.   That’s not necessarily what we need to track, but we do need to see the results, to look for opportunities (adding value as facilitators, not just content producers).

It’s clear that ‘in world’, we can have the equivalents of most social media, e.g. collaborative persistent spaces with representations and annotations are a richer form of wiki.   A shared element was the ‘overhead’ in virtual worlds, so the question is whether the affordances of virtual worlds are worth the investment.   I still believe that’s an issue of whether the domain/task is inherently 3D and/or that this is a long-term relationship so the investment is amortized.   There are lots of factors.   Still, it’s an intriguing idea to think that we will be able to interact, communicate, and collaborate in technology-augmented ways that aren’t possible in the real world. Of course, we’ll be able to do those in the real world too, largely, via ARGs (as I previously commented on the connections).

There’s a broad gap between what our tools enable, and what standards are ready to support.   The ultimate question was what the role of ADL would be.   I reckon it’s early days for VWs, so the role in this regard is, to me, track what’s happening and look for patterns that can be extracted and codified for ways to add value.

It’s the wild west or a goldrush right now, and the outcome is still to be decided.   However, the learning potential is, quite frankly, awesome, so it’s an exciting time.   Here’s to adventure!