Learnlets
Clark Quinn's Learnings about Learning

Rob Penn, CEO of SuddenlySmart (makers of SmartBuilder, one of the new breed of authoring tools), interviewed me last fall about engaging learning: game design, simulations, etc.  It followed one by Professor Allison Rossett of SDSU (also available at the site).

I always find it hard to listen to myself (my voice sounds much better in my head :), and the audio is a little murky, but I hit the usual important notes about focusing on decisions that learners need to be able to make, getting challenge right, capturing misconceptions, and more.

Rob also gets me to discriminate between simulations, scenarios, and games (simulations are just models, scenarios have an initial state and a goal state learners should get to, you can tune a scenario into a game), and I also elaborate how you go from multiple choice, through branching scenarios, to full simulation driven engines (jumping off from Rob’s question instead of first answering it, mea culpa!).

Feedback welcome!

Here’s my mind map of John Romero’s keynote on social gaming (again, done with OmniGraffle on my iPad) (smaller then Kay, as he only talked for half an hour):

The Learning Circuits Blog Big Question of the Month for May is “What will workplace learning technology look like in 2015?”  This is a tough question for me, because I tend to see what could be the workplace tech if we really took advantage of the opportunities. Consequently, my predictions tend to be optimistic, as the real world has a way of not moving near as fast as one could wish.  Still, I actually prefer to think on what could be the possibilities, as it’s more inspiring.  Maybe I’ll answer both.

The opportunities on the table are immense.  Mobile technologies are taking off, we’re getting real power in technology standards (and still some hiccups), and we’re crossing boundaries between reality and virtual worlds.

Smartphones are on the rise, and new portable devices (e.g. tablets) are expanding the possibilities.  It’s highly plausible that we’ll have expanded the performance ecosystem to be location independent, and be providing the 4C’s in ways that allow powerful access, sharing, and collaboration.

Virtual worlds provide a different approach, where instead of augmenting reality, we’re re-contextualized in an artificial but enhanced space where capabilities that don’t exist in the real world are available to us.  We can build 3D models, communicate in micro or macro spaces (within molecules or between galaxies), and open up the hidden components of real spaces.  Again, we can leverage the 4C’s to go beyond courses to a fuller definition of learning.

This can be facilitated by standards.  If HTML 5 coalesces as it should, we can and should be delivering rich interactivity, not just content delivery.  Similarly, if we can move beyond ebook standards to capture interactivity, we can make easy marketplaces to deliver capability that is available regardless of connectivity. Virtual world standards are emerging too, and hopefully some convergence will have happened by 2015!

Also, if our backend systems progress as they can (and should), we should be able to move to Web 3.0 where instead of producers or users, the systems generate content.  We can use semantic technologies to do customized delivery of information, pulling together what we know about the learner (e.g. from a competency map or learning path), about the content available (from a content model), and their tasks (from a job role) and their current context (their location and what’s on their calendar) to serve up just the right information.

This is all possible.  What’s probable?  We’ll have seen major progress in mobile tools, whether companies wake up or it’s just individual initiative to accessorize the brain.  Virtual worlds will also be more prevalent, though not ubiquitous.  Social media systems will be much more integrated into the workflow, and LMS will have become just a cog in the ecosystem, not the ecosystem. The social media will be available whether you’re in-world, in the world, or at your desk.

Semantics, however, are likely to still be nebulous. People are beginning to take advantage of powerful content systems leveraging tagging and flexible delivery, but it’s still embryonic.  There’ll be more pockets, but it won’t be a groundswell yet.

I’m probably still be optimistic, but a guy can hope, and of course strive to make it so.  This is what I do and where I like to play. I welcome more playmates in this great playground of opportunity.

Eric Zimmerman spoke eloquently on games as the second day keynote at DevLearn.  In it, he talked about how systems thinking was important, how games are systems of rules and consequently develop systems thinking.  He talked about how our play brings meaning to the rules, and that creating spaces of possible outcomes allow us to explore.

He ended up advocating that we design for possibilities of unexpected outcomes to create meaning for our learners.  Cammy Bean has blogged the presentation too.

How do you systematically design learning experiences that effectively engage the learner?

This was the question I set out to address more than 5 years ago.  Based upon years of deep investigation into learning & instruction theories and design processes, and practical experience in designing games, I wrote Engaging Learning: Designing e-Learning Simulation Games.

The book was based upon work I’d published as an academic, but was focused very pragmatically.  There were already a few books out about the value of computer games to support learning, notably Marc Prensky’s prescient Digital Game-Based Learning, and Clark Aldrich’s Games and the Future of Learning was also already out.  Subsequently, books by Gee, Shaffer, and others have highlighted the opportunities.

However, I thought and think that my book had a unique contribution, being quite specific around:

• the principles that underpin why games are the best learning
• how to modify your design processes to successfully design games

Having looked at the books out there, I still feel it does the best job of making the case.

At core was an alignment between what makes effective learning practice, and what makes engaging experiences.  Looking across educational theories, repeated elements emerge. Similarly with experience design.  It turns that they perfectly align.  If you recognize that, and can execute against it, your learning will be greater than the sum of the parts, and will both seriously engage and truly educate.  Learning can, and should, be hard fun!

The workshops I’ve run based upon the book have been very well received, reinforcing the value of the book.  Similarly, the content has been solicited as a component of both Silberman’s Handbook of Experiential Learning, and the Guild’s Immersive Learning Simulation report.  I’ve now heard Tony O’Driscoll talk about the design principles for learning experiences in Virtual Worlds (in his and Karl Kapp’s coming book on the topic), and they’re the same principles!

So why hasn’t the book penetrated corporate learning more than it has?  There are several contributing factors.  First, the work I published as an academic didn’t hit the mainstream.  I was part of the international society on computers, and a member of the group specifically about learning through computers (IFIP WG 3.3). They’d just started their own journal, and I wanted to support it (and get a publication). In retrospect, it would’ve been better to publish in one of the more recognized journals on the topic.  As I was overseas, the work never hit the US academic awareness.

Second, I didn’t really understand book marketing then, and trusted that the publisher did.  At the time, they weren’t very pro-active in developing a joint understanding of responsibility (that’s changed), and my book fell through their cracks (and I’m not a marketing person).  (Still, I’m going to be a bit more proactive on the mobile learning book, and they have promised likewise.)

I still firmly believe that the book is the best guide to designing meaningful learning experiences that are centered on deep practice, and a guide for everything from better multiple choice questions to full on simulation-driven serious games.  I’ve tracked the rest of the books out there, and they do a good job of arguing why games are a powerful learning environment, why they make business sense, and more.  However, Engaging Learning is still the best book out there that tells designers how to go about making them.  Sure, I recommend having the workshop to actually get a chance to practice the skills (you know, get your whole team to lift their game), but many who have read it have told me they found value in the book on it’s own.

I don’t say this to generate sales; I get so little it’s not going to make a difference.  I say this because I really worked hard to ensure there is a lot of value in it for you.  I’m just trying to make sure there’s better learning out there, and there’s a lot more need than I can service individually.  There are other good books, Michael Allen’s Guide to eLearning being one, but my book focuses specifically on helping you make more meaningful practice, and that’s a big area of needed improvement, and a major opportunity in making your learning more meaningful.

Please, wherever you draw inspiration, however you figure it out, make more engaging learning. Align the elements of effective practice and the elements of engaging experiences, and make your learning rock. For your learners’ sake, please!

Many years ago, I read of some work being done by Valerie Shute and Jeffrey Bonar that I later got a chance to actually play a (very small) role in (and even later got to work with Valerie, definitely world-class talent).  They had developed three separate tutoring environments (geometric optics, economics, electrical circuits), yet the tutoring engine was essentially the same across all three, not domain specific.  The clever thing they were doing was tutoring on exploration skills, varying one variable at a time, making reasonable increments in values to graph trends, etc.

Subsequent to that, I got involved again in games for learning. What naturally occurred to me was that you could put the same sort of meta-cognitive skill tutoring in a game environment, as you have to digitally create all the elements you’d need to track anyways for the game reasons, and it could be a layer on top.  While this would work in a single game (and we did put a small version into the Quest game), it would be even better on top of a game engine.  I even proposed it as a research project, but the grant reviewers thought that while  a good idea, it was too ambitious (ahead of my time and underestimated :).

The reinterest in so-called 21st century skills, the kind Stephen Downes so eloquently calls an Operating System for the Mind, reawakens the opportunity.  These skills are manifested in activity, and require an understanding of the activity to be able to infer approaches and provide feedback. In a well-defined arena like a designed game environment, we can know the goals and possible actions, and start looking for patterns of behavior.

Game engines, with their fixed primitives, make it easier to define what goals are and consequently to specify the particular goals and makes looking for patterns more generally definable.  Thus, in a game, we can see whether the learners’ exploration is systematic, whether their attempts are as informative as possible, and possibly more.

This is also true of virtual worlds, although only when designed with goals (e.g. from a simulation to a scenario, whether tuned into a game or not).  The benefit of a virtual world is, again, the primitives are fixed, simplifying the task of defining goals and actions.

Of course, building particular types of interaction (e.g. social), particular types of clues (e.g. audio versus visual) and looking for patterns can provide deeper opportunities.  Really, such performance is initially an assessment (one of the facets of what we were doing on the Intellectricity project was building a learner characteristic assessment as a game), and that assessment can trigger intervention as a consequence.  For any malleable skill, we have real opportunities.

Given that much of what is necessary are abilities to research , evaluate the quality of sources, design, experiment, create, and more, these environments are a fascinating opportunity.  I’m not in a situation to lead such an initiative, but I still think it’s a worthwhile undertaking.  Anyone ‘game’?

I was talking with Gina Shreck, who I’d known through Twitter, at a Sun-sponsored happy hour about new learning environments. She’s been quite active in Virtual Worlds (VW), and I was describing an Augmented Reality Game (ARG), and it came to me that there are some really meaningful similarities.

We know from research like John Bransford’s Anchored Instruction and Brown, Collins, & Duguid’s Situated Cognition that learning works better in context (even if you spread across contexts to generalize).  What I realized is that both approaches are really using technology to bring context for learning into vivid relief.  I’ve been active in games for learning because it provided meaningful practice, and of course VW’s can be used to host games in (realizing that VW’s aren’t inherently games, but instead are just environments), and so are ARG’s.

Even when designed for learning, the point is to try to enrich the context.  Web-based games are the easiest, but there are times when more full contextualization is necessary, and the different environments offer different affordances or capabilities.

Despite the overhead, VWs are immersive in that your avatar is totally ‘in world’, and you can design that world to be anyplace/anytime you want it to be.  You can design the contingencies the way you want.  While most valuable for 3D, it may also be important for when total difference is necessary.  Specific examples include building real world structures that must be explored or investigated, for learning purposes.

On the other hand, ARGs are set in the real world, but specific constraints can be introduced.  You can have specific events, materials, and people (real or virtual) appear in the world you want.  Again, you want to develop associated decision making for those explored contexts.

The reason to use an ARG is to develop the ability to develop the capability in situ, that is, as close to the real world context as possible, whereas VWs can add extra dimensions, or work for contexts that are too expensive or dangerous to do live.  That’s also true for non-VW games as well, of course.

The point is to minimize distance and maximize transfer from learning context to real world application.  The overhead to take advantage of these sorts of capabilities is dropping quite rapidly. The goal is to discover the degree and type of contextualization needed (as well as pocketbook, of course), and decide what environment offers the necessary depth and value to achieve the outcomes you need.  However, you need to understand the full repertoire of tools available, and their affordances, to optimally choose an approach.  So, game on!

Yesterday I had the pleasure of working with a team on a grant to use games as a context to conduct high stakes cognitive assessments.  The cognitive tasks for assessment are remarkably abstract, e.g do this particular discrimination task (ie look at a string of four characters and signal if one is a vowel), sometimes while monitoring another situation or as an attention-distractor from another task, but the tasks that they are matched to range from very expensive to life-saving..  The goal is to establish a baseline, and then look for decrements at particular instances before a crucial task, indicating lack of readiness.

The interesting thing is the challenge of placing these tasks in a meaningful context.  It’s creative, and consequently fun.  It’s also collaborative, and when you get divergent contributors in a safe environment, you can really get productive synergy going.  We got together the night before for a meal and some social lubricants, and the next day spent hours in a conference room discussing, whiteboarding and generally designing.

One of the problems is that there have been diverse project specifications from the granting organization, and lack of access to the intended audience. We had some feedback that there should be minimal ’story’, and very clearly that if the audience doesn’t perceive value, they can ignore the activity completely.  Also, there were some important constraints on how much we could change the core task without invalidating the deep research base.  Fortunately, a background in cog psych as well as having the minds behind the tests with us allowed a reasonable guess.  Still, a bit of a challenge.

We focused in early on the value, and I brought up that if the cognitive activity produces improvements in ability, that it’s training as well as assessment, and the audience cares very much about being able to do the job.  That hadn’t been determined to date, but may be available.  We also talked about marketing the value, and if the assessment can in this case (as it has in the past) serve as a very accurate predictor of performance (e.g. detecting a decrement in performance without prior knowledge), that may provide the necessary motivation.

When it comes to design, I’ve made a claim before that you can’t give me an objective I can’t design a game for (I reserve the right to raise the objective ‘high’ enough, but have yet to be proved wrong; it’s an outcome of the engaging learning framework), and this isn’t an exception. In fact, we came up with numerous possible settings, originally for what we were told of the mission, and then for a more near-term mission.  We also came up with relative degrees of abstraction from real (e.g. closely aligned to real task) to essentially arbitrary (like Tetris has little correlation to real time).  The fact of the matter is, you can embed meaningful tasks in appropriate contexts, and tune into a game no matter what the objective is.  It just takes systematic creativity (not an oxymoron), as in the heuristics I’ve talked about previously in two spots.

Since we don’t yet have access to the audience (though we know who they are), I suggested that we need to mock up several different plausible looks and trial them when they do get access (they’re working on that).  They had talked to some stakeholders, but that’s not reliable, for reasons I related to them.  In the process of designing the Quest game, we talked to the counselors who worked with these ‘at risk’ youth, who suggested this issue was smart shopping and cooking.  Fortunately, we then got to talk to some of the youth themselves, who responded “yeah, that’s important, but what’s really important is…” and proceeded to give us a set of relationships that then became key to the game.  Lesson: don’t just listen to the managers, or just the trainers, or just…all those are important, but they may not be right.

It was easy to consider a number of degrees of story ‘depth’, and visual styles to go with each.  At least, they’re in my head, but I went out and grabbed screenshots of various things that could serve as models.  You want the game mechanics to reflect the cognitive task, but you can wrap a number of different looks round that. We’ll pull together our notes, get some storyboards generated, but we managed to sketch out five separate games for what evidence suggests are likely to be the most important skills.

And that’s the real lesson, that it can be done, reliably and repeatedly.  And that’s important, because if you can’t, then it’s all well and good to talk about the value of games as learning environments, but it’s a waste of time if you don’t have an associated design process.  Fortunately, I can still comfortably say: “learning can, and should, be hard fun“.

Several things got up my nose yesterday (and I don’t mean literally :).  I listened in on the Corporate Learning Trends event in the morning, and in the evening participated in #lrnchat.  Don’t get me wrong, both events were great: great presentations organized by Tony Karrer, with examples coordinated by Judy Brown on mobile, Bob Mosher on performance support, Karl Kapp on games & simulations, and Tony on asynchronous elearning (all folks I know and respect); and a great lrnchat session as always with Marcia Conner coordinating fantastic participation by a whole host of great folks.  It’s just that several continuing beliefs surfaced that we’ve really got to address.

The first one was the notion that games and simulations are about tarted up quiz shows.  Let me be clear, these are a last resort!  When you’ve addressed the important decisions, and there’s still some knowledge that absolutely has to be memorized, not looked up, they’re ok.  But they’re not your starting point!  Games should be first thought of as your best practice environment for skills, not knowledge recitation.  What’s going to make a difference in learner (and organizational) performance is not rote knowledge, but meaningful decisions.  That is where games shine.

Ok, as Treena Grevatt pointed out, these ‘frame games’ may serve as the easiest entry point for organizational acceptance, but only if you ‘get it’ really, and are only using them as an entry point to do meaningful stuff.  Otherwise, it’s still lipstick on a pig.

The problem is, we already have a problem with our formal learning being too knowledge focused, and not skill focused, and a tool to make drill and kill easy isn’t going to help us remedy the problem.  So, please: first get that games are really deeply contextualized, immersive, challenging skill practice.  Then, when your analysis has addressed that and there still are knowledge components, bring in the quiz show games.  If you ‘get’ that, then you might use a stealth policy, but only then.

The second problem had to do with mobile learning.  There were still notions that mobile learning could be about courses on a phone  and that there’s not really an audience.  Look, depending on what metrics you pay attention to, the mobile workforce can be anywhere from 20-40% of your workforce.  Sales reps, telecommuters, field engineers, execs, the list goes on. And that doesn’t even tap into the folks who want access for convenience!

And it’s not about courses.  It has been, and can be done, but that’s not the real win.  As an adjunct to a course, absolutely.  Reactivate knowledge (developing learners), update it with podcasts (Chris von Koschembahr had a nice way to interview yourself, controlling the outcome :), review stories, solve problems, review with mentors, etc.

The real win, however (as Judy and Bob both pointed out), is performance support. This can include references, job aids, how to videos, connections to experts, and more.  This is huge, yet people don’t seem to be seeing this opportunity yet.

Mobile is ready for primetime. There are ways to deal with screen sizes, security, and cross-platform differences.  Next to social learning, I reckon it’s the greatest missed opportunity going.

Speaking of performance support, I do have to admit how surprised I was that people were thinking that single sourcing content to populate help systems, manuals, and training was a new idea.  This really isn’t a misconception, it’s just surprising.  I led a project developing such an approach years ago now, and it’s another big opportunity.  Still ahead of the curve, though, more so than the other two.

The point being, the more you tie these together, the greater the synergy: the whole is greater than the sum of the parts. And having been out saying these things for years, it continues to surprise me that the meme hasn’t propagated any further than it has.  And that’s my learning, that changing minds is a tough job.  But still an important one.  Evangelism, anyone?

On last week’s #lrnchat, which I missed most of for my lad’s band concert, I tuned in during a break and saw that Marcia Conner (@marciamarcia) had asked a question I wanted to answer (but couldn’t in 140 chars :).  She asked: “Would someone explain diff between sims (often used well for ed) and VWs?”  She was concerned that some people were using them interchangeably, and I do think it’s important to have some clear definitions.

I stipulate (and would love to get agreement on) a definition that works like this:

• A simulation is, technically, just a model.  It’s captures the relationships of some part of the world (real or virtual), typically not all.  It can be in any potential state, and be manipulated to any other valid state.
• When we put that simulation into an initial state, and ask someone to take it to a particular goal state, I want to call that a scenario.  And, typically, we wrap a story around it.
• We can tune that scenario into a game.  Not turn it, tune it.  A game is a scenario that’s been optimized to have just the right (subjective) level of challenge, a story learners care about, and a bunch of other elements that characterize an engaging experience.

So what’s a Virtual World?  In the above definition, it’s a simulation with the particular characteristics that it’s 3D, and typically also can host many individuals within it.  Now, the infamous World of Warcraft has been turned into a game by a) embedding a bunch of quests (initial states where you try to achieve certain goal states) and b) tuning the experience to be compelling (even addictive).

It gets interesting when we start talking about learning in the context of sims, games, and Virtual Worlds.  A simulation, for a motivated and effective self-learner, is a powerful learning environment.  They can explore the relationships to their desired level of understanding.  The only problem is that motivated and effective self-learners are unfortunately rare.  So, we more typically create scenarios.

When you choose an initial state, and properly choose the goal state, you can ensure that they can’t achieve the goal state until they fully have grasped the nuances of the relationships and can act upon them in specific ways.  That’s the essence of serious game design! This is, I argue, the best learning practice next to live performance with mentoring.  The benefits to scenarios, of course, are that live performance can have costly consequences (e.g. losing money, breaking things, or killing people) and individual mentoring doesn’t scale well.

Are there reasons to tune a scenario into a game?  I want to argue that there are.  First of all, there are the motivational aspects, keeping the learner’s interests.  Second, optimizing the challenge means that the learner is moving through in the minimal amount of time.  Finally, we can alter the storyline to make it more meaningful – exaggerating characters or motives or context – which actually brings the practice environment closer to the urgency likely to be felt in the real world, when it matters. Truly, learning can and should be ‘hard fun’!

How about learning in virtual worlds?  I’ve talked about this before, but certainly, I believe, if the learning objectives inherently support 3D reasoning, whether industrial plant arrangement and operation, molecular structure, or architecture, absolutely.

However, a virtual world is just a simulation, and if you want learning outcomes, you need either self-directed and motivated learners, or embedded scenarios.  Which is what I have been seeing, for example I have seen a very nice demonstration for insurance adjusting.

In addition, when social interaction matters, there are some interesting opportunities.  Individuals can represent themselves as they please, and can create the contexts they wish as well.  (However, I have also seen what are, essentially, slide presentations in a virtual world, and think that’s ridiculous.)

On the other hand, virtual worlds currently have some overhead issues: learning to be effective in them has a learning curve, and there are technical overheads as well.  Consequently, I have been loath to recommend them for many situations where they could be used, if there isn’t an inherently 3D rationale.

However, I do believe that a) the overheads are rapidly being dropped by advancements both UI and technical and b) that there are some ephemeral things that are still fully to be realized.  People I trust, including Joe Miller and Claudia L’Amoreaux of Linden Labs, Karl Kapp of Bloomsburg University, and Tony O’Driscoll of Duke University, continue to express not only the available, but also the untapped potential.

Still, I think the definitions are solid, and am comfortable with the current assessment of virtual worlds.  I’m willing to be wrong, on the latter :). I welcome your thoughts.