In my preparations for writing yet another book, I found myself rereading an article I had published with ISPI back in 2002:
“lean-ISD — Designing for the ISD Life Cycle” – ISPI’s Performance Improvement Journal (August 2002).
The “trigger” spurring me to write this (in 2001) was that Training Magazine (April 2000) had an entire issue addressing the problems with ISD – Instructional Systems Design: The ATTACK on ISD.
I addressed this last year – but my book writing efforts – and some posts from others on Social Media have me “at it” again.
I had added my business colleagues’ names to that original ISPI article after I had them review and provide me feedback. Subsequently, I removed them for republishing it in my quarterly newsletter.
But the Attacks on ISD continue.
This must be my 4th or 5th rodeo, as that saying goes.
Training Magazine raised the complaints about ISD to another level of our consciousness in their
April 2000 and February 2002 issues. To summarize the complaints that they brought forth and
our response to each, the attack on ISD is about:
ISD is too slow and clumsy to meet today’s training challenges
Yes, ISD’s pace is glacial in an Internet world demanding speed and adapting to constant
change. Statements like “the analysis itself will take a month and a half” make our
clients and critics lose patience. But ISD can move quickly, deliberately and systematically.
Our approach, and we are sure others,’ is very visible, predictable, repeatable, and
systematic It is “lean.”
There’s no “there” there
This questions whether there is an instructional “technology” for training in the first
place, because too often people have learned from “stuff” that was created in processes
that didn’t follow the ISD-ADDIE model. We disagree. What did they “learn?” Did
they become aware? Were they entertained and slightly enlightened? Were their expectations
low in regard to knowledge or skills to be transferred? Do you want your airline
pilot or surgeon to be taught in a non-structured, non-systematic approach?
Used as directed, it produces bad solutions
Yes, too often ISD begins without a business purpose in mind, and therefore can be applied
poorly. Or it overreacts to a fraud, like designing for “learning styles” (a concept
easy to like but thoroughly debunked by actual research) resulting in wasted effort and
time. Or it breaks the learning process into ridiculously tiny increments and forces unnecessary
exercises and assessments.
It clings to the wrong world view
Training Magazine’s article suggests that ISD arrogantly assumes a “stupid learner” that
needs constant handholding in learning anything, and then designs instruction to the
lowest common denominator. But that’s if the “product” was intended to teach to the
lowest common denominator, either because that’s where the bulk of the learners were,
and/or the enterprise simply couldn’t afford multiple versions, or the ISD’er didn’t know
how to chunk it and create multiple entry points in the ‘learning process,” or the deployment
method wouldn’t allow for that. We don’t think that ISD clings to the wrong
While we disagree with most of these blanket statements, we know there is some truth in these
for many of the ISD approaches we’ve seen in action, or seen in the results thereof.
Those complaints in “The ATTACK on ISD” resonated with us too, because we’ve heard them
before. Other similar issues brought to our collective attention by meaningful ISD customers
over the years include:
- Content of the product line elements (courses, CBT, OJT programs, etc.) may be redundant across programs while still leaving critical gaps in other important content
- It is costly to produce the T&D in the first place, and even more costly to maintain
- T&D is costly to deploy
- It is impossible to predict development schedules and costs and then predict return on investment (ROI)
- The look and feel of the T&D varies across the product line, and chunks of potentially shareable T&D aren’t designed with reuse in mind.
Again, we agree with much of what’s been generalized about the majority of ISD methods. But this attack on ISD presumes that there is only one ISD model/approach being used. That, of course, is ridiculous! In a department of 10 ISD’ers, we too often have encountered 10 different ISD approaches in use. These varied ISD approaches are typically not predictable in terms of the quality of the T&D outputs produced, or their costs and schedules, and they are not in control. They are often not very visible processes for T&D management or for T&D customers. And therein lies the rub. Too much variation in the ISD processes being used, a bad thing we all should have learned from the ongoing global quality movement.
But we think that there is even more to complain and worry about than just these.
Blanketing versus targeting ISD efforts – Too often the focus is on providing T&D
opportunities for everyone. ISD efforts and resources are often wasted on low-value
projects, with little chance for significant ROI for the shareholders.
Performance Impact – Performance is often understood in the most generic terms, perhaps
driven by a generic competency model.. Generic models cause ISD’ers to create
generic products, with little chance at real impact back on-the-job. Communications
skills, or presentation skills, or problem solving skills apply very differently for shop
floor workers, their bosses, the sales force, the process engineers, the ISD’ers, and the
company lawyers and accountants. One size-fits-all products don’t have much impact
compared to targeted content (with perhaps some shareable components/objects). The
costs of lost opportunity of really impacting on-the-job performance because the content
and design did not focus ultimately on someone’s real job performance requirements
can be significant.
Reuse of content –Too often instructional content is not designed to increase sharing
where appropriate, and for non-sharing when unique content is needed. Even in multiple
targeted communications skills training products for varied audiences there are
common content pieces/chunks/objects. The costs for not improving reuse capability d
can result in significant additional costs to the enterprise. Imagine if your car didn’t
share any components with the cars built by your manufacturer; the cost to produce
your car would be significantly higher. Remember the “platform” design approach that
helped save Chrysler in the 1980’s?
Development – The costs for developing are artificially too high due to a lack of available,
or reluctance to use, standard but flexible rules, tools, and templates, and to employ
a rationale reuse strategy and approach. The end result can be redundant content
that will cause higher “first costs” than necessary and will lead to higher “life cycle
costs,” some of which are explained next.
Inventory – The costs for storing and retrieving content are too high due to lack of a
rational, logical “dewey decimal-type system” for products and their sub-assemblies,
much like the bar coding SKU (stock keeping unit) schemes in place everywhere in our
daily personal lives. If content exists within your current, total product line, can anyone
find it quickly for reuse or maintenance?
Administration – The costs are too high for communications/marketing, registration
and scheduling (for those T&D products needing to be scheduled) or ordering (for those
T&D products that need to be ordered) because the product line of T&D for any target
audience is overlapped, gapped and a mess in general and hard to present as a unified
system of instruction?
Deployment – The costs to deploy the T&D are often too high given the probable returns;
and recently when the cheaper, total “e” learning strategy has failed to produce
results for many buyers, we now find ourselves back to a more blended approach, that
still too often focuses on low hanging fruit content that still won’t move performance
levels higher at an adequate ROI.
Maintenance – The decentralized ISD systems and processes that typically exist, including
the lack of design rules and tools, and the lack of a rational inventory scheme, will
drive up the costs for keeping content up-to-date. But if the content isn’t really improving
performance anyway, maybe it’s better left hidden with the hope that any subsequent
effort may get luckier; just don’t share that with the shareholders.
There’s more to the article – here – and my response led me to create a session presentation for the next ISPI Conference in 2002 – which was then brought back in 2003 for an Encore Session after being rated in the top 5 sessions for 2002. You can see the set of slides from 2003 here.
Design for the Life Cycle – ISPI – 2003 – 40 page PDF – delivered at the ISPI conference in Boston in April 2003. This covers 7 value variables that need to be addressed to improve the ROI of a training/ learning function.
The updated model – for one of my next books, follows.
The battles continue.