Ventana software support forum

Hi,

You could read Professor Sterman's work called ""A skeptic guide to computer models"".

Here you can download it. www.millenniuminstitute.net/publications/Skeptics.pdf

Be well...

Fabian Szulanski

Posted by Fabian Fabian <f_fabian@yahoo.com>
posting date Sun, 23 Jan 2005 06:49:30 -0800 (PST)

Posted by John Morecroft <jmorecroft@london.edu>
For an insightful philosophical comparison of system dynamics and
econometrics, still highly relevant today, I suggest you read Dana
Meadows' article:

Donella H. Meadows (1980), The Unavoidable A Priori, chapter 2 in
Elements of the System Dynamics Method, (Randers editor), MIT Press.

John Morecroft
Adjunct Associate Professor of Decision Science
London Business School

Tel +44 (0) 20 7262 5050 x3252 Fax +44 (0) 20 7724 7875
jmorecroft@london.eduPosted by John Morecroft <jmorecroft@london.edu>
posting date Mon, 24 Jan 2005 12:16:28 GMT

Posted by ""Brian Dangerfield"" <B.C.Dangerfield@salford.ac.uk>
Marziyeh:

Comparisons of SD & econometric methodology are in:

DH Meadows & JM Robinson ""The Electronic Oracle"" Wiley, 1985

DH Meadows ""The Unavoidable a priori"" in ""Elements of the
System Dynamics Method"" J Randers (Ed) MIT Press 1980 (now
available from Pegasus Communications, Waltham, MA).

Best

Brian.



Prof Brian Dangerfield
Professor of Systems Modelling &
Executive Editor, System Dynamics Review
Centre for OR & Applied Statistics
Faculty of Business & Informatics
Maxwell Building
University of Salford
SALFORD M5 4WT
U.K.
Posted by ""Brian Dangerfield"" <B.C.Dangerfield@salford.ac.uk>
posting date Mon, 24 Jan 2005 17:58:26 -0000

Posted by ""Goncalves, Paulo"" <pgoncalves@exchange.sba.miami.edu>
You may find these references useful to inform the differences and
similarities between system dynamics and econometrics as well as to
inform how statistics and econometric results can be integrated with SD
models.

Senge, P.M. (1977). ""Statistical estimation of feedback models.""
Simulation. 177-184.

Meadows, D.H. 1979. The Unavoidable A Priori. In J. Randers (Ed.),
Elements of the System Dynamics Method. MIT Press: Cambridge, MA.

Senge, P. (1980). ""A System Dynamics Approach to Investment Function
Formulation and Testing."" Socio-Economic Planning Science. 14: 269-280.

Sterman, J.D. (1984). ""Appropriate Summary Statistics for Evaluating the
Historical Fit of System Dynamics Models."" Dynamica, the international
journal of system dynamics. 10: 51-66.

Homer, J.B. (1993). ""Partial Model Testing As A Validation Tool for
System Dynamics."" Pine Manock College - Chestnut Hill - MA: 1-26.

Oliva, R. (2003). ""Model calibration as a testing strategy for system
dynamics models."" European Journal of Operational Research. 151:
552-568.

Regards,

Paulo Goncalves
paulog@miami.edu
Posted by ""Goncalves, Paulo"" <pgoncalves@exchange.sba.miami.edu>
posting date Mon, 24 Jan 2005 09:10:14 -0500

Posted by ""Corey Lofdahl"" <clofdahl@bos.saic.com>

This is an important question for several reasons, not the least of which is
that econometrics is the established methodology and system dynamics (SD),
in comparison, is a relative upstart. I'll make three observations.

First, from a philosophy of social science perspective, SD is deductive and
econometrics is inductive. SD is less tightly bound to actuarial data and
thus is free to expand out and examine more complex, theoretically informed,
and postuated relationships. Econometrics is more tightly bound to the data
and the models it explores, by comparison, are simple. This is not to say
the one is better than the other: properly understood and combined, they're
complementary. This is not obvious as I've recently been in an academic
forum where voices were raised and accusations made over precisely this
topic. The basic disagreement was over the fact that one researcher was
very good at a deductive technique and the other an inductive technique.
Methodological overspecialization is not a defensible position from which to
argue.

Second, SD is able to perform scenario analysis by representing causation.
Econometrics examines historical relationships through correlation. It is
useful to think about why this is so. Most econometric relationships use
linear models of the y = mx + b variety and then use least squares to
compute the fit. In contrast, consider a simple overshoot and collapse SD
model and its shifting loop dominance. The initial growth portion of say
population is driven by the amount of food available. So there is a
correlation between population level and food. What makes SD more powerful
and enables it to do scenario analysis (bordering on prediction) in a way
econometrics cannot is that it implicitly contains the limiting loop logic
that changes the direction of the growth curve. For a time population
grows, then it falls. This behavior is hard for an econometrics model to
capture.

Third, there is another, more practical reason to think about both SD and
econometrics. Many students interested in SD have to write theses for
faculty who are not SD experts. This is a very real problem for students
who need to graduate and are interested in SD but their advisors have been
trained in econometrics. In this regard I can only speak to my own
experience, but I have found it useful to frame the question using SD --
specifically using the reference mode -- from that perspective create an
interesting, unique, and defensible econometric model, and then use and SD
simulation to explain why the results are not statistical artifacts. The
faculty can then sign off on the econometric results independent of their
understanding of the SD. Moreover, as noted initially, the two
methodologies -- SD simulation and econometric models -- combine in a
deductive, theoretical and inductive, empirical fashion. For those who wish
to learn more, an example can be found at,

Lofdahl, Corey. 2002. _Environmental Impacts of Globalization
and Trade: A systems study_. Cambridge, MA: MIT Press.

best,

Corey Lofdahl
SAIC
Posted by ""Corey Lofdahl"" <clofdahl@bos.saic.com>
posting date Tue, 25 Jan 2005 10:40:36 -0500