REPLY Models, Problems and Systems (SD7042)

[SDMAIL John Morecroft]

Posted by  "John Morecroft" <jmorecroft at london.edu>

Bill
Thanks for your reflections (quoted below) stemming from system
engineering.  Yes, parsimony is a good word to describe the convergence
and focus that modellers ultimately seek.  However the advice to
iteratively model the dynamical problem situation is probably still the
most reliable way in SD to achieve parsimony..... plus lots of practice
with illustrative models and real world situations.  

------------------------------------------------------------------------
--
"Coincidentally, I was thumbing through my Fink and
Christiansen Electronic Engineer's Handbook (Second Edition) recently
and found their chapter on Systems Engineering.  Section 5.12 is called
"Parsimony and Validation," and I think the parsimony idea may relate to
this issue.  While we wrestle explicitly with validation from time to
time (there's a chapter in Business Dynamics devoted to the subject), I
wonder if we condense the idea of parsimony into "Model the problem, not
the system."  Parsimony sounds at once more artful and more craft-like,
something emerging from experience at creating models that err on one
side or the other".

_________________________________

John Morecroft | Senior Fellow
Management Science and Operations
Posted by  "John Morecroft" <jmorecroft at london.edu>
posting date  Wed, 14 May 2008 08:58:16 +0100