Learning to Model LO5231

Richard Karash (rkarash@karash.com)
Tue, 30 Jan 96 16:59:30 0500

Replying to LO5141 -- was: Intro -- Maarten Sierhuis

Maarten, I face this question regularly when I teach systems thinking.
And, incidently, I do a middle management program in your company several
times a year.

I have been thinking lately about two different states of mind in which I
find people. There's a "pre" state which is (hopefully) different from the
"post" state after a program. When I'm successful with a group, and I
usually am, these are some differences I see in the "pre" vs. "post"
states. What I'm trying to get at here is to understand the essential
element that can make learning to model a "transformative" experience.

I see/hear people in the "pre" state:

- asking for a *method*. "Just tell us the steps, what we have to do, to
get a good result."
- judging the model by the criterion "Does it match reality" with the
implicit addition "completely." (Generally they are looking at "face
validity", does the structure match their view of the world, does it make
sense, not whether conclusions or predictions match the observed
behavior.)
- asking, "Is my model *right*?"
- asking, "How do I *know* what to put in my model? How do I *know*
whether this part of the model is OK?"
- stuck at each step; concerned they might "have it wrong"; looking for
confirmation at each step, "Is this OK?"
- stuck on questions such as, "How do I know when I've put in enough
things? What if someone else's model has it differently; don't we need to
reconcile them?"
- very nervous about presenting their work to others
- uncertain of the rules and roles in a group process of model building
("Are we looking for 'truth' or to see who's 'right' or to see who
'wins'?")

I see/hear people in the "post" state:

- noticing that any model is someone's assertion of a theory of how
things work in a certain area.
- willing, interested, curious, and ready to make their own assertions,
"Well, I think it probably works like this..."
- judging the model in terms of it's suitability for some purpose
- open to have their models/assertions examined, tested, and enhanced in
conversation with others.
- testing their model both in terms of face validity and also in terms of
its conclusions/predictions vs. observed behaviors
- using data appropriately (observed data can "reject" a model, can never
"prove" a model, but several instances of matching data can "support" a
model
- able to have better group conversations in a group model building
process
- accepting that two different people might have two different models of
the same area; that these might reflect different purposes, different
views, or different mental models
- able to use the differences as a launch pad for productive reflection
about own's own mental models and discussion about these
- curious to keep testing and refining their model over a period of time;
accepting it as "helpful" while still seeing it as "provisional"

There's also some much deeper material which is often part of our work,
but may not be relevant to yours (root cause vs. structural determinism
for example).

I'd like to understand better the essence of the difference between the
"pre" and the "post" state, and I'll stop here to see what you or others
might add at this point. Later, I can comment on the methods I use to get
from "pre" to "post."

Maarten wrote:
> subject: Learning to Model
>
> I am starting some research in which I am interested to investigate
> the
> problems in having a group of workers use a modeling paradigm (such
> as
> data flow modeling, object oriented modeling, or business process
> modeling) to create models. The experience we have had is that it is
> very
> difficult to 1) teach people the modeling paradigm (language +
> techniques), and 2) after they have learned the language and
> techniques,
> have them applied it appropriately.
>
> Maarten Sierhuis NYNEX Science & Technology, Inc.

--
      Richard Karash ("Rick")    |  <http://world.std.com/~rkarash>
  Speaker, Facilitator, Trainer  |     email: rkarash@karash.com
"Towards learning organizations" | Host for Learning-Org Mailing List
(617)227-0106, fax (617)523-3839 |     <http://world.std.com/~lo>