Modeling in groups: who does that?

[Bill Braun]

I draw a distinction between conceptualizing a model (by working at a flip
chart with people, developing thinking lists and connections, causal loop
diagrams and SD model sketches) and actually building the SD model (I work
in PowerSim) and testing it.

My opinion is that group think is indispensible for development but group
SD model building in real time would drive me insane. As I built the SD
model I would bring it back to the group frequently for critique and
revision. Id also have the group test the model upon completion.

Bill Braun
From: Bill Braun <medprac@hlthsys.com>

["Raymond T. Joseph"]

I like starting off with the modelers to do a group overview - top down.
Once a model is broken down into modules directly convertible to code,
everyone takes their pieces and runs. Working in sufficiently close
proximity allows for productive interaction and problem resolution.

Ray
--
Richard G. Dudley
rdudley@indo.net.id
http://home.indo.net.id/~rdudley

["Richard G. Dudley"]

I often read that the best models are built by groups of people working
together, but most modelers I know like to work alone.

I was wondering what techniques people have successfully used to get the
two types of activities (group discussion and at the keyboard modeling)
meshed in a productive way?

--
Richard G. Dudley
rdudley@indo.net.id
http://home.indo.net.id/~rdudley

[Stephen Shervais]

Richard,

At my former company this was done primarily through a series of meetings
between the development staff and the technical representitives of our
DoD customer. The key questions centered on what kind of statistics they
wanted -- what was being measured -- so that the developers could decide
if they needed to model something in detail or just "capture the effect."

Operationally, it often came down to a cycle of "what if we did this.."
discussions. "Suppose we just set a flag, and measure the time from then
until the flag is cleared."..Well, but..."OK, suppose we set a flag, and
then poll the modules to see if any of them have had an event..." etc.

The models I was associated with were all discrete event, rather than SD,
and were pretty mature by the time I came on the scene, so the original
process may have been different.

Regards,
Steve

Steve Shervais shervais@acm.org
Graduate Student shervais@sysc.pdx.edu
Systems Science PhD Program psu00872@odin.cc.pdx.edu
Portland State University http://www.sysc.pdx.edu
Portland, Oregon http://www.ee.pdx.edu/~shervais
(503) 725-7344 / 725-4997

[Anthony Gill]

Group Model Building by Jac A M Vennix published by Wiley 1996 ISBN
0-471-95355-5 should answer some of your questions.

Tony Gill phone: +44 (0)1295 812262
Phrontis Limited
Beacon House fax: +44 (0)1295 812511
Horn Hill Road
Adderbury email: TonyGill@phrontis.com
Banbury
OXON. OX17 3EU URL: http://www.phrontis.com/
U.K.

[Niall Palfreyman]

Hi,

In object-oriented software design there is a technique which is
sometimes rubbished as not "serious", but which can have amazing results
in terms of developing a group understanding of the requirements on a
software project. This technique is called "CRC"
(Class-Responsibility-Collaboration) - see "Using CRC Cards : An
Informal Approach to Object-Oriented Development" by Nancy M. Wilkinson
for a good coverage.

What CRC basically does is to assign to participants various roles
(stocks) which are found within the system under investigation. The
participants then pass information to each other (flows) in order to
enact a simulation of the system. In the process they discover new
components of the system (which are then in turn assigned as roles to
further participants) and also uncover divergences in their
understanding of the system and its requirements. The result is a system
which everybody understands in (more or less) the same way (a major
problem in software design!), and in which everybody feels personally
involved. I personally think CRC is a major development tool which we
have not yet investigated to its full potential, and Im convinced it
could be a great help in system dynamics - not that I can claim much
experience in system dynamics model development.

Hope this helps,
Niall.

--
Have fun!

Niall Palfreyman email: Niall.Palfreyman@assyst-intl.com
assyst GmbH, Henschelring 15a
85551 Kirchheim bei Muenchen Tel: ++49-89-90505-230
Germany. Fax: ++49-89-90505-102/3

[carolus]

Richard,

(regarding group modeling) try:

1.
Group model building : facilitating team learning using system dynamics
/ Jac
A.M. Vennix. - Chichester [etc.] : Wiley, cop. 1996. - XIII, 297 p. :
ill. ; 24
cm
Lit. opg.: p. [278]-292. - Index.
ISBN 0-471-95355-5 cloth

and:
2.
the special issue of the System Dynamics Review on Group Model Building
Volume 13, nr. 2, summer 1997

greetings,
--
Carolus Gr=FCtters

+++++++++++++++++++++++++++++++++++++++++++++++++++++
Recht en Informatica Law and IT
Faculteit der Rechtsgeleerdheid Faculty of Law
KU Nijmegen, University of Nijmegen

URL: http://www.jur.kun.nl
it/
phone: +31 -24 -361.57.01 P.O. Box: 9049
fax: +31 -24 -361.61.45 6500 KK Nijmegen
email: carolus@jur.kun.nl The Netherlands
+++++++++++++++++++++++++++++++++++++++++++++++++++++