[From Fred Nickols (971022.1702 EDT)]
Bruce Gregory (971021.2045) EDT) in response to Fred Nickols (102197.1800)
Fred:
I've not chimed in on the generic problem solving discussion because I
noted the initial reference to Polya and to teaching math problems.
<snip>
BruceG:
You may understand PCT, but you've got a lot to learn about CSGNet.
Discussions "center" on whatever you feel like saying. So out with it
Okay; here goes... >:-( <---- Howzzat for a grumpy face?
I consider a problem to be a situation requiring action and in which the
action to take is not immediately apparent. In short, a problem requires of
you that you figure out what to do (which is why one writer called problem
solving "What you do when you don't know what to do"). Figuring out what to
do is only part of solving a problem; it's what I call the investigation
phase. To actually solve a problem, you have to do something, you have to
intervene. Ordinarily, we call this "implementing the solution." I call it
the intervention phase.
Said a little differently, to solve a problem you have to change things.
Enter now the complex nature of the situations in which I most often find
myself charged with intervening; namely, the workplace. Organizations,
processes, functions, markets, and all the other "stuff" associated with the
workplace are labels for some pretty complex situations. Yet, it is these
situations in which we intervene to as to solve problems, especially those
known as "business problems." In these kinds of situations, change is
indirect; that is, you don't change it (e.g., profit), you change something
else and "it" changes as a result.
Given that change is necessary, and given also that change is frequently
indirect (at least in complex systems and situations), it is useful to think
of two points or places in the structure of the situation being
contemplated. The first is the Point of Intervention, and the second is the
Point of Evaluation. In simpler language, where will you intervene, and
where will you gauge the success of your intervention? These two "points"
are connected by way of the structure of the situation in which a problem
may be said to be embedded.
So, to solve a problem (or at least the kinds of problems I'm generally
saddled with solving), you first identify where and how you will determine
that the problem has been solved. These are the Points of Evaluation (POE).
Next, you try to define/draw/diagram the structure of the situation in which
those Points of Evaluation are embedded. Then, you go looking for places in
that structure where you can intervene and see the direct and immediate
effects of your intervention. These, of course, are the Points of
Intervention (POI). Thus it is that changes introduced at the Points of
Intervention ripple outward through the structure of the situation, making
themselves felt at the Points of Evaluation.
Generally speaking, I've found that there are only three basic kinds of
structures with which I need to concern myself: financial, operational, and
behavioral (the last, of course, giving rise to my interest in Fred Skinner
and Bill Powers; or, more precisely, in operant conditioning, and in PCT).
All this kind of comes together in a general problem solving approach that I
call "Solution Engineering."
One of the interesting (to me) aspects of this is that Solution Engineering
almost always requires working backward from the desired result (Point of
Evaluation) to the means of achieving (action taken at the Point of
Intervention). This flies in the face of a lot of research into solving
math problems, which almost always finds that experts work forward from the
problem to the solution and novices work backward from the solved state to
the solution. I think what that research is really finding is that if the
problem has a known solution, experts invoke it and novices try to figure it
out. On the other hand, if you present experts with problems that have no
known solutions, you make novices of them, and they, too, must work backward.
My point, then, is that many problems encountered in the workplace are, to
the people who encounter them, novel problems, with no known solutions.
They can, of course, cast about for people who know of solutions (these
people are known as "experts"), or they can try to work things out
themselves. If they choose or are forced to work things out themselves,
they must engineer solutions to the problems they encounter. I think the
process of figuring out what to do in novel situations can be taught.
Hence, I also think once can successfully teach generic problem solving
(although I don't think Polya had it quite "nailed").
There ya go, BruceG...
Regards,
Fred Nickols
The Distance Consulting Company
nickols@worldnet.att.net