QUERY Models, Problems and Systems (SD6995)

[SDMAIL Monte Kietpawpan]

Posted by  Monte Kietpawpan <kietpawpan at yahoo.com>

PROBLEM

However novice modelers carry out SD projects,
they will rarely get simple models. This is a big problem
when they should build a simple model.
The problem has been discussed in the recent issues of
SDR, but its root cause has not yet been clearly stated.
As far as I see, the problem is caused by the adoption
of conventional modeling processes proposed in
classic SD texts. Following these texts, modelers will
always define the model purpose BEFORE they
build a model. And by accepting the ultimate purpose
of SD (i.e., to improve the system performance),
they will set similar purposes and will conceptualize all
components considered relevant to their purposes.
Such a model purpose reflects an implicit assumption
that modelers are either managers or managers' consultants
who are interested in solving the existing dynamic
problems that jeopadize the sustainability of the systems
they are in. They are not scientists who are free to select
issues to investigate and to publish their investigation
results in journals. They are not those who should build
simple models that are easy to develop, to write, to review,
and to publish quickly.

MY SOLUTION
To get a simple model, I altered the SD paradigm by
inversing some initial steps in the SD modeling process.
The model purpose was defined only AFTER an initial
model was developed. This process led to the simple
model (first-order system) circulated via
http://www.gotoknow.org/file/philosophist/P-003.pdf
 
Here, it is essential that we ask not what structure
might address the stated model purpose, but ask
for what purposes a given simple model can serve.
In this way, I am more like a scientist who builds
a (simple) theory looking for concrete application than
a manager who identifies concrete problems looking for
(usually necessarily complex) theoretical formulation.

Regards,
M. Kietpawpan
Posted by  Monte Kietpawpan <kietpawpan at yahoo.com>
posting date  Wed, 7 May 2008 02:00:19 -0700 (PDT)