Where Do Babies Come From

This forum contains all archives from the SD Mailing list (go to http://www.systemdynamics.org/forum/ for more information). This is here as a read-only resource, please post any SD related questions to the SD Discussion forum.
Locked
Bill Braun
Senior Member
Posts: 73
Joined: Fri Mar 29, 2002 3:39 am

Where Do Babies Come From

Post by Bill Braun »

> for example, has babies emerging
>from a "cloud", a device I find extremely counterintuitive, when its
>obviously a circular flow.

I had the same difficulty until someone suggested that I think of the
clouds simply as a convention for indicating the boudaries of the model. As
she put it, You could have included a lot more in the model and you didnt.

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

The Health Systems Group
- Physician Leadership Training
- Simulation Modeling for Healthcare
http://www.hlthsys.com
Ford Stone
Junior Member
Posts: 4
Joined: Fri Mar 29, 2002 3:39 am

Where Do Babies Come From

Post by Ford Stone »

Bill Braun wrote:

> I had the same difficulty until someone suggested that I think of the
> clouds simply as a convention for indicating the boudaries of the model. As
> she put it, You could have included a lot more in the model and you didnt.

Not to belabor the point, I do appreciate - and
support - the use of "clouds" for
this purpose. But I had in mind the standard
population loop, in which babies
are definitely "in the model". And anybody on
this board almost certainly knows about
alternating between stocks and rates of flow.
All this notwithstanding, to all respondents to
this issue, your points are well taken, and your
considerable efforts are appreciated.

I guess that where Im coming from is as a
teachertrying to make what I have found to be a
rather awkward transition from causal loop models
(simple in appearance, albeit,
yes, fraught with modelers pitfalls) to Stella,
which I have used for years.
I guess I yearn for a simpler graphical
representation that still captures , but perhaps
hides, the subtlties inherent to all this . Maybe
such is not possible, but I can still yearn.

My final comment is that perhaps we might try
giving this new spreadsheet software
the benefit of the doubt. Surely with any such
program rates and delay factors can be specified.
Afterall, in Dynamo and Stella there are really
mathematical expressions running the show
underneath it all. Another, and i think, not
inconsiderable point to be made here is that there
are a variety of other plug-ins for Excel that
might be useful to system modelers, such as fuzzy
logic, neural nets, and genetic optimizers, and, I
think, but am not sure, inputs from the real
world. I long ago suggested to Stellas
programmers that hooks to the real world would be
extremely valuable. After all, whats really
instructive is to play ones model off against the
real world process one is modeling. The
discrepancies force us to question how our model
is distorting or failing to capture some aspect
of reality.

f.
From: Ford Stone <
fordston@caverns.com>
"HARRIS,BILL (HP-LakeStevens,ex1
Junior Member
Posts: 5
Joined: Fri Mar 29, 2002 3:39 am

Where Do Babies Come From

Post by "HARRIS,BILL (HP-LakeStevens,ex1 »

Ford Stone wrote:
....................................................................
My final comment is that perhaps we might try
giving this new spreadsheet software
the benefit of the doubt...


Im all in favor of giving people and tools chances, but I did discover
something that has made me a bit more leary of spreadsheets as tools.

Recently I pointed in this list to a paper on an expense management tool I
did based on an SD model. The first implementation of the actual tool (not
the SD model) used a spreadsheet as the calculation engine. It was easy for
the person who actually created it to build it. It basically worked. It
was also easy to modify for temporary, ad hoc analyses. Unfortunately, it
was hard to see that such modifications had been made, and so it was hard to
spot errors in the system thereafter.

A revised model in a functional array programming language made it pretty
easy to spot changes in the system. One statement contained all the "logic"
inherent in a significant chunk of the system, so it was fairly obvious to
someone who could program in that language if something were wrong. Errors
decreased.

Stock and flow diagrams are also good at that. The picture you see
represents the model or calculation, with only some of the details being
hidden. Those details are almost always invariant across time. (It takes a
bite of extra effort to craft a rate equation, for example, that has enough
conditionals to be a different calculation at a small number of points in
time.)

A spreadsheet, IMHO, hides more of the details (e.g., are the cell formulae
the same in each cell in a range or did one get changed ever so slightly?)
and makes it harder to see whats really going on.

So, use a spreadsheet if its the right tool for the job. Just beware -- it
carries some risks, and it may be better to teach stocks and flows or even
programming than to hide too much in "invisible" cell formulae.

Bill
--
Bill Harris Year 2000 Program Office
mailto: bill_harris@am.exch.hp.com Hewlett-Packard Company
phone: (425) 335-2200 M/S 330C
fax: (425) 335-2483 8600 Soper Hill Road
web: http://hpweb.lsid.hp.com:8080/~billh/ Everett, WA 98205-1298
Locked