The opening session for the eLearning Guild’s Performance Support Symposium was an insightful panel moderated by Allison Rossett, with panelists Rose Lawyer of Huntington, John Low of Carney, and Matthew Henzel of American Express on their experiences with performance support (PS) initiatives. Interesting points included taking a holistic perspective, emphasizing change management, and future directions. I note that in some cases my interpretation is not their words :).
I continue to have a problem with the term content as a component of what our field does. I think there are potential problems with the label, so let me make them clear.
What we do is create content. In elearning, we create introductions and concept presentations, we portray examples, and we make interactivities that provide practice. Even in F2F training, we have content and structure around actions we ask our learners to take. At the end of the day, much of what we’re working on is content that is communicated or triggered by learner actions.
However, I think there’s a problem with thinking of it that way. I believe we need to focus on the activity, not on the content. What’s important is the learner’s experience that is created by sequencing content and learner actions, not the content itself. You could present content in different ways (for instance, labeled slides, narrated slideshow, or video) and it’s be semantically equivalent (and please don’t bring up Clark & Kozma), at least for our purposes here.
The problem is that if we focus on content, it becomes too easy to think that content presentation is equivalent to learning. Even if we test knowledge of the content afterwards, it’s not going to lead to meaningful outcomes. Thinking about producing content makes it easy to go astray.
The alternative, however, is still uncertain. Technologically, it makes sense to talk about content management systems, but learning management systems above that is the wrong language. While ‘course management systems’ addresses the real function of such systems, ideally we’d instead be thinking about ‘experience management systems’. Except I don’t think we really have those right now. You might say that trainers or mentors or coaches are that, and I might agree in the latter sense, certainly, though I’m looking for a better branding for the technology infrastructure.
There’s now an Experience API that provides some infrastructure for creating such an experience management system, but there’re still some intermediate steps needed. Fingers crossed.
Ok, so I’m thinking out loud about our language and what the implications are, but I’m a big fan of reflection and I think it’s useful to stop once in a while and think about where we’re at and how we’re doing. I welcome your thoughts.
Our brains are good at certain things, and not so good at others. We’re good pattern-matchers and meaning-makers, but not so good at doing things by rote. We make mistakes, almost by necessity (evolutionary advantage: if you do something a little different by chance and it’s better, it can get rewarded and more likely). And we simplify the world, partly to save energy for what we care about, but also because complexity is taxing.
And, in general, this is good. Our simplifications help us cope, make us more effective. However, given our nature, at times this can fail us. We may think we’ve taken a necessary step when we haven’t. Henry Petroski, in To Engineer is Human, helps us understand that we continue to push boundaries and take consequent risks. Atul Gawande, in The Checklist Manifesto, helps us understand the usefulness of support if we’re not going to make mistakes.
But sometimes this expediency can mask complexities and lead us astray. For a simple example, the term ‘learning management system’ can actually lead us to believe we’re achieving learning, instead of courses. And just because you have a course doesn’t mean something was learned.
There are many ways we can mislead ourselves. We can talk about a concept that we all realize has to be true, that learners differ, and then believe we can identify how someone learns. We may eventually be able to do so, but existing instruments aren’t valid, and learners change in different contexts. Plus, if we label learners as X or Y, we may limit them. When I humorously compared the ‘generational differences’ argument to age discrimination, someone deeply involved in that field corrected me that real age discrimination is a serious problem not to be taken lightly!
It may seem like an ‘angels dancing on the head of a pin‘ type of argument, but we have to be careful of the words we use and their import. We have to carefully consider the ways in which phrases can be used, or misused, and perhaps structure our use of language appropriately. It’s branding, and perhaps we need to treat it as such. At least, be careful of what terms you use and what inferences you’re making easy and which you might be inadvertently making hard.
It’s clear that our brains aren’t the logical problem-solvers we’d like to be. The evidence on our different thinking systems makes clear that we use intuition when we can, and hard thinking when we must. Except that we use intuition even when we shouldn’t, and hard thinking is very susceptible to problems. Yet we need to have reliably good outcomes to solving problems or accomplishing tasks. What can we do?
The answer, of course, is to use technology to fill in the gaps, when we can. We can automate it if we totally understand it, but the best solution is to let technology (and design) do what it can, and let our brains fill in what we do best. So, when there’s a problem or task that needs to be accomplished, and it requires some decision making, we should be doing several things.
To start, we should be looking at the scope of possible situations, and determine what’s required. We should then figure out what information can be in the world (whether a resource or in other’s heads), and what has to be in the performer’s repertoire. We want to design a solution system, not just a course.
Recognize that getting things into human heads reliably is problematic at best. It takes considerable work to develop that expert intuition: considerable practice at least. So the preference should be to design either a really good support system that helps in characterization of the problem, and a dialog that helps determine what of the possible solutions matches up with the situation.
It can just be information in the world, such as a job aid or checklist, or an interactive decision support tool. Or, it could be a social network of resources such as tools and videos created by others that’s usefully searchable and the ability to ask questions of the community and get responses. It’s likely a probabilistic decision here: what is the likelihood that the network has the answer, versus what’s the possibility that we can design support that will cover the range of problems to be faced?
The point is that support design is a necessary and very viable component of performance solutions, and one that isn’t being used enough. I’m looking forward to the upcoming eLearning Guild’s Performance Support Symposium in Boston as a way to learn more, and hope to see you there!
One of the things I’m wrestling with right now is the transition from where we’re at to where we should be. I’ve already said we have to change, the question is how. A number of us have talked about what the future could and should look like (including but not limited to my ITA colleagues), but how do you get there from here? Whether it’s 70:20:10 or some other model, there simply has to be a better way.
Right now, many organizations have training operations going on: people traveling to a location to sit down and have knowledge dumped on them, perhaps including a knowledge check: show up and throw up. With smile sheets to evaluate the outcome. And/or we have elearning: knowledge dump and test, or virtual classrooms where we do the same spray and pray. And it doesn’t work; it can’t. Well, it certainly is not going to lead to any meaningful outcome.
Granted, this is a gross stereotype, and in particular situations the training is hands on, there’s lots of practice, and maybe blended so the knowledge necessary is learned beforehand. Yes, this can lead to some real skill acquisition.
And don’t even bring up ‘compliance’. Sure, I know you have to do it, and until we get government to stop thinking seat time and content presentation is the same as behavior change, and lawyers to only care about CYA, we’ll be stuck with it. That, apparently, is just the cost of doing business. I’m talking about the things that matter to your company. Your skills, your ability to do.
And we need to stop it. I think the first step is to kill all training. When you raise your head up and start talking about “but how are we going to meet this need”, then you start thinking anew: how can we go ‘world first‘, performance support 2nd, and meaningful practice last? Once you find you’ve exhausted all other options, then you might realize that formal training is needed, but it could and should be focused on doing, not telling.
This is only part of the picture, and perhaps I’m being deliberately controversial (who, me?), but it’s one way to be forced to think anew about what we’re doing and how we can help. We have to change, and we need to start figuring out how to get there. I’m open to ideas. I’ve heard some others, but that’s another post. What are your ideas?
A few weeks ago, I posted that we should think social first. I want to amend my statement (I reserve the right to improve my recommendations :) to say “world first”, and clarify what I mean. Earlier, I posted about working backward and forward, and err on the side of putting information in the world first, and only put things in the head as a last resort (because formal learning is expensive).
What I’d suggested is that you should go to networks first, if the answer is out there. If not, you would try to use performance support (which, though not necessarily cheap, may be less expensive than formal learning in terms of time off task, etc). Formal learning is the recourse of last resort. However, I missed one element, which came up in a conversation.
The conversation had to do with not developing resources at all, except core ones. When someone wanted help with something, the option was to first try to point them to a video or book, or person. The goal is not to reinvent the wheel, reproduce resources, etc. Use the world first, and only pick up a secondary approach if the world doesn’t have it. In some sense that’s the social network, but it might be that the L&D department, in the course of their continual self-learning (hint hint, nudge nudge), would’ve curated a relevant resource, so it could come either from a pointer to a resource, or a question of others.
Then, the core content for an organization would be meta-learning: how to find resources, how to solve problems, etc. Something we were pushing about a decade ago. Ok, so you might also develop the values and mission of the organization as well. And of course proprietary processes, formulas, etc.
The focus, however, is and should be the Least Assistance Principle: what’s the least we can do for someone to get them safely back into their flow. The reality is we can’t meet all the needs in an era of increasing complexity, and we need to be more efficient and effective: giving good support but in the right modality. So, point to it if you can, develop the minimal support to move forward, and only put stuff into people’s heads when it absolutely has to be there. Fair enough?
At lunch last week with my colleague Jay Cross, we riffed on the most important word was for learning in the organization. I chose ‘meaningful’, he went with ‘experiential’. They’re both important, but I thought I’d tease them out a bit.
I take meaningful in two senses: what you’re doing (read: learning) is directly related to what your goals are, and it’s something you care about. So, for example, in designing serious games, you want to focus on key skills, not on irrelevant material. And, to work as engagingly as possible, you also are also choosing a context that the learner cares about. Taking something relatively abstract like coaching, you could be providing support for developing a coaching model (rather than using an existing one), or you could be figuring out how to help a person separate out person from behavior (commenting on the latter is to be preferred). Similarly, it could be in the context of being a better accountant (uninteresting except to accountants), or it could be for sports (which might be of interest to a broader segment). The point is to be focusing on relevant skills in interesting contexts.
Now, I take experiential to have some overlap, but to be addressing two senses as well: both the context you are learning in, and the nature of the learning experience. That is, you can be learning away from work, or in the work process, the latter being more ‘experiential’. And you can be learning by doing in either context, learning ‘how’, as opposed to learning ‘about’. I think there’s overlap here in being contextually relevant , but separate in the sense of personally interesting and the learning being applied.
I mapped it out, with lack of experiential and meaningful being disconnected content (which I see far too much of in workplace learning), or where we start providing knowledge how (not about) to be meaningful, providing activity-based learning to be experiential, and of course the ultimate being the intersection of both.
I’m sure Jay would argue that if it’s experiential, learning through real work experience, it’s inherently meaningful. And I’d argue that if it’s suitably meaningful, it naturally has to be experiential. Yet overall I’m happy take either one or both versus neither!
It used to be that the L&D model was to prepare people, then send them out to perform. There would be some data collection from the result, including debriefing perhaps, and then the training and personnel would be reviewed. In that sense, L&D was outside the performance loop, in a separate loop. And that made sense is a world where couldn’t do on the job scrutiny, and things were more predictable. We’re not in that world anymore.
The world we’re in is changing faster: we’re getting more data, companies can move faster, and customers expect more. And we now can have much more insight into what’s happening (and more’s on the way, courtesy of xAPI), and be much more closely coupled to performance. What does this mean we can and should do?
I think it means a different loop relationship with performance, where a second loop is integrated on top of the loop of performance. In this loop, L&D is more closely monitoring individual performance, looking for opportunities to support outcomes. L&D could be reviewing correlations between resource use and performance, finding those that are not performing as expected to remove or redesign, they can be looking at interactions to see how to facilitate, and they can be monitoring for emergent knowledge and skills that can be captured and possibly developed.
A concept from cybernetics is double loop learning, where you’re reviewing your goals as well as your methods, and it’s been a valuable contribution to thinking and action. Here we’re reviewing our approaches to a goal, which is a synergistic concept. And this is a role where L&D both gets more strategic in supporting business goals, more integrated into the operations by being more coupled to operations, and more facilitative in role by helping facilitate at the moment of need. This, I will suggest, is a possible and necessary shift to the ways in which organizations can start being more nimble, and the way that L&D can be directly responsible for that shift. Does this make sense to you? And is this something you think can be done?
A few posts ago I posited that formal learning should be expensive. Then, as a consequence of my subsequent posts on social first and performance support second, I was in a discussion where the question was shouldn’t performance support be first. It intrigued me, but my response was that while it’s not just about saving resources, using people is more appealing in general and given the expenses for developing performance support it should be a secondary response. So why do I say Performance Support (PS) shouldn’t be cheap?
First, there’s a lot that goes into good job aid design. I visited a bank many years ago, and was investigating their processes. During the course of it they mentioned a job aid, and after I examined it I pointed out that it did a good job of structuring the information process, and the employees I was talking to weren’t aware of how well it did. However, it was structured to support getting the information, but not structured for effective use. First, the employees were supposed to ask the questions by rote, which would lead to a bad customer experience. No one wants to interact with a person reading rote questions: that’s a role for tech. Yet There wasn’t a lot of highlighting to help focus on the important points, and gradually support personalizing the communication. It was graphically designed appropriately, with spacing and color, and informationally designed but not usably designed.
And that’s the point, there are nuances to usable design that you shouldn’t assume you can wing, so it takes time and skills to get it right. You need to do the right analysis, right information design, graphic design, and usability design. You may also need to do the right information architecture, so it’s findable both by search and browsing. If it’s not findable when needed, it’s not going to get used and won’t help.
Second, don’t assume you’re going to get it right the first time. As Atul Gawande said in The Checklist Manifesto, he iterated many times to refine simple checklists to support performance. You’ll want to test and refine iteratively until you’ve documented that the design works and yields the desired reports. Don’t assume (as we do with much elearning too) that ‘if we build it, it is good’. You need to test it in context, with real users, to see if it meets their needs. Really, the designer should ultimately be the user!
Yet often resources created by performers are useful. Why is that? Because they’re in context, and are creating what they need. It may still benefit from some support, but this shifts the L&D role to facilitation, not design. And this is good.
At the end, the more we can help people help themselves, the more overall help is being provided, so we’re optimizing our time and impact. And that’s what we want, right?
A number of years ago, I discussed a useful model that talks about how we solve problems in the world. In the piece, I talked about how when we can’t act, we try to find the answer (and if we do, we go back into action). Then if we can’t find the answer, we have to go into a problem-solving mode: we need to do research, experiment, and generally discover the solution. If we find a solution we should update the resources to help other people find the answer rather than having to rediscover it.
I was thinking about this in terms of the ways in which L&D can support this process, and started noting the ways in which we can help besides courses. I broke it up into two different forms of support: direct, and supporting the associated skills.
When we have an information need, we might need directories to people with various expertise (associated with Communities of Practice, presumably). We might design or curate useful information resources (how-to videos, job aids), and occasionally, when a significant skill shift is needed, we might design courses.
There are associated skills here, so communicating successfully with the experts who have the answer, or information literacy to develop the performer’s ability to find the answers themselves.
If the answer doesn’t already exist, then we might support learner with tools about problem solving, and research and problem-solving skills, as well as communication skills again to deal successfully with collaborators.
Finally, when the answer is found, might have tools to create resources and skills to edit existing resources.
First, this is by no means a complete list, as even in writing this I thought of design to create resources to support problem-solving, and information architecture to go along with information resources, and…you get the idea. The point is two-fold: we need to recognize how people actually act in the world, and we need to then find ways to support all the points of need, not just the ones we can design a course for. There are lots of opportunities!