Tony Karrer's eLearning Blog on e-Learning Trends eLearning 2.0 Personal Learning Informal Learning eLearning Design Authoring Tools Rapid e-Learning Tools Blended e-Learning e-Learning Tools Learning Management Systems (LMS) e-Learning ROI and Metrics

Friday, December 01, 2006

Significant Work Needed to Help Instructional Designers

The November LCB Big Question was "Are ISD / ADDIE / HPT relevant in a world of rapid elearning, faster time-to-performance, and informal learning?" I was planning to try to write up a summary at some point (similar to last month's eLearning Technology: Top Ten Reasons To Blog and Top Ten Not to Blog), but I must say this month it was a lot harder. My personal summary is very briefly (more below):

  • There’s general consensus that ISD, ADDIE, HPT provide a good foundation, but that they need to evolve.
  • There’s little practical advice that can assist Instructional Designers in what the evolved forms really are.
  • There are a few skeptics.

And I find that second bullet especially troubling given that the crux of the question as Mark Oehlert from e-Clippings put it:

If the question of whether or not your job is relevant fails to stir up some considerable heat - then I despair for us.

Well Mark start despairing for us...

I’m beginning to wonder if we are expected to just wing it until we become marginalized. As a Forbes article stated –

One thing is clear: In two decades, your job probably won't exist, at least not in the same form.

So let me step back and give some foundation for my summary. Oh and before I go any farther, I should point out that, as is always the case, there is general consensus that as Jay Cross put it:

This is the wrong question.

And as Russ Crumley put it:

we’ve already answered that.

So, again, Dave Lee and I are continuing our consistency in asking bad questions.

Also as background, there were quite a few people who said, ISD/ADDIE/HPT – huh? That was actually quite a big surprise to me. Maybe they are already irrelevant?

ISD, ADDIE, HPT are Good Foundations

There was general consensus that they provide a good foundation for the future.

Phil Charron:

More models will emerge, but they'll only be improvements on existing models. We'll always take shortcuts, but it's important to know what the original route was in case you get lost.

Harold Jarche:

SAT, ISD and ADDIE are excellent methods to develop training that is stable ... The Internet is forcing us out of our self-constructed disciplinary boxes. ... As work and learning become connected online, the barriers are blurring between organisational development, HR, training, education, HPT, etc. A new, amalgamated field of practice requires better tools and integrated theories from which to base our practice.

Russ Crumley:

The challenge we now face is not the creation of more tools, but better use of the tools we have, while applying the fundamentals of human behavior and performance, including the role of emotion, curiosity, discovery, and the desire to improve.

Dennis Coxe:

The role of the instructional designer will be to use the ADDIE model to determine what baseline structure can be built into the formal piece of the learning and what parts of the knowledge base are fluid and need to maintained delivered in an informal venue, be it a blog, a podcast, or talking points delivered by a project manager to his or her team.

Karl Kapp

We need these models now more than ever ... Would you ask a builder to skip the design step for a building? No need for the architect, just build the building.

Clive Shepherd

We need to be reminded of this discipline to avoid jumping to solutions, without having a proper understanding of the problem.

ISD, ADDIE, HPT need to Change

There was also general consensus that while they provide a foundation, they need to change.

Jacob McNulty:

In a world where products, targets and strategies adjust constantly the application of these models (ISD, ADDIE, HPT) will be greatly diminished. As the roles of knowledge workers expand and require them to use information that changes quickly it will be more important for the workforce to have easy access to information rather than them be required to retain it as the result of a well-designed course.

I use the term ‘greatly diminished’ because I don’t feel the models will become obsolete or completely irrelevant.

The models listed in this month’s Big Question were designed for a type of training that was relevant for the needs of an environment different from today’s.

Geetha Krishnan

my argument: A, D, D, I, and E are relevant. ADDIE may not be

Karyn Romeis

some method still needs to be applied to decide what features to provide and the look and feel of the vehicle -if we are completely without strategy, we are likely to wind up with a resource like Homer's car

Little Practical Advice

While the consensus was that the models were changing, I found myself wondering what that really means in practice. How will they change? What do I need to do differently? And I didn’t see a lot of advice. Here's a few of the things I saw...

Karl Kapp

You can do the model in an abbreviated format. For example, for the analysis phase of ADDIE, hold a one hour focus group. Yes, one hour for analysis. Now, will you get the best analysis in the world? No, but you will get some insightful information. Instead this step gets skipped, even just doing a one hour analysis can save learners hours of time in terms of focusing the instruction.

Tony Karrer

  • Figure out Rapid HPT, ISD, ADDIE
  • Increase the Breadth and Improve Your Understanding of New Models / Tools
  • Become Meta-learning Experts
  • Learn to become guides, aggregators
  • Anol Bhattacharya:

    Don’t work in Silos – integrate with learners, produce rapid prototypes - the content development team work in silos and deliver the final product to the ‘client’ for evaluation.

    Clark Quinn

    Without knowing what we’re trying to achieve, metrics that let us know how we’re doing, and ownership of the outcomes, among other things, you don’t have any idea what you’re doing.

    Anil Mammen

    the problem lies not with the models, but in how we approach them and what we take out of them. models like ISD, if used creatively, can help produce highly effective learning programs.

    Clive Shepherd

    I would recommend anyone interested in the design and development of learning experiences to explore the various models relating to instructional design and to develop their own views on the strengths, weaknesses and general applicability of these models.
    Jay Cross
    Think like a great chef. - Great designers use the models as a great chef uses a recipe.

    Skeptics Among Us

    There were several notable comments that suggested that we should move away from these models and these terms:

    Wendy Wickham:

    Even in the classroom these models disconnect from reality ... The longer I work, the more I realize that the old roles of "Trainer" and "Instructional Designer" are becoming irrelevant.

    Clark Quinn

    training and instructional design don’t cut the mustard ... Keeping the labels is not good ... It’s not going to fall into a course or job aid.

    This Leaves Me With More Questions Than When I Started

    I’m questioning whether we really know what the next generation of ISD, ADDIE, HPT really looks like. I’m being told to not just do D & I, but I’m also not given time to do A, D, E like the models suggest. I’m told to at least spend an hour with the learners or in a focus group. Is that enough? How do I know if that’s enough? When should I demand more?

    The kinds of interventions I’m using are changing. How does informal learning change what I do in A&D? How about in E? How does Rapid Learning change it?

    I’m personally surprised at the lack of depth in the responses. I’m also surprised that some of the real visionaries such as Allison Rossett doesn’t have better suggestions. Do the experts really just think we should learn the models as a good foundation and wing it from there? That is what we are doing, but is that it?

    I hope this isn't the end of the discussion.


    Harold Jarche said...

    The old ISD looks like the Waterfall Method in software development, so I wager a guess that a new method might look more like Agile programming.

    Thanks for the synthesis of this conversation, Tony.

    Tony Karrer said...

    Great insight Harold! Especially for a software process person like me.

    Agile (in my mind) is rapid iteration over successively larger solutions. Does that hold here? Do we make a small, fast version of a solution and see if that works and then modify it? In some ways that's what we are doing, its just that we never end up going back and changing it - no one ends up asking for more features.

    Wow, it just feels like there's a whole world to be figured out here.

    Anonymous said...
    This comment has been removed by a blog administrator.
    Dave Lee said...

    I agree with Tony. Great insight, Harold. Regarding the Waterfall Methodology, Anol Bhattacharya give a great analysis of ADDIE as Waterfall Methodology in his post. Hit the link in the list of participating posts and check it out.

    Agile programming is a great model for us to consider for development of learning content and experiences moving forward. My experience with Agile was that the cornerstone of success is building in explicit and regular feedback loops into the content. feedback from users, the sales team, various project teams, SME's, technology thought leaders, the receptionist, etc. The coordination necessary is tremendous and it takes a strong project manager to handle the volume of feedback, but when it works, it's powerful.

    As an alternate model, Anol suggest Michael Allen's Successive Approximation process.

    Karl Kapp said...


    While the world of ISD, ADDIE and HPT is in flux, it is NOT standing still. There are a number of alternatvies and a lot of thinking going on in this area. I have create a post called Alternatives to the ADDIE Instructional Design Model which lists some work people are doing on alternatives to ID.

    One of the people doing the most work is M. David Merrill who has developed a number of models that are extremely effective for developing instruction.

    One of the things that the field needs to do is work to bring research and practice closer together. Will Thalheimer's working on that process.

    So, there are some deep thinkers working on the issue of the relevance of our models and the development of alternatives.

    The scary part, I think, is the lack of knowledge of the current models as well as the lack of recognition that alternative models exists and are viable.

    As the field continues to evolve, our thinking needs to evolve from simple linear models to non-linear models and to models like successive approximation or agile or rapid development.

    However, unless the basics are known, then we can't pull off a Jackson Pollock. Pollock pulled off his "art" because he understood what the rules of art where and then he consiously decided to break them.

    Designers must know the rules of good design and then consiously break them. You can't go into it without knowing the rules and, from some of the responses, teaching the basics might be a good place to start.

    Wendy said...

    I wonder if we are seeing the beginnings of a paradigm shift - how often do we have to break the model to realize that it may not work?

    From what I can tell, fast development is not the only thing being asked of us - fast results are too. How quickly can we teach so that the "student" can put the material in practice?

    The increased demand for customization, the decreased authority of the "expert" and the expectation of "just-in-time" information (never mind learning) makes me think we need to keep the focus of the model on rapid learning rather than rapid development.

    I'm not entirely sure that Agile programming, successive approximation, or any other developer-focused model fits the bill. It may help us build faster, but are we really being more effective?

    The current learner-focused models (cite by Karl Kapp in a previous comment) in practice require more time and attention than people (and organizations) are willing to give.

    I wish I had a tidy model that I could hang onto for a solution. I haven’t found it yet.

    Thanks for the summary and support.

    Tony Karrer said...

    Good comments. I'll have to take a deeper look at some of the alternatives cited such as successive approximations and the 4C/ID which looks interesting.

    Karl its interesting to me that I had never even heard of the 4C/ID model.

    Further, if the reality is that rapid development is a common pressure in corporate environments, why is there not more around this topic?

    Wendy, your comment is quite interesting. I would agree that there's need on both ends. The focus seems to be rapid development rather than rapid learning / rapid performance. You would think that there's great opportunity around this especially given the lessening expectation of knowledge transfer as opposed to information look-up.

    Does it seem strange that there's not more discussion around this or have I just been missing it?

    Mark said...

    Tony - I have absolutely had this post open on a Firefox tab since you posted it. I think the post and the comments should be used as instruction in all the college-level ISD programs.

    My 2 cents come from my inability to tell my history grad student genes to keep their mouth closed. I think that if we are looking at industry folks to correct ISD then we are looking to far downstream. The answer I posit lies all the back in the undergrad and grad programs and that means a big part of the solution has to come from re-engineering the way professors who teach ISD understand the field and consequently change the way they teach it - otherwise we will just keep having this discussion every year as new grads enter the marketplace with ideas that we all agree need to change.

    I feel a full post coming on put just want to say that I think that tone of the most important skills I learned as a history grad student was to question everything - especially those who held themselves out as authorities. That doesn't mean disrespect everyone but rather know your theories, know your facts, know your context and demand the same of everyone else. This discussion we are having on your blog needs to going on in faculty boardrooms across the country/globe. We need to teach students theory about learning,cognition, neurobiology, sociology, economics and toss in some Senge and Argyris just for kicks! They need to be masters of 1000 disciplines...enough for now...will post longer more coherent version on e-clippings - thanks for starting this up though.

    Tony Karrer said...

    Mark - it's interesting that you feel this will/should come from the academic side of things as opposed to evolve out in the real-world (probably the consumer world) and then have us figure out the patterns and structures and formalize it. I feel a post coming on as well.

    Mark said...

    Tony - I hear you - couple of things (actually me just using your comment section to organize my post :-))

    #1 I think we need a pretty good review of the capabilities that people in this field need to have at the genetic level.

    #2 That those capabilities include a broader, more interdisciplinary theoretical background and a stronger cultural orientation (as a field) to an ongoing critical review of that theoretical base. Those sound to me like pieces that need to be part of a collegiate degree program.

    #3 That a "real world" focus is a must but that it must be mediated through a different lens than we have now. I have this real feeling of "if all you have is a hammer, then everything looks like a nail" right now. I think that people need to be born into this career with the greatest background possible so that we can be masters of the improvisation - be like the rodeo clowns who are actually the best rodeo riders and are sooo good that they cam improvise, on the fly, and still make people laugh. You can't do that when all you teach people is which model to select.

    I hope this doesn't sound like I'm yelling at anyone - this is just how its spilling out right now.

    Tony Karrer said...

    I have some of the same feelings about the necessity of handling great variations in structure. At the same time, the skills of looking at how people are doing things today and extracting from that is still going to be hugely important. There's still a lot of the analytic stuff that's needed.

    And where's that post? :)

    Mark said...

    OK...full post it here

    Camille B. Price said...

    So it's 3 years later and I'm still hearing the same argument go around and around. "IDS takes too long." "ISD is outmoded." And yet, it seems to me like people are chasing their tails and really don't see what ISD IS.

    Instructional Systems Design is a domain of knowledge the purpose of which is to increase the EFFICIENCY and EFFECTIVENESS of instruction. It is composed of PROCESSES (ADDIE, Dick & Cary, et. al.), PRINCIPLES, and STRATEGIES (both macro and micro). Regardless of whether the formal ISD process is used, ISD principles and learning strategies still have tremendous value and folks tend to just move right on by the relevance of that body of knowledge in the zeal to draw and quarter the ISD process.

    But, back to ISD process. ADDIE was originally derived as a descriptive model and only later made into a prescriptive model. In other words, if Gandolph 2867 decides that there's a better way to do the "Toad to CEO" spell (analysis), then types it up (design, develop), posts it on the Goober game forum (implement) and Frodo 23468 says "That sucks" (evaluation), the world has just experienced ADDIE.

    However informal or minimal, an instance of an instructional development model has occurred. I don't understand why folks don't get this. Formal or informal, intentional or unintentional, ADDIE happens. To me the issue is not one of "should we ADDIE," but rather HOW WELL.

    Systematic development processes evolved in the industrial age as a way to increase the potential for successful outcomes. In other words if a very rigorous, high quality process is used there is a higher probability that the outcome will be successful. If one decreases the rigor of the process the risk of a less effective outcome increases.

    There are situations where the time required to implement a highly rigorous process is not justified. And there are times when it is. So the real questions are, "How much rigor does a particular training need justify?" And, "How can we optimize the process for the different levels of rigor required?" Not, "Is ADDIE dead?"

    The contemporary approach to assuring quality outcomes appears to be different. Rather than relying on rigorous processes, the approach seems to be to throw out a lot of stuff and use the wisdom of the masses to filter the best stuff to the top.

    In the final analysis, I don't think this approach is really much more time efficient, but it does get the chum pool churning sooner and builds excitement.

    Personally, I think there's a place for both approaches and too many people are too anxious to throw the ISD baby out with the bathwater.