Archives for April 2019

Surprise and safety

30 April 2019 by Clark Leave a Comment

As I reflect further on the improved surprise model, I realize there’s one thing I missed. The model gives a motivation for learning, and an implication for design. But there’s one thing more in the model, and one more implication for design. And this has to do with safety.

So, first, the initial model says that we learn to minimize surprise. We’re driven to remove the mismatch between what we expect and what occurs. This could lead to a desire to do nothing, or as little as possible, but a further elaboration says we also want to maximize outcomes. Thus, we won’t just sit around, but explore.

That means that helping learners know 2 things: that they want to know this (it’s to optimize what they care about), and that they don’t know it (the gap they have to minimize). If we do that, they’re ready to learn. But there’s one more thing.

We won’t explore other alternatives to see if they’re a better solution if the consequences are high. We’ll only explore if the cost of this exploration isn’t higher than the benefits we gain if it’s better. So that one other things is safety. If it isn’t safe, we’ll stick with the known solution.

Which means that we need to make it safe to explore in our learning. And, that includes both formal learning and informal. Mistakes in learning must be expected and accepted. In formal learning, mistakes are learning opportunities. Have alternatives that represent reliable ways folks go wrong, and it’s ok if they choose those because you have feedback specifically for that selection. And informally, mistakes (not the same ones, or obvious ones, there’s accountability too) are fine when the lesson’s learned.

Understanding how, and why, we learn is critical to optimizing learning. And I think that’s a valuable goal. It’s too important to leave to chance, or old habits. It’s time to be alert to what we know, and put it into practice.

What’s the next buzzword?

24 April 2019 by Clark 2 Comments

I was perusing an old list of potential column topics, and came across one that asked about MOOCs. Now, you probably recognize that the term is pretty much evaporated from any list of top L&D concerns. That’s kind of funny, to me (ok, so you may question my sense of humor). And it makes me wonder what topics are current and are on the horizon. What is the coming buzzword?

I talked in a column about the problem with chasing shiny objects. In short, it’s easy to get swayed by the latest hot topic, and want to be seen to be on top of things. But, as I’ve said repeatedly, a gilded dud is still a dud. If we get the core right first, then we can move on to see what’s real. And, of course, we need to dig into the real affordances, not just the hype (PowerPoint in Second Life, anyone?).

So there are some buzzwords already on the wane. Such as MOOCs. A good sign is if someone’s trademarked it, it’s jumped the shark. Frankly, that already characterizes microlearning. And we’ve had someone recently claim to have invented workflow learning (though it’s been talked about for years). When they’re fighting about ownership, it’s done.

What’s waxing as opposed to waning? How about ‘bots? That’s the topic du jour! Often, as part of AI; as is Machine Learning, Deep Learning, and so on. Also Analytics (I think Big Data is already in the last paragraph’s category). Not necessarily bad, but part of this phenomenon is a lack of clarity about what we mean when we use any of these terms. So, maybe it is like AI: if you know what you’re talking about, it’s no longer new and shiny! And of course, AR and VR are very much now. And personalized and adaptive! (Time for some ownership moves!)

So here’s the question: what’s the next buzzword? Would that it were learning science! Ok, there’s been a bit of a resurgence (time to plug the coming Science of Learning Summit, with the usual caveat), but not near enough. C’mon, folks, lets get together and work on taking your design approaches and tuning them up! Of course, I could wish we’re talking IA instead of AI, too. What else? Contextual. Content Systems. Those are my thinking (and I’ve been talking about these things for years; maybe it’s time).

So, what’s on your list? What’s next? What’s ready for primetime? Wearables? Post-AI? (I just made that up.) I look forward to hearing your thoughts!

Reflection on reflection

23 April 2019 by Clark Leave a Comment

Of late, there’ve been a few dialogs on Twitter. As I opined in the recent podcast I was interviewed in, using Twitter for a dialog is kind of new. I’m not talking about a tweet chat like #lrnchat (which I think is a great thing), but a out-loud dialog with others weighing in. And it’s fun, and informative, but occasionally I need to go deeper. So here’s a reflection on reflection.

In that podcast interview, I opined, as I often do, about action and reflection. The starting point is a claim is that our own learning action and then reflection. What I mean is that we act in the world, and if we reflect on it we can learn.

One of the pushbacks was that we can learn without reflection. And, yes, I agree. We can learn without conscious feedback. In fact, in Kathy Sierra’s insightful Badass, she talks about chicken sexing, a task which no one’s been able to make consciously accessible. Things can go below consciousness.

This was related to another pushback: do we really learn differently from chickens and rats? And the answer is no, but what we learn is different. And, further, what we can learn is different. I’ve yet to see rats sending rockets up to the moon to see if it’s made of cheese.

Conscious representations facilitate learning, particularly for things we learn that aren’t strongly tied to our evolved survival. Learning about cognition itself, for instance, the ability to think about our own thinking, is just something that separates us fundamentally. And, to do that well, conscious artifacts facilitate it.

We’ve found that creating conscious frameworks to facilitate our understanding and acquisition are helpful. So, specifically, models and examples are two things that help us develop skills. We use models to guide and review our performance, to guide us. M0dels are conceptual relationships that we can compare to our performance. Examples show how those models play out in particular contexts.

There’s a followup: if learning is action and reflection, then instruction should be designed action and guided reflection. That is: do, get feedback, but also more. To me, models and examples are that additional reflection. We can present them ahead of time (but see Problem-Based Learning), but we should use them as part of the feedback, pointing out how flaws in performance didn’t align with the models, and further examples that illustrate those nuances.

Ok, so I may be playing fast and loose with the notion of reflection here, lumping in models and examples and feedback. However, my point is to try to keep learning not being information dump and knowledge test. We know that won’t lead to meaningful change. If I label it action and reflection, we have a better chance to push for an application-based instruction.

So, I’ll stick to my claim about (designed) action and (guided) reflection, with the caveat that my ‘reflection’ is more than just noodling. And, yes, it’s for learning goals beyond ‘hitting your head on rocks hurts’. But the goals I’m focusing on are the types of goals that will make a difference in individual and organizational success in our society. If I’m pushing too far and too hard, let me know.

Exploration & Surprise

17 April 2019 by Clark 2 Comments

Some weeks back, I posted about surprise. That is, a new model that says that that our brains work to minimize surprise. We learn so as not to be wrong. And that made sense in one way, but left another gap. Another article explains (well, partly; the mathematics are more than I want to wade into) further, and that gives me a new handle on thinking about designing transformative experiences. It’s about the value of exploration to accompany surprise.

The problem with the original story of us just minimizing surprise is that this leads to another inference. Why wouldn’t we want to just hang in a dark warm room? The notion of minimizing surprise did explain people who don’t seem keen to learn, but many of us are. And, as Raph Koster told us in A Theory of Fun, the drive to play games seems to be learning! We want exploration, and the outcomes aren’t certain. This is in conflict.

The new article posits that there’s another factor, the expectation of value. We also want the optimal outcome. The theory says that we’ll be willing to try several relatively equal predicted value outcomes to learn which to choose in the future (if I’ve understood the article correctly). So we will explore even under uncertainty if there’s a benefit to learning.

This doesn’t quite explain things to me. I think it’s missing some emotional aspect. Why would we do things like try out Escape Rooms or The Void (as I’ve done with colleagues)? There’s no real outcome, except perhaps to know about such experiences. But horror movies or thrillers? All we know is that we’ll have our emotions raised and then settled. But maybe that fits into a good outcome.

Still, this gives me a new handle. When I was preparing the Learning Experience Design workshop I gave at Learning Solutions last month, I was talking about ensuring surprise. That is, the learning experience should make learners aware that they didn’t know what the outcome would be. But I knew, and suggested, that there had to be more. They had to care about the outcome. And now we have the hook.

They care about the outcome, because it’ll be a higher value situation once they do! If we do this right, we let them know that they care about the outcome, and they can’t do it now (either they know already, or we have them fail). Then, we can offer them the path to achieve this outcome. If they explore, they’ll learn? If we’ve got a truly meaningful outcome (you’ll now be able to do X) that they truly care about (you do want to be able to do X), you’re now set with emotionally ready learners. Cognitive science models suggest that this should work! :)

To turn it around. the point is that you should create a goal that they should desire, and then demonstrate that they don’t already know it. It’s simplistic, but I think it’s part of creating a transformative experience, one where they are not just ready for the outcome, but eager. And I think that’s desirable. What do you think?

Quinnovations

16 April 2019 by Clark Leave a Comment

I was talking with my lass, and reminiscing about a few things. And, it occurs to me, that I may not have mentioned them all. Worse, I confess, I’m still somewhat proud of them. So, at the risk of self-aggrandizement, I thought I’d share a few of my Quinnovations. There’s a bigger list here, but this is the ‘greatest hits’ list, with some annotation. (Note, I’ve already discussed the game Quest for Independence, one of my most rewarding works.)

One project was a game based upon my PhD topic. I proposed a series of steps involved in analogical reasoning, and tested them both alone and then after some training. I found some improvement (arguing for the value of meta-learning instruction). During my post-doc, a side project was developing a game that embedded analogical reasoning in a story setting. I created a (non-existent) island, and set the story in the myths of the voodoo culture on it. The goal was a research environment for analogical reasoning; the puzzles in the game required making inferences from the culture. Most players were random, interestingly, at a test, but a couple were systematic.

With a colleague, Anne Forster, we came up with an idea for an online conference to preface a face-to-face event. This was back circa 1996, so there weren’t platforms for such. I secured the programming assistance of a couple of the techs in the office I was working for (Open Net), and we developed the environment. In it, six folks reknown in their area conducted overlapping conversations around their topic. This set up the event, and saw vibrant discussions.

A colleague at an organization I was working for, Access Australia CMC, had come up with the idea of competition for school kids to create websites about a topic. With another colleague, we brainstormed a topic for the first running of the event. In it, we had kids report on innovations in their towns that they could share with other towns (anywhere). I led the design and implementation of the competition: site and announcements, getting it up and running. It ended up generating vibrant participation and winning awards.

Upon my return to the US, I led a team to generate a learning system that developed learners’ understanding of themselves as learners. Ultimately, I conceived of a model whereby we profiled learners as to their learning characteristics (NB: not learning styles) and adapted learning on that basis. There was a lot to it: a content model, rules for adaptation, machine learning for continuing improvement, and more. We got it up and running, and while it evaporated in 2001 (as did the organization we worked for), it’s legacy served me in several other projects. (And, while they didn’t base it on our system, to my knowledge, it’s roughly the same architecture being seen in Newton.)

Using the concept of that adaptive system, with one of my clients we pitched and won the right to develop an electronic performance support system. It ended up being a context-sensitive help system (which is what an EPSS really is ;). I created the initial framework which the team executed against (replacing a help system created by the system engineers, not the right team to do it). The design wrote content into a framework that populated the manual (as prescribed by law) and the help system. The client ended up getting a patent on it (with my name on too ;).

Last one I’ll mention for now, a content system for a publisher. They were going to the next generation of their online tool, and were looking for a framework to: incorporate their existing texts, guide the next generation of texts, and support multiple business models. Again pulling on that content structure experience, I gave them a structured content model that met their needs. The model was supposed to be coupled with a tech platform, and that project collapsed, meaning my model didn’t see the light of day. However, I was pleased to find out subsequently that it had a lasting impact on their subsequent works!

The point being that, in conjunction with clients and partners, I have been consistently generating innovations thru the years. I’m not an academic, tho’ I have been and know the research and theories. Instead, I’m a consultant who comes in early, applies the frameworks to come up with ideas that are both good and unique (I capitalize a lot on models I’ve collected over the years), and gets out quickly when I’m no longer adding value. Clients get an outcome that is uniquely appropriate, innovative, and effective. Ideas they likely wouldn’t have come up with on their own! If you’d like to Quinnovate, get in touch!

Learning Tools and Uni Change

11 April 2019 by Clark Leave a Comment

As part of a push for Learning Engineering, Carnegie Mellon University recently released their learning design tools. I’ve been aware of CMU’s Open Learning Initiative for a suite of reasons, and their tools for separate reasons. And I think both are good. I don’t completely align with their approach, but that’s ok, and I regularly cite their lead as a person who’s provided sage advice about doing good learning design. Further, their push, based upon Herb Simon’s thoughts about improving uni education, is a good one. So what’s going on, and why?

First, let’s be fair, most uni learning design isn’t very good. It’s a lot of content dump, and a test. And, yes, I’m stereotyping. But it’s not all that different from what we see too often in corporate elearning. Not enough practice, and too much content. And we know the reasons for this.

For one, experts largely don’t have access to what they do, consciously, owing to the nature of our cognitive architecture. We compile information away, and research from the Cognitive Technology Group at the University of Southern California has estimated that 70% of what experts do isn’t available. They literally can’t tell you what they do! But they can tell you what they know. University professors are not only likely to reflect this relationship, they frequently may not actually be practitioners, so they don’t really do! We’ve compounded the likely focus on ‘know’, not do.

And, of course, most faculty aren’t particularly rewarded for teaching. Even lower tiers on the Carnegie scale of research institutions dream and hire on the potential for research. There may be lip service to quality of teaching but if you can publish and get grants, you’re highly unlikely to be let go without some sort of drastic misstep.

And the solution isn’t, I suggest, trying to get faculty to be expert pedagogues. I suggest that the teaching quality of an institution is perceived, except perhaps the top tier institutions, as a mark of the quality of the institution. And yet the efforts to make teaching important, supported, valued, etc, tends to still be idiosyncratic. Yes, many institutions are creating central bodies to support faculty in improving their classes, but those folks are relatively powerless to substantially change the pedagogy unless they happen to have an eager faculty member.

CMU’s tools align, largely, with doing the right thing, and this is important. The more tools that make it easy to do the right thing, rich pedagogies, the better. It makes much more sense, for instance, to have a default be to have separate feedback for each wrong answer than the alternative. Not that we always see that…but that’s an education problem. We need faculty and support staff to ‘get’ what good learning design is.

Ultimately, this is a good push forward. Combined with greater emphasis on teaching quality, even a movement towards competencies, and rigor in assessment, there’s a hope to get meaningful outcomes from our higher education investment. What I’ve said about K12 also holds true for higher ed, it’s both a curriculum and a pedagogy problem. But we can and should be pushing both forward. Here’s to steps in the right direction!

#LSCon 19 Reflections

5 April 2019 by Clark Leave a Comment

It’s hard to think of now, but last week I was at the Learning Solutions conference. And I had a really great time. I didn’t see as much as I’d like (as you ‘ll see, I was busy), but there were some really worthwhile learnings, and some fun as well. Here are my conference reflections.

For the first time, I rented a scooter. That was a learning all in itself. I’d been having pain, and walking was the worst. The scooter was a way to address that, and it did. I scooted around and avoided much walking. Not all, but a lot. And it was fun to zoop around, but…it was hard to maneuver in small spaces. Like the necessary elevators. And my room. I tried to slow down and do it carefully, and that worked to an extent, but it wasn’t pretty.

The great part was that, having heard of my plight, some friends descended upon my ride and tarted it up with glitter and dangly things. And, best of all, caution tape. Very appropriate. Very much appreciated! And it wasn’t even too hard to take off at the end.

Thus, I was happy to zoom to my room to run my pre-conference workshop on learning experience design. It was designed as an integration of Engaging Learning and the Serious eLearning Manifesto. I snuck a bit of ‘transformation‘ in there as well. The evaluations aren’t back yet, but I think overall it achieved the purpose. One attendee later suggested an improvement that I’d agree on (allowing learners to choose from the topics to workshop on). Always learning!

That evening, we did something I’d never done, Presentation Roulette. The speakers (I agreed to be one, without having seen it before; I do like experimenting [read: living dangerously]) choose a random title out of a sock (well, it was clean) and are then given a deck that Bianca Woods of the Guild had developed for that title, including the silliest pictures she could find on the web. As she describes it, a mashup of presentations and improv comedy. It was very fun, and in particular extremely funny; the other presenters did great jobs. I’ll attend again even if I don’t present!

Tuesday was a normal day (e.g. I didn’t present). As usual, I mindmapped the keynotes (several posts back), cruised the floor, and attended some sessions. The panels were good. I attended the one on the Future of ID, and the comments were insightful about how the tools and goals were changing. Similarly the one on the Future of Work had a convergent message I resonated with, that we need to focus on using tech to augment us on the stuff we’re good at, not try to fight off automation of rote tasks. I also took some time off for calls and work.

That evening, after dinner, some friends and colleagues (they’re the same folks) came over to my suite. (I have gotten lots of accommodations for my situation; and I’m very grateful.) Fueled by libations, we proceeded to gin up an evil plan to control the world (or at least the market). Politically correct it wasn’t, fun it was. Too late to bed.

The next day I was part of the Guild Master panel with about 14 participants. Too many! Great thoughts, and I tried to stifle myself and only make the most cogent points. Apparently I still spoke a little too much. I blame it on this blogging, it gives me lots of thoughts. :) The points I wanted to make were, not surprisingly, about the need for getting back to basics in learning design, and to look beyond optimal execution to continual innovation.

I also sat in an ARK Kit presentation. It made AR seem almost within reach. At this time you still do need some coding, but if it progresses like many tools, much will soon become at a higher level of ability to describe what you want and make it so.

I still wasn’t done, as later that day I also gave my ‘professionalism and myths’ talk. The audience was small but enthusiastic. I do believe we made some converts. I added in not just debunking myths, but how to talk to folks who buy into it. There’s a little learning science in it as well. We really do need to be on a sound basis before we can have credibility.

I have to say, delightedly, that I continue to have folks say that my books have helped them. Different books for different folks, but something I love to hear. As an author, you get some idea of the sales, but none of the impact. Some of these were small effects, and some were “I’ve used this to change my/our practice.” That’s what it’s about, after all, you write a book to effect change. I’m grateful for those who share this insight! In particular, I hear lots of folks using the Myths book in their orgs to counter employees/customers’ misguided intentions. The Revolution book still (or, perhaps, now) has influence. And I still hear about the Games book!

I also slipped away with some more conspirators and experienced The Void. It was hard on my legs (I went with cane, not scooter), mostly because they didn’t have anywhere to sit while you waited!?!?! (I gave them a serve in the too-long post-experience survey.) However, it’s very cool: a compelling experience and great implications for learning. Embedded performance? That would be ‘yes’.

The keynotes, by the way, were excellent AND… I’ve heard over the years that conference organizers say it’s hard to have diversity in speakers. All white males (e.g. me ;), or at least white. This time, there were two women, and two blacks, out of three people. With good messages. It was inspiring to hear and to see! Kudos to David Kelly and the Guild for managing to debunk the barrier.

There was some discussion of whether there was a place for those who proselytize learning science or it was all going commercial and cheap. I feel like there’s a growing interest in the science, but I’m frequently a year or several ahead of the market. In this case, I want to yell “make me right!” This is a field I care about, and we can be doing so much good. I want us to capitalize on that potential. There were new folks looking for solutions and the opportunity to grow. I hope we can make that happen in a positive direction.

Overall, it was a success. I had time with smart colleagues, saw interesting sessions, and met new folks. I presented and got feedback, which is a great cycle. And it was another chance to immerse myself in the state of the industry. Here’s to continual improvement.

Violating Expectations

4 April 2019 by Clark Leave a Comment

As some of you may know, last week I had a surgical procedure. I don’t want to share details, but while it was non-trivial, it went fine. (I’ve talked earlier about the situation beforehand.) What did not go fine was the recovery. I’m good now, but there were a harrowing few days. And I think the reason is of interest, and there’s a lesson. So I thought I’d share.

Now, my only previous experience with surgery was outpatient knee surgery. And it was amazing; I went off pain killers the 2nd day, and recovery was rapid. This, too, was outpatient, and while not as ‘micro’ as the knee surgery, I had no other frame of reference. And that caused a problem.

So, the day of the surgery went about as you might expect. I went in, lay down, woke up somewhere else, and was told things went well. With the benefit of meds, I let folks know I’d lived :), and proceeded to sleep away the afternoon. Come the evening, I was more clear-headed. With good meds, I looked forward to a night’s sleep, and better in the morn.

That night’s sleep was not good. I couldn’t get comfortable, and so couldn’t sleep. Specifically, my left (not bad) side was uncomfortable and so was my right side that was supposedly fixed. I was awake all but maybe 2 hours. Yet, I’d gotten used to sleep deprivation.

I was bothered that the side that had had the surgery, that was now supposed to be free of the cause, still hurt. Differently, perhaps, but still hurt. This was dismaying (to put it mildly). I called Saturday night, and was told that the right side could still hurt for some days. Er, okay…

Saturday night wasn’t really better. I slept maybe 3-4 hours, but lack of comfort meant I was still worried and in addition now sleep-deprived. This wasn’t all, but worth recounting is by Sunday, my whole right leg wasn’t working. Any sort of moving but standing hurt. Not good.

This continued into Monday. Little and bad sleep meant I was going into a mentally challenging state of sleep-deprivation. The lack of right leg action began to make me feel like the whole experiment had failed, and I was going to have to face this again. I put in a call, but my doctor was in surgery. You can imagine I was discouraged and distressed. I was headed to bed when my doctor finally called. They’d scheduled me to see him the next day, and I hung on to that.

When I got to see him on Tuesday morning, I was a wreck. Spaced out from lack of sleep, distressed about my leg, and so on. I first talked to the PA, and then the doc came in. And, I found out a lot more. They’d completely removed the material pressing on the disk, but in so doing they’d likely irritated the nerve. And there could be some bleeding doing that, too. So, my leg hurting was explained. When we talked about meds, they were reminded what I had been on, and how the sudden cessation of that could be problematic. With explanations, and revised recommendations for medications going forward, it seemed promising.

Low and behold, after the visit, things began to fall into place. The medication revisions kicked in, and I felt a lot better. Not good, mind you, but many times better. Finally, I could see how this was all working, and I was progressing!

I awoke this morning and verified that yesterday wasn’t a fluke; I’m on a path to recovery. I still have a backlog of things to deal with, but I can get on that now. And I still hurt in various places. But it’s the right hurt, I now know.

The short version of all this is that expectations matter. Stephanie Burns did her Ph.D. research on the people who succeeded with their goals (vs those who don’t), and found it was the ones who managed their expectations appropriately. Set goals, rewarded them, realized it was a long haul, etc. Similarly for learning; you want expectations to match experience. A mismatch can induce barriers to successful learning. If the experience will be typical, it may not matter so much, but you want to be wary of any ways in which people can find their expectations mismatched. Yes, you want some surprise, but you don’t want people to lose their comprehension of who they are and where they’re going.

I don’t actually blame the doctor. I think they could’ve set my expectations better, but I fear I come across as someone who has an idea of what’s going on. And I should’ve asked more questions. Further, I think there weren’t any flags that I needed such support. Still, it perhaps ought to be automatic. So consider setting expectations. Deliberately. Systematically. You can let them know there might be some surprise, without giving it away. Don’t leave people open to making inappropriate expectations, or you might be unpleasantly surprised.