Responding to Barry Mallis in LO28750 --
Note: I've changed the title of this strand of the thread from History of
Uncovering the Act Learning to Learning from Solving Problems because that
seems a better fit with the direction this strand is headed. I've also
snipped a lot from Barry's post because I want to respond to the central
notions, not the detail.
After recapping the highlights from Barry's post (done below in snipped
fashion), I'm going to provide a description of a problem I once solved
that I believe is a very good fit with what Barry describes. However,
I'll let Barry be the judge of that.
Barry, responding to Dileep in LO28747, begins his post as follows...
>Dileep's comments stimulate my memories of working with W.J.J. Gordon and
>Tony Poze, the creators of "synectics" and founders of SES Associates in
>Cambridge, Massachusetts.
snip...
>Bill Gordon said to me over and over again that all ideas contain a
>paradox.
snip...
>Paradox lies at the
>core of a problem. In fact, a conflict that requires resolution is the
>very thing that makes a problem a problem.
I take a slightly different view. Paradox might well lie at the core or
heart of a problem but it is the not knowing what to do about it that
makes a problem a problem. As one wag wrote many years ago, "Problem
solving is what you do when you don't know what to do." If it is
uncertainty regarding the action to take when action is required that
makes a problem a problem, then problem solving (as I also happen to
believe) is a process that serves to reduce uncertainty. Further, because
action is required, that is, you must actually do something to solve a
problem, there are two basic phases of problem solving: (1) analysis (or
investigation) and (2) action (or intervention). More on this later.
snip...
>"The first step in the SES synectics process is to identify the inherent
>paradox within a problem. This step is mostly tough-minded examination of
>specifications with only an implied metaphorical component. After that
>begins the metaphorical process for developing a new context by which to
>view the problem. It is this new context that yields the creative, new
>viewpoint and the SES synectics process constitutes the step-by-step
>procedure for accomplishing this. The first new-context step is to
>generate an Analogue, a thing or situation that parallels the Paradox. In
>SES synectics, this is brought about at a purposeful, explicit level.
>
>"Before SES research identified certain explicit skills, Analogues were
>uncovered by accident.
The problem solving case I'm going to present is exactly that; a problem
in which insight (an accidentally discovered analogue) played a key role.
How that insight was triggered is kind of humorous (at least to me).
snip...
> "If you review each of the historical examples, you will notice that only
>a central aspect of the Analogue actually was used. That essential
>function of the Analogue is called its Unique Activity.
snip...
>"In the Equivalent step, the problem is considered in terms of the
>Analogue's Unique Activity.
Interestingly, I only articulated the "analogue's unique activity" just
now, many years after solving the actual problem.
snip...
>Therefore, in an Analogy Equation the
>Paradox must describe the Analogue and the Equivalent must parallel the
>Unique Activity of the Analogue. In the potato chip example, potato chips
>that would conform and stack so as to take up less space when packaged
>(the Equivalent) paralleled the way wet leaves conform to take up less
>space (the Unique Activity of the Analogue).
>
>"The four steps of the Analogy Equation--Paradox, Analogue, Unique
>Activity, and Equivalent--form the heart of the the SES Problem-Solving
>Worksheet....It constitutes the sequence for moving from one step to the
>next as you attack a problem."
I have for many years now been meaning to dig into W J J Gordon's
synectics and Barry's post finally prods me to do so. You will see why
shortly.
Now, on to the case in question.
About a year after joining Educational Testing Service (ETS), my boss, the
VP of Operations, walked into my office and posed the following problem:
"Given an array of rooms of known or determinable capacity, and
given a set of rules regarding the number of students and number of
proctors per room, is there an algorithm that will allow you to determine
the optimum (i.e., least costly) staffing level for any and all test
administrations?"
I replied confidently enough, "Sure."
My boss wanted to know how I could be so certain.
"There has to be," I explained, "there's nothing ambiguous in the
situation."
I then set off in search of the "algorithm" (by which the VP
meant a simple procedure, not necessarily an actual algorithm). Several
approaches were taken, none of which panned out because we were always
able to come up with a real test administration situation that the
proposed procedure couldn't accommodate.
Then, one day, a couple of weeks later, I had occasion to review
some sample test items in a test bulletin. One of them involved being
given a specified amount of liquid in a pitcher and then distributing that
liquid amongst several smaller containers in a way that would satisfy
certain given requirements for each container. I almost shouted "Eureka!"
There was my staffing problem albeit in a different form. The problem of
filling up those smaller containers so as to satisfy certain capacity
specifications was essentially the same as the problem of filling up an
available array of classrooms with a given number of test takers so as to
minimize staffing costs. I won't go into the actual test administration
solution here because my point is to support Barry's notion of the
important role played by analogues in solving problems in a creative
fashion.
The resultant "algorithm" (actually, a simple little five-step
procedure) led to annual savings in test administration costs of $800,000.
Over the past 10 years, that's amounted to a pretty good chunk of change.
The person in the testing program for which the solution was developed was
positively ecstatic and, over a celebratory dinner, he told me that that
particular problem had stumped the company for many years. He, of course,
wanted to know how I had done it and I had to confess that it owed to a
flash of insight (an accidentally discovered analogue).
My boss was quite pleased, too, but for a different reason. You
see, the reason he wanted the "algorithm" was to resolve tension between
two competing approaches to reducing the cost of test administrations.
One camp wanted to provide the test center supervisors with monetary
incentives to reduce staffing levels. The other camp thought that giving
the test center supervisors monetary incentives to hold down staffing
costs would lead to understaffing and, consequently, jeopardize test
reliability owing to non-standard conditions. One solution to this
conflict was an "algorithm" that would enable the test center supervisors
to determine the minimum staffing level that would also satisfy test
administration requirements. This "algorithm" entailed interaction
between and among (1) an array of available rooms (always a variable, in
terms of number and capacity), (2) the number of test takers (also a
variable), (3) the room seating requirements and (4) the staffing
requirements (these last two were givens). Needless to say, the solution
to the problem he gave me was also the solution to the problem with which
he was wrestling. How's that for "parsing" a problem? He was a very
smart fellow.
Thanks again to Barry for a post that struck a chord with one of my better
experiences. I've certainly learned something from it and I hope others
might too. My questions to Barry are these:
"Might a knowledge or application of Synectics helped me discover
the pitcher-liquid-container analogue (or a similar one) in a structured,
disciplined way instead of fortuitously stumbling across it?"
"Where is the 'paradox' in what I have described?"
Regards,
Fred Nickols
740.397.2363
nickols@safe-t.net
"Assistance at a Distance"
http://home.att.net/~nickols/articles.htm
--Fred Nickols <nickols@safe-t.net>
Learning-org -- Hosted by Rick Karash <Richard@Karash.com> Public Dialog on Learning Organizations -- <http://www.learning-org.com>
"Learning-org" and the format of our message identifiers (LO1234, etc.) are trademarks of Richard Karash.