Intuitive understanding of system dynamics concepts

["Tzur Levin"]

Dear all,
My research concerns the intuitive understanding of system dynamics concepts. I focus on the way people grasp the dynamic behavior of model structures. The questions I present students are of the form: "Given a static description of a systems structure (a causal loop diagram + natural language story), draw a graph describing the behavior of a certain variable". The required qualitative graph may be of exponential growth, goal seek, S-shaped growth ect.

In particular I contrast continuous models with hybrid models (continuous+discrete). I present two groups of students with two types of models describing similar phenomenon: a continuous (causal loops diagrams only) and a hybrid (causal loop diagrams + finite state machine). Then I compare their performance in drawing graphs and answering questions for each case.

I will be grateful for any information concerning
1.. studies on the pedagogical value of causal loop diagrams (besides Richardson 86, 97; Sterman 2000)
2.. studies on cognitive aspects of deriving dynamics from static structures (besides Sterman&Sweeney)
3.. cases where hybrid modeling is applied to social systems
4.. any comments concerning the research
Thanks,
Tzur Levin levintzu@post.tau.ac.il
Tel Aviv University, Israel

[Bill Harris]

For your second question, check out Alan Grahams PhD dissertation,
"Principles on the relationship between structure and behavior of
dynamic systems," 1977. You might also find
http://www.albany.edu/~gpr/LoopDom.pdf of interest.

Bill
From: Bill Harris <bill_harris@facilitatedsystems.com>

--
Bill Harris 3217 102nd Place SE
Facilitated Systems Everett, WA 98208 USA
http://facilitatedsystems.com/ phone: +1 425 337-5541

["Paul Atkins"]

Dear Tzur,

My interests are similar to yours.

> I will be grateful for any information concerning
> 2.. studies on cognitive aspects of deriving dynamics from
> static structures (besides Sterman&Sweeney)

I presented a poster at the Atlanta SD conference on an instrument I
developed called the Dynamic Forecasting Questionnaire. This was used with
approx. 500 participants drawn from school, first year uni, 3rd year
engineering and MBA populations. The target system was slightly more complex
than Booth-Sweeney & Sterman (2000) being a population with births and
deaths and two feedback loops. However, instead of having participants draw
their responses, I had them choose from list of multiple choice
alternatives. Participants were still very poor at picking the correct
dynamic behaviour. In one condition, where participants had the option of
choosing from graphs showing the stock moving in either direction, 39% of
engineering students chose the wrong direction!

I have not pursued this research further. With this approach it is very
difficult to isolate specifically what the problem is. I believe that in my
research at least, there was a problem associated with the use of the
bathtub metaphor (ie. understanding of the concept of births flowing in and
deaths flowing out of the stock). I was intending to follow up with tests
involving natural language without either graphs or SD representations to
see if performance was as bad but have not done so as yet.

I would be happy to supply the poster text to anyone who is interested.

Cheers,
Paul
From: "Paul Atkins" <paul.atkins@anu.edu.au>

-----------------------------------------------------
Dr Paul Atkins
Director of Research and the MBA Program
National Graduate School of Management
120 McCoy Circuit
The Australian National University
Canberra ACT 0200

["Jim Hines"]

Bill Harris is right about the usefulness of Alan Grahams Ph.D.
dissertation for understanding the relationship between structure and
behavior. If you havent read it, you have something to look forward
to: Put it at the top of your list of what to read next.

Another wonderful dissertation about "where behavior comes from" is
Nathan Forresters -- it will be particularly valuable for those people
who have a bit of background in linear systems and ordinary differential
equations. (Alans dissertation has no pre-requisites).

Jim Hines
From: "Jim Hines" <jhines@MIT.EDU>

["Jim Hines"]

A couple of people have written to ask how to get Alans and Nathans
theses. Bill Harris is probably fielding similar queries. Heres the
scoop:

The thesis are available from the M.I.T. Libraries Theses Collection.
The titles and dates are

A dynamic synthesis of basic macroeconomic theory : implications for
stabilization policy analysis . Nathan Blair Forrester. (1982-21).

Principles on the relationship between structure and behavior of dynamic
systems.. Alan Karl Graham. (1977-17)

The general link for MIT theses is
http://libraries.mit.edu/docs/theses.html. Nathans and Alans can be
viewed over the link for web Alan Grahams thesis is

http://theses.mit.edu/Dienst/UI/2.0/Des ... ?abstract=.

The link for Nathan Forresters is
http://theses.mit.edu/Dienst/UI/2.0/Que ... art+search+.

In this same important thread is also Bob Ebeleins excellent thesis:
Simplifying dynamic models by retaining selected behavior modes . Robert
L Eberlein. (1984-26).

The link to Bobs thesis is:
http://theses.mit.edu/Dienst/UI/2.0/Des ... ?abstract=.

A foundation work for both Nathans and Bobs efforts was a thesis by a
fellow in electrical engineering:

Selective modal analysis with applications to electric power systems .
José Ignacio Pérez Arriaga. (1981-92).

The link to Arriagas thesis is
http://theses.mit.edu/Dienst/UI/2.0/Des ... ?abstract=.


Jim Hines
jhines@sloan.mit.edu