September 1998
This is a monitored list for discussion of system dynamics in K-12 education.
Send contributions and all requests to subscribe and unsubscribe to:
k-12sd@sysdyn.mit.edu
Date: Wed, 9 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: "Jay W. Forrester" <jforestr@MIT.EDU> (by way of k-12sd)
Subject: REPLY "ownership" in the public sector (SD1622)
Below is a message from Donella Meadows on the system dynamics discussion
group and my reply that may be of interest to the K-12 list participants.
>Date: 10 Aug 98 10:39:08 EDT
>From: Donella.H.Meadows@Dartmouth.EDU (Donella H. Meadows)
>Subject: REPLY "ownership" in the public sector (SD1622)
>To: system-dynamics@world.std.com
>Sender: system-dynamics-approval@world.std.com
>Precedence: list
>Reply-To: system-dynamics@world.std.com
>
>I am fascinated by the ongoing discussion on this list about "ownership,"
>"buy-in," what we used to call "implementation," and what in a larger context
>we could call "social change" or even "paradigm shift."
>
>I have several questions for those of you who have developed wisdom in these
>matters from business consulting:
>
>Would your experience translate directly into the public sector? If not, what
>would have to change to create significant systems insight and change there?
>
>Given that many of the most dysfunctional systems and inadequate mental models
>and insufficiently long time horizons are outside of the business world,
>how do
>we go to work on those?
>
>Given that the business world itself is imprisoned in an increasingly
>dysfunctional super-system, namely the market with all its externalities,
>success-to-the-successful loops, blindness to nonpriced values, growth
>fixations, shrinking time horizons, increasing monopolies, and growing hubris,
>where are the levers for "buy-in" for systems analyses of the market?
>
>The way I see it, the businesses so many of you are successfully working with
>are embedded in larger systems that REALLY need help. I don't think I need
>elaborate for this list the many global trends that are accelerating in
>terribly destructive directions, some of them driven by perfectly logical
>business decisions. No matter how successful you are in transforming the
>behavior of a single business client, the client is still forced by the larger
>system into behaviors that are disastrous for society (and ecosystems) as a
>whole, and therefore, after a lot of long-term feedback loops come home to
>roost, disastrous for that business as well. The Natural Step is trying hard
>to enlarge the context of business thinking, but can those who are trapped
>within a system really challenge the basic structure of that system?
>Don't we
>need to spend at least as much time and energy on public sector clients and on
>the general public?
>
>How can we direct more excellent systems analysis and change-expertise to the
>big systems (where the money isn't, where there are no obvious single clients,
>where there isn't a whole lot of perceived need for change -- at least not on
>the surface, and where the very mention of change or the honest discussion of
>problems causes instant loud denial)?
>
>What is going to happen if we don't?
>
>Is the predominance of industrial consulting in our field due to the greater
>immediate financial rewards? Or to the fact that we have given up on the
>overarching, difficult, ultimately determining systems?
>
>Donella Meadows
>From: Donella.H.Meadows@Dartmouth.EDU (Donella H. Meadows)
>
------------------------------------------------------
From Jay Forrester:
In: REPLY "ownership" in the public sector (SD1622), on August 10, 1998,
Donella Meadows wrote:
>Given that many of the most dysfunctional systems and inadequate mental models
>and insufficiently long time horizons are outside of the business world,
>how do
>we go to work on those? ........
>Don't we
>need to spend at least as much time and energy on public sector clients and on
>the general public? ........
>
>How can we direct more excellent systems analysis and change-expertise to the
>big systems (where the money isn't, where there are no obvious single clients,
>where there isn't a whole lot of perceived need for change -- at least not on
>the surface, and where the very mention of change or the honest discussion of
>problems causes instant loud denial)? ........
>
Or, have we
>given up on the
>overarching, difficult, ultimately determining systems?
We have had an interesting discussion of "Systems from the top" to which
Donella was responding. But the problem of understanding systems cannot
start from the top in the public sector and seldom works in the private
sector.
In corporations and governments, system dynamics has often been introduced
at the top on the assumption that "decision makers" actually make decisions
that will be implemented. Usually the lower levels of an organization are
successful in seeing their roles as protecting the organization from
capricious and unwanted decisions at the top. But the difficulty is more
fundamental.
One probably must work with the top managers of a company for at least five
years before they can internalize the radically new paradigm of dynamic
systems. By that time they either retire or die and one must start over.
In the public sector, the major threats to which Donella refers have their
solutions only in policies that are contrary to conventional wisdom and
popular expectations. There is no government led by people who can both
develop personal understanding of systems and also educate and lead a
public in unexpected directions.
I believe that change of the magnitude that Donella describes can not be
inserted from the top down. To be effective there must be a large
constituency providing support from the bottom up to the top.
Donella asks if we have, "given up on the overarching, difficult,
ultimately determining systems?" The answer is, NO. But one must go about
the change in a basic fundamental way, even if such does not give quick
results.
It is my belief that the answer to Donella's concern lies in the present
pioneering activities to bring system dynamics into K-12 education. That
effort is only beginning, but the start looks most promising. The
conference on system dynamics in K-12 education held in June of this year
was attended by about 270 very enthusiastic teachers. Several dozen
schools are doing good work.
Given another 20 years, enough schools should be converted to a dynamic
education to make a significant impact on the student output stream. In 20
years beyond that, the output of such schools will be finding their way
into leadership positions in corporations and government.
So, you may say that 40 years is too long to wait? It has already been 40
years since the beginning of system dynamics. If we continue to focus only
on introducing the systems paradigm from the top, another 40 years can
elapse with little change from conditions that exist today. Time is on the
side of doing the job right.
Based on the above reasoning, I decided some ten years ago that I should
focus much of my own effort on helping to create a new kind of
system-dynamics-based
education at the K-12 level. It is much easier for an elementary school
student to understand dynamic feedback systems than for someone who is 45
years old, because the young student has so much less static, nonfeedback
baggage to unlearn.
---------------------------------------------------------
Jay W. Forrester
Professor of Management, Emeritus
and Senior Lecturer, Sloan School
Massachusetts Institute of Technology
Room E60-389
Cambridge, MA 02139
tel: 617-253-1571
fax: 617-258-9405
Home office:
tel: 978-369-9372
fax: 978-369-9077
Date: Fri, 11 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Niall Palfreyman <Niall.Palfreyman@assyst-intl.com> (by way of
k-12sd)
Subject: Re: REPLY "ownership" in the public sector (SD1622)
Jay W. Forrester (by way of k-12sd) wrote:
>
> Based on the above reasoning, I decided some ten years ago that I
> should focus much of my own effort on helping to create a new kind of
> system-dynamics-based education at the K-12 level.
I agree that change at grassroot level is the only way real, effective
change can occur, but I think Jay is also doing something else in
parallel.
In the area of interpersonal communication, one of the most powerful
tools (or maybe even THE most powerful tool) for rapid, effective change
is the use of spiritual awareness - the sense each of us has in one form
or another of relationship to something in some sense beyond ourselves:
maybe Community, maybe God. If one person can relate an issue in the
minds of people to such spiritual awareness, then that person can effect
grassroots change.
And that, to my mind, is another part of what we are trying to do. We
ourselves represent system dynamics to the extent that it enters into
public awareness, and I assume system dynamics is for most of us a
compelling, even spiritual, vision of the future. Consequently it comes
across that way when we interact with others in our various committees,
organisations and social groups. I contend that this is as important a
way of changing the world as going into schools. After all, how can we
influence education unless teachers feel swept along by our vision of a
system-aware world?
Niall.
If we gain something, it was there from the beginning.
If we lose something, it is hidden nearby.
Dr. Niall Palfreyman mailto: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
Date: Fri, 11 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: teresa@northwest.com (Teresa Hazel) (by way of k-12sd)
Subject: Introduction and Inquiry
Hello, I am Teresa Hazel. I teach at Valley Catholic High School in
Beaverton, Oregon.
I have been working with Systems Thinking/Dynamic Modeling in my classes
off and on for the past four years, mostly in my previous classes, at my
previous school, LaSalle High School.
Now that I have moved to another school I am the "seed" there for Systems
Thinking as Jay Forrester describes those people trying to get things going
at their school. It is an uphill climb. I am now in the process of
getting a computer system together with a video projector just to be able
to do some large group systems work. I don't have software right now as I
am changing to PC format and I only have access to a computer lab for two
of my six classes. I not only plan to introduce Systems Thinking to my
students but I also plan to team teach some sessions with other teachers to
introduce both them and their students to Systems Thinking. I am
interested in hearing from others who have attempted the same work, what
difficulties you encountered, helpful hints, etc. I am also interested in
making this my action research project to complete requirements for my
Masters degree so....the proverbial question will come up I'm sure that we
have all encountered and that is..."how do we assess these projects in
terms of improved student learning. We all know that definitely happens
but I'm sure I will have some hoops to jump through regarding measurement
of the improved learning. Any helpful hints here would also be greatly
appreciated. Thanks, TERESA
Date: Tue, 15 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Jean-Louis Cordonnier <jlcord@wanadoo.fr> (by way of k-12sd)
Subject: LH, FSH, oestrogens & progesterone
Hi everybody.
This year, I have to teach about sexual hormons, ...
The production of ovarian hormons is regulated by LH and FSH, produced by
the hypophyse. The production of those two hormons depends on ovarian
hormon concentration ; it's a beautiful example of a non homeostatic loop,
that produces fluctuations (menstruations, ...).
Does any of you know if there is a model (Stella, Vensim or other) of this
physiological feed-back ?
Thanks for your expected replys.
Jean-Louis Cordonnier
jlcord@wanadoo.fr
36, rue Lavisse
66000 PERPIGNAN
FRANCE
jlcord@wanadoo.fr
Date: Tue, 15 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Charis Dube <charis@ns1.pacsci.org> (by way of k-12sd)
Subject: Greetings
Hello all,
I just found your list and I'm very excited to be a part of this discussion.
I am a teacher trainer for Pacific Science Center in Seattle. We train
teachers all over the state with a focus on science education reform. Part
of that message is that science is all about systems. The concepts are
related, the different sciences are related and science is related to all
other subjects.
I am also a student at Antioch University - Seattle in the Whole Systems
Design: Organization Systems Renewal program. Most of the people in my
program are organization development consultants for business or
non-profit. My personal focus in this program is the application of
systems thinking to educational organizations. I am a consultant on
organizational development within schools.
I have a theory that as a society we are moving from cooperation (and
cooperative learning) to collaboration (and collaborative learning). The
best way that we can help our students make this shift is to help teachers
make this shift. If teachers are working in a more collaborative manner
they will find it easier to teach in a more collaborative manner.
If any of this is of particular interest to you, please feel free to send
me a note. I'd love to network more on this and here what more people have
to say.
Thanks so much and I look forward to the discussion.
Charis Dube
Teacher Education Supervisor
Pacific Science Center
(206)443-2385
charis_dube@pacsic.org
Date: Wed, 16 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Ed Gallaher <gallaher@teleport.com> (by way of k-12sd)
Subject: Re: LH, FSH, oestrogens & progesterone
I don't know of explicit materials, but we will be teaching some of these
concepts in a graduate course in Physiologyy and Pharmacology in the next
few months. At what level will you be teaching this material?
We may be able to develop some of this material over time, but correct
models of such systems need to be well thought out to avoid errors, so it
won't happen quickly.
Please keep in touch.
Ed Gallaher
Assoc Prof Physiol-Pharmacol and Behav Neuroscience
Oregon Health Sciences University
Date: Tue, 22 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Mary Ellen Verona <mverona@mvhs1.mbhs.edu> (by way of k-12sd)
Subject: correctness of models
Hi all,
We have been thinking a lot about how to *tune* model correctness to model
purpose. Several of our teachers are familiar with Ed G's "rain barrel"
models. We also had a teacher develop a flow model using a burette. Data
is collected to determine the net flow rate as a function of volume. That
is then fit to a linear equation. The experimental situation was set up
so that data was not collected on the outflow of the last of the liquid.
After some discussion, some of us became unhappy about the linear equation
predicting that there would be outflow even when the stock was empty,
since in fitting the data, the flow equation mx+b, b was not 0.
This led to some research - we read about the idea of assuming that all of
the potential energy of the liquid is transformed into kinetic energy. So
we could solve for the velocity of the flow depending on height (and on
volume of the burette as a proxy - given a constant girth of the
container). Thus we came up with flow as dependent on sqrt of stock -
instead of 1st power. We were able to fit our data nicely.
Here is my question - how do we determine whether to introduce this more
complicated formulation - which of course also neglects many factors of
flow? There are more questions - e.g. the idea above introduces some
minimal physical theory to guide curve fitting. When and how do we need
to do this?
Mary Ellen Verona
mverona@mvhs1.mbhs.edu
***** new address ******
Maryland Virtual High School
Montgomery Blair High School
51 East University Boulevard
Silver Spring, MD 20901
301-649-2880
Date: Tue, 22 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Kardi Teknomo <kardi@peter.petra.ac.id> (by way of k-12sd)
Subject: Intro
Hi all!
My name Kardi Teknomo. I am a new member of this discussion group.
I'm teaching trafiic engineering and transport planning, Dept. Civil Enggr.,
Petra Christian University, Surabaya, Indonesia. I have read some materials
of System Dynamics and its seem interesting. I hope I can learn more from
this group.
Kardi
e-mail:kardi@peter.petra.ac.id
Date: Tue, 22 Sep 1998 To: k-12sd@sysdyn.mit.edu
From: Niall Palfreyman <Niall.Palfreyman@assyst-intl.com> (by way of
k-12sd)
Subject: Re: LH, FSH, oestrogens & progesterone
Jean-Louis Cordonnier (by way of k-12sd) wrote:
>
> This year, I have to teach about sexual hormons, ...
Hi Jean-Louis,
I've just bought a book of STELLA models in the Health Sciences. You can
find it at Amazon under the following URL:
http://www.amazon.com/exec/obidos/ASIN/0387949968/qid%3D906017317/002-6501762-45
61253
It doesn't have any models directly relating to hormones, but you may
still find it interesting.
Good luck!
Niall.
--
If we gain something, it was there from the beginning.
If we lose something, it is hidden nearby.
Dr. Niall Palfreyman mailto: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
Date: Tue, 22 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Niall Palfreyman <Niall.Palfreyman@assyst-intl.com> (by way of
k-12sd)
Subject: Re: Greetings
Charis Dube (by way of k-12sd) wrote:
>
> I have a theory that as a society we are moving from cooperation (and
> cooperative learning) to collaboration (and collaborative learning).
Hi Charis,
This sounds fascinating to me - I'm very interested in using systems
thinking to help people solve problems in their everyday lives. I'd be
very interested to hear more about how precisely you understand this
distinction between "cooperation" and "collaboration".
Look forward to hearing from you.
Best wishes,
Niall.
--
If we gain something, it was there from the beginning.
If we lose something, it is hidden nearby.
Dr. Niall Palfreyman mailto: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
Date: Tue, 22 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: "sergio romagnoli" <patagonia5@hotmail.com> (by way of k-12sd)
Subject: reply from argentina
I see this topics in samples of STELLA, in the page of High Performance
Systems( www.hps-inc.com).
When you download the Stella Software, you will see a directory named
samples, where you find a exercise of homeostasis in insuline &
diabetes.
Good luck.
Sergio Romagnoli.
Argentina
Patagonia.
Date: Wed, 23 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Randy McDaniel <rmcdaniel@usd257.net> (by way of k-12sd)
Subject: introduction
My name is Dr. Randy McDaniel and I'm the Supt. of Schools in Iola, KS (a
small rural area in SE Kansas). I first got interested in system dynamics
through my professor at Wichita State University and his love of "Chaos"
theory. I am interested in how school systems do not fully understand how
their independent actions affect students as they progress through the
school system. I am trying to set up "models" that can demonstrate how this
occurs and would welcome feedback.
Thanks, Randy McDaniel
Date: Wed, 23 Sep 1998 To: k-12sd@sysdyn.mit.edu
From: George Richardson <gr383@cnsvax.albany.edu> (by way of k-12sd)
Subject: Re: correctness of models
On Tue, 22 Sep 1998, Mary Ellen Verona wrote:
> Here is my question - how do we determine whether to introduce this more
> complicated formulation - which of course also neglects many factors of
> flow? There are more questions - e.g. the idea above introduces some
> minimal physical theory to guide curve fitting. When and how do we need
> to do this?
This is not a question about good system dynamics. It is a question about
science. And good science teaching.
I would think the experiment and the modeling described by Mary Ellen are
great opportunities to get at what science is -- whether or not the
students get to what an expert would say the correct functional form ought
to be.
Building a linear model, comparing it to data, and testing it with the
thought experiment of "no water in the burette" are all great steps in
what a scientific approach would be. I'd involve the kids it in for sure.
- If the kids at some level can't handle more than the linear model, then
they can at least see that the mathematical model is only approximate, and
they can use it where it fits and discard it where it doesn't.
- But if the kids can handle more complex mathematical relations (and
that should be easy in a simulated word -- flow is constant, versus
flow depends on x, versus flow depends on x*x or 1/x), then I'd ask the
kids for their thoughts on how to make the model fit the data better.
Scatterplots of flow versus height of the water in the burette might bring
out the nonlinear character, and they could guess at what might produce
that.
- If the kids aren't kids anymore and can actually do some of the
physicial theorizing Mary Ellen sketched, that that would be a wonderful
follow on the the observations that the linear model does't fit well.
They might get a theoretical nonlinear functional form that they could
then test in experiment and in simulation.
At some levels I hope we would not care if they got it "right" as long as
they were thinking rightly and gaining the habits of mind of good science.
Still, Mary Ellen's last question [when and how should we use theory to
guide curve fitting] deserves a better answer. Mary: are you asking how
do we know when kids are ready for using theory to guide their empirical
searches, or are you asking if there is a time when all science programs
should do this sort of thing and what's the best pedagogy?
...GPR
-----------------------------------------------------------------------
George P. Richardson G.P.Richardson@Albany.edu
Rockefeller College of Public Affairs and Policy Phone: 518-442-3859
University at Albany - SUNY, Albany, NY 12222 Fax: 518-442-3398
http://cnsvax.albany.edu/~gr383/
Date: Wed, 23 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Ed Gallaher <gallaher@teleport.com> (by way of k-12sd)
Subject: Re: correctness of models
This is long, but I got on a roll; sorry . . .
Mary Ellen (Verona) raises several important issues here.
First, there is a lack of material available that allows teachers and/or
students to move between the lab and the modeling exercise to test the
"correctness" of models. Although this seems like a great idea (it is!),
I'd like to suggest caution, in that we learn how to do this before asking
our students to do it. As Mary Ellen is finding, it may not be that easy.
I am guilty of a minor sin with the rain barrel, but it should get resolved
in the near future. Although the height of the water creates a decreasing
pressure head as the barrel drains, in fact, a pure exponential decay is
NOT what one finds. Gravity plays a role, and the equation for draining a
barrel through a simple hole is not quite that simple. This confused me
for a couple of years when I tried to confirm the model with a 2-gallon jug
draining into the sink. It did not fit quite right.
I believe (?) that what will work quite accurately is a long hose attached
to the base of the barrel. Imagine a threaded pipe at the bottom of the
barrel with a 25-foot (50-foot; 100 foot) hose attached. The hose could
vary in diameter (1/2 inch, 3/4 inch, 1 inch). In this case the primary
impediment is -resistance- of the water as it flows through the hose,
rather than the downward flow of gravity pushing through a hole that is too
small for the water to pass through.
If this is correct, then we should be able to conduct this experiment in
the lab with a milk jug, but the output will need to be a length of plastic
tubing, not a simple hole in the jug. A little more difficult to set up,
but more in agreement with the simulation results. (or more accurately,
the simulation is a more accurate model for the bottle with the hose than
the bottle with the hole. . . )
The beauty of the rain barrel exercise is that it leads easily to
discussions of other exponential processes, but we should be careful that
we don't do things incorrectly for expediency.
We could of course build the correct model for the influence of gravity,
but this requires more difficult physics, and then does not lead easily
into the elimination of drugs, etc.
Going one step further: If one sets up the simple jug with a hole in it,
sitting under a water faucet, with equal inputs and outputs, there WILL be
a constant volume of water in the jug. A certain fraction (let's say 3.5%)
of the volume WILL be entering and leaving each minute. As stated above,
the model for the water draining is not quite correct, but an absolutely
correct demonstration of drug elimination can still be accomplished.
Insert a conductivity probe into the bottle (Vernier Software, about $75).
This probe will measure the concentration of ions in the water. Add 10 ml
concentration salt solution to the jug. You will see a spike (pulse) on
the conductivity detector. Assuming moderate mixing, about 3.5% of the
salt will be removed each minute as the water flows through the system.
This will exhibit exponential decay, and the halftime can be calculated as
indicated in some of the rain barrel materials. This is an -excellent-
analogy for drug elimination. (The same experiment could be done with
colored water and a colorimeter if available.)
We will be working on some of these experiments over the next few months
and hopefully can expand on these materials with real experience (including
kids' experience!).
Regarding the buret, you are quite right to be concerned about the "linear"
flow. I'm not sure of the physics here either, but gravity is certainly a
factor. And of course you cannot have linear flow from an empty buret,
leading to a negative volume!
When teaching the RB materials I assume absolute beginners with no
knowledge of SD or modeling. I want to make it VERY simple. I start with
an empty barrel with only an input; no output. I then turn on a faucet
(linear input), dump in buckets (pulse), or turn the faucet up and down
(step function). Going to the simplest output example I start with a
barrel with 40 gallons in it (no input). You'd think the simplest output
would be linear (a pump), but this leads to the same problem Mary Ellen
describes, with possible negative volume.
The solution to this dilemma is absolutely NOT to use the non-negative
function in STELLA!! The solution is to determine the correct physical
structure of the model to accurately describe what is happening.
The solution is to insert an IF-THEN statement. Consider a vertical pipe
extending down into the barrel. It stops near the bottom and is attached
to a hose and pump at the top. Turning on the pump removes 2 gallons per
minute, UNTIL the barrel is emptied to the remaining 4 (or whatever)
gallons, at which time the pump starts sucking air.
The statement would thus be "IF VOLUME >4 THEN 2 ELSE 0"
This statement is correct and is perfectly valid modeling practice, but it
creates another potential problem. IF-THEN statements are used way too
much in all kinds of programming, and they are frowned upon by good
programmers. IF-THEN should be used rarely, and only if the
programmer/modeler -knows- it is appropriate. (This is one of those
occassions.)
For this reason, I usually skip this and move right to the barrel with a
hole in it. Also, when presenting this material, I mention that the hole
is not quite correct, but it works for our current purposes. I need to add
this information clearly to written descriptions of the RB exercises so
that others are not enticed down the wrong path.
One more issue -
Back to the burette - if you consider the burette is holding 100 ml of
water suspended in the air and you then cut off the bottom of the burrette,
now you have 100 ml of water simply falling to the ground under the
acceleration of gravity. It might as well be a tennis ball.
You mention that the square root of the stock fits the data nicely. This
would be a great demo if it is correct. Can you document and/or derive
this equation from a physics text? If not I'd be very cautious about the
message.
If we can collectively develop several models that allow students to move
back and forth from the lab to the computer we will have some VERY powerful
tools! But we better be absolutely sure that the relationships are
accurate.
Ed
Date: Wed, 23 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Mary Ellen Verona <mverona@mvhs1.mbhs.edu> (by way of k-12sd)
Subject: Re: correctness of models
Here is the information on using sqrt in the flow model:
Here is information that may be used with the flow model. I
took it from a modeling book - it is called the leaky bucket model.
In our situation, a, the area of the hole, is probably the same as A,
the cross sectional area of the bucket. This is a crude model.
Let h(t)=height of the water remaining in the bucket at time t, assume
that a and A are constant; v(t) is the velocity of water passing through
the hole.
1) av(t)=Ah'(t) (in our case v(t)=h'(t))
2) To derive an additional equation, use conservation of energy. First,
find the change in potential energy in the system, assuming that the
height of the water in the bucket decreases by an amount delta-h and
that the water has density rho. Then find the kinetic energy
transported out of the bucket by the escaping water. Finally, assuming
all the potential energy is converted into kinetic energy, derive the
equation v^2=2gh.
3) Combining 1 and 2, show that h'=-c*sqrt(h) where c=(a/A)*sqrt(2g)
In our case this means that h'=-sqrt(2gh) or flow=-sqrt(2g*stock)
Mary Ellen Verona
mverona@mvhs1.mbhs.edu
***** new address ******
Maryland Virtual High School
Montgomery Blair High School
51 East University Boulevard
Silver Spring, MD 20901
301-649-2880
Date: Wed, 23 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Mary Ellen Verona <mverona@mvhs1.mbhs.edu> (by way of k-12sd)
Subject: Re: correctness of models
Thanks George and Ed for *lots* of food for thought. George - about using
theory to guide modeling - yes, I am concerned about when this should be
done, since I think it is important. So often I have seen teachers using
graphical analysis and automatically picking the "best fit" giving a
functional form that does not relate at all to what they know
theoretically.
Mary Ellen Verona
mverona@mvhs1.mbhs.edu
***** new address ******
Maryland Virtual High School
Montgomery Blair High School
51 East University Boulevard
Silver Spring, MD 20901
301-649-2880
Date: Wed, 23 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: RICHARD TURNOCK <Richard_Turnock@pgn.com> (by way of k-12sd)
Subject: Big Systems
There is a Big System based on demographics defined in the books
"Generations" and "The Fourth Turning" by Strauss and Howe. As the Baby
Boomers, GenX and the new Millennial generation fill their next age cohort
group, their mental models will be shaped by events and they will shape
events based on their mental models. A consequence will be social change
and paradigm shifts.
(In reply to a question to him by the list moderator, Richard wrote....)
Hmmmmm, (wait, wait, I'm thinking), I was using the term "Big System" that
Donella Meadows used in the email forwarded by Jay Forrester. I think it
means large, dynamic, socio-economic systems at the scale of a country or
on a global scale. The Generations model is one example.
My point? Donella's note asked about how to intervene in "Big Systems";
using terms like "...create significant systems insight and change there";
"....how do we go to work on those?"; "can those who are trapped within a
system really challenge the basic structure of that system." I think we
need to develop a model of the "Big Systems" before we intervene in them.
I'm suggesting "Big Systems" have patterns, based on events, shaped by
mental models that lead to social change and paradigm shifts. Also, people
use their mental models to shape events. I listed one example of a "Big
System" in the US using a model of generations defined by birth year.
Another example is when China tried to intervene in a "Big System" with
their one baby policy. Without going into details in this email, a person
could show the relationships between events, mental models, social change
and paradigm shifts as a result of the Generations model or the Chinese
policy, and tell a very interesting story about the consequences.
Richard Turnoch
Date: Thu, 24 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: ASeaman970@aol.com (by way of k-12sd)
Subject: Re: correctness of models
Hi,
Mary Ellen Verona
mverona@mvhs1.mbhs.edu wrote:
>Here is my question - how do we determine whether to introduce this more
>complicated formulation - which of course also neglects many factors of
>flow? There are more questions - e.g. the idea above introduces some
>minimal physical theory to guide curve fitting. When and how do we need
>to do this?
Don't lose sight of the fact that all models neglect some factors, be they
known or unknown to us. Some models are more complete than others, but even
very crude models may be useful to us. There is never a complete model or a
"right answer." It is important to stress this to students. We tend to bring
children up to believe that there is always a "right answer" to any question.
This is, IMHO, a good opportunity to introduce the notion of a "good enough
answer" to the students. They simply must understand that all models have
shortcomings, and that the more we know about the shortcomings, the more
useful the model becomes. From there they can move forward. Surely by the time
a student reaches high school he/she can understand the concept and some of
the implications of the square of a number, so that should not be an
impediment. I have introduced squares of numbers (and square roots) to
children as young as 7-8 with satisfying results. Don't expect them to
completely grasp it at that age, but 7-8 is not too early to introduce the
concept with some simple application. Perhaps you would like to have the
students experiment with both the linear and sqrt models and decide for
themselves the utility of each compared to the amount of effort expended and
desired accuracy? How about letting them decide what the important questions
are, such as "What range of burette volume are we interested in modeling?" Are
we really interested in that last drop? It all should be put into context.
Good luck!
Art Seaman in Jackson, TN
Date: Thu, 24 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: George Richardson <gr383@cnsvax.albany.edu> (by way of k-12sd)
Subject: Re: correctness of models
On Wed, 23 Sep 1998, Mary Ellen Verona wrote:
> Thanks George and Ed for *lots* of food for thought. George - about using
> theory to guide modeling - yes, I am concerned about when this should be
> done, since I think it is important. So often I have seen teachers using
> graphical analysis and automatically picking the "best fit" giving a
> functional form that does not relate at all to what they know
> theoretically.
I, too, would say teachers who pick a "best fit" functional form that
contradicts what they know theoretically are making an unfortunate choice.
It may be forced on them because they can't make the theory accessible to
their students, so they chose to ignore it, but it's still unfortunate.
BUT, in those cases where relatively simple thought experiments can reveal
weaknesses in say, a linear fit, I would beg the teacher to think through
the thought experiments with their kids and amend the functional form as
best the group can with their added insight.
There is a long, strong tradition of thought experiments in science, and
they have become a crucial tool for system dynamics model builders. In
the science traditions we have Maxwell's demon and Einstein's lovely
thought experiments about an elevator accelerating at g and moving near
the speed of light. [From the latter Einstein got gravity to bend light!]
In system dynamics our "extreme condition tests" are such thought
experiments, and they are crucial for uncovering weak formulations. What
if my workforce goes to zero? Will the system behave properly? What if
some resource doubles or is halved? Will the dynamics of the model be
credible?
I'd say systems modeling gives us as teachers a vivid way to help people
learn some crucial "scientific thinking" skills. One of those skills may
be curve fitting. But a much more powerful one is thought experiments.
...George
-----------------------------------------------------------------------
George P. Richardson G.P.Richardson@Albany.edu
Rockefeller College of Public Affairs and Policy Phone: 518-442-3859
University at Albany - SUNY, Albany, NY 12222 Fax: 518-442-3398
http://cnsvax.albany.edu/~gr383/
Date: Thu, 24 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: George Richardson <gr383@cnsvax.albany.edu> (by way of k-12sd)
Subject: Re: correctness of models
On Wed, 23 Sep 1998, Ed Gallaher wrote:
> When teaching the RB materials I assume absolute beginners with no
> knowledge of SD or modeling. I want to make it VERY simple.
[...]
> You'd think the simplest output
> would be linear (a pump), but this leads to the same problem Mary Ellen
> describes, with possible negative volume.
>
> The solution to this dilemma is absolutely NOT to use the non-negative
> function in STELLA!! The solution is to determine the correct physical
> structure of the model to accurately describe what is happening.
Indeed! Relying on STELLA's nonnegative constraint means your model
leaves out some important part of reality.
> The solution is to insert an IF-THEN statement. ...
Ed goes on to say that this sort of thing is frowned upon by "programmers"
[we should say "modelers"] and should be rarely used but is appropriate
here.
I don't think it's appropriate even here. I'd disagree that the nicest,
or even simplest way out of this is to use an IF-THEN here. And I don't
think it fits simple observations about what water pouring and then
dribbling out of a pipe looks like.
I'd use a graphical function (of, say, the height of the water in the
burette) to modify the outflow rate down to zero. The equation might be
OUTFLOW = MAX OUTFLOW * EFFECT OF HEIGHT ON OUTFLOW
where the EFFECT OF HEIGHT is a graphical function the kids in class could
argue over and experiment with. It has two certain values: When the
burette is full (height = 100 mm or whatever), the EFFECT = 1 so that the
OUTFLOW is maximum. When the burette is empty, the EFFECT = 0. In
between we can theorize and debate and try different functional forms.
[In the graphical function I just tried, I had a very slight linear
decline from 100 to 10 mm, followed by a steeper and steeper decline to 0
at 0 mm. Try it -- the dynamics over time of the Water in the burette and
the outflow are a bit surprising.]
The important things here are
1) Graphical functions as multiplers like this are the modifying tools we
want young (or old!) modelers to learn. Most of the systems we are
dealing with are aggregate systems in which an IF-THEN may work for one
actor in the system but be utterly meaningless for many actors. We need
to wean people away from IF-THEN thinking in social system modeling, and
I'd suspect also in natural science modeling.
2) The graphical function formulation draws kids in naturally to
combining theory (the 1 point and the 0 point) with observation (what
actually happens to the stream as the height declines? what happens near
the end?).
I think we'd get Mary Ellen's theory/observation stuff pretty
automatically arising in the classroom from the kids themselves. And I
would think that would be cool.
...George
-----------------------------------------------------------------------
George P. Richardson G.P.Richardson@Albany.edu
Rockefeller College of Public Affairs and Policy Phone: 518-442-3859
University at Albany - SUNY, Albany, NY 12222 Fax: 518-442-3398
http://cnsvax.albany.edu/~gr383/
Date: Thu, 24 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Ed Gallaher <gallaher@teleport.com> (by way of k-12sd)
Subject: Re: correctness of models
>Thanks George and Ed for *lots* of food for thought. George - about using
>theory to guide modeling - yes, I am concerned about when this should be
>done, since I think it is important. So often I have seen teachers using
>graphical analysis and automatically picking the "best fit" giving a
>functional form that does not relate at all to what they know
>theoretically.
>
You are absolutely on track with this concern! This has been done in
biomedical science to some extent, and more often in behavioral
(psychology) research. There are extensive statistical tests to help one
decide whether the data fit to one equation is significantly better or
worse than that fit to another. This "statistical decision" can be made
with not a shred of underlying theoretical justification!
There are a few instances when this procedure is actually OK (but very
few). e.g. When we look at brain slices with radioactive drug attached we
want to know the density (and thus the concentration) of the drug in
various regions. We run a set of known radioactive standards on the film
along with the brain slice. We see an increasing density (darker gray)
with increasing radioactivity (8-10 data points), but the shape of the
curve is very complicated due to exposure times, silver particle density,
linearity of the film, etc. etc. The point here (and it is the ONLY point)
is that we want an equation, any equation, that lets us reliably
interpolate between the standard data points. We thus search for an
equation that fits the data well, with NO theoretical reason to choose one
equation over another.
This is one of the few examples I can think of where a somewhat mindless
curve-fitting routine actually provides us with something useful. Most
other cases fall into the category that Mary Ellen describes above.
Ed Gallaher
Date: Thu, 24 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Ed Gallaher <gallaher@teleport.com> (by way of k-12sd)
Subject: Re: IF-THEN statements
This is in response to George's comments about If-then statements:
I hope this isn't getting too convoluted, but I may not have been clear in
my earlier message.
Mary Ellen inquired about water running out of a buret. I agree with
George here; this is NOT an appropriate place for an if-then statement.
The dynamics should be modeled carefully.
The example that I -do- think is appropriate (but I want to know if George
concurs . . . ) is not where water is dribbling out of a tank, but where a
pipe extends down into the tank to pump water out. As long as the water
level is above 4 gallons (the bottom of the pipe), the pump will remove
exactly 2 gallons per minute (constant; not a function of volume). As soon
as the water drops below the end of the pipe the pump will begin to suck
only air, i.e. 0 gallons pumped.
Therefore:
IF VOLUME > 4 THEN 2 ELSE 0
It seems to me this is the simplest, clearest, most accurate description of
the actual dynamics here.
George?
Ed Gallaher
Date: Thu, 24 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Quaden@aol.com (by way of k-12sd)
Subject: Re: Re: Greetings
Charis,
Niall's question is exactly the same as mine. In addition I would like to hear
how you developed the theory. In other words what events lead you to see the
development of this structure? I just joined the sysdyn group, so I dont' know
if you mentioned this in your original statement.
Rob Quaden
Carlisle Public Schools
Date: Thu, 24 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: "Jay W. Forrester" <jforestr@MIT.EDU> (by way of k-12sd)
Subject: Water flow through an orfice
My 65-year-old hydraulics text book has a 25-page chapter on water flow
through an orfice. An orfice is defined as a sharp-edged hole, or a hole
in a plate that is thin compared to the diameter of the orfice.
Simplifying the chapter, one can say that in general velocity is
proportional to the square-root of the height of the water. Of course,
that leads to zero flow at zero height.
---------------------------------------------------------
Jay W. Forrester
Professor of Management, Emeritus
and Senior Lecturer, Sloan School
Massachusetts Institute of Technology
Room E60-389
Cambridge, MA 02139
tel: 617-253-1571
fax: 617-258-9405
Home office:
tel: 978-369-9372
fax: 978-369-9077
Date: Thu, 24 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: <dburke@nsf.gov> (by way of k-12sd)
Subject: Greetings
Hi.
My name is Dan Burke and I am a Senior Staff Associate in the
Directorate for Education and Human Resources, National Science
Foundation. Among my responsibilities is to work with our education
systems reform programs, including the Urban, Statewide and Rural
Systemic Initiatives. These are multimillion dollar awards to urban
school districts, states, and consortia of rural districts
respectively. As a condition of these awards we require the recepient
to reform their K-12 math and science curriculum, student assessments,
professional development, funding streams, policies, and community and
other partnerships. As a biologist by training, a systems approach
and modeling seems quite natural, but what I have seen is that while
isolated individuals in many school districts have good ideas
concerning the pieces of a system, few really understand what it means
to think about the system as a whole. What I would like to do is
involve our awardees in systems thinking and modeling as they plan and
implement their reforms. I am developing a few simple models to
demonstrate and I would be very interested in learning about school
districts that use system dynamics in their planning that I could use
as examples.
Thanks,
Dan Burke
Date: Thu, 24 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Linda Booth Sweeney <Linda_Booth_Sweeney@harvard.edu> (by way of
k-12sd)
Subject: Stories as a Means of Exploring Causality
Greetings to you all --
I am conducting a semester-long research study on "Stories as a Means of
Understanding and Exploring Causality." The work is being done in
conjunction with Harvard's Project Zero.
Here's my question for you:
What stories can or do educators use to help students explore causality?
For example, Dr. Zeuss's The Butter Battle Book is a story of escalation
(similar to the arms race between the US and Russia). The Giving Tree can
be viewed as an example of the Tragedy of the Commons archetype, a
universal pattern of exhausting limited resources (i.e. the Georgia's
fishbanks).
For each story, I will give a brief description of the story line and
identify framings and insights relevant to causation and feedback. The
results will then be used as part of a larger research project.
Any ideas? As in the past, I'm happy to send you the results of my work -
especially if you are able to contribute story ideas.
My best, and thank you in advance,
Linda
Linda Booth Sweeney
Harvard Graduate School of Education
18 Fernald Drive #22
Cambridge, MA 02138
617-354-1390
e-mail: Linda_Booth_Sweeney@harvard.edu
Date: Fri, 25 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: George Richardson <gr383@cnsvax.albany.edu> (by way of k-12sd)
Subject: Re: IF-THEN statements
Are we all having fun or gaining something from this?
If so, Ed has asked for a response, so...
On Thu, 24 Sep 1998, Ed Gallaher wrote:
> The example that I -do- think is appropriate (but I want to know if George
> concurs . . . ) is not where water is dribbling out of a tank, but where a
> pipe extends down into the tank to pump water out. As long as the water
> level is above 4 gallons (the bottom of the pipe), the pump will remove
> exactly 2 gallons per minute (constant; not a function of volume). As soon
> as the water drops below the end of the pipe the pump will begin to suck
> only air, i.e. 0 gallons pumped.
I see now this is a different situation, in which we've tried to remove
the effect of the height of the water in the tank. I'd say the IF-THEN
probably comes closest to what we could get a simulation to do to mirror
the actual situation. So I'd say Ed's right. (By the way, I know almost
nothing about fluid dynamics, except what I observe, so my opinion on the
accuracy of this approach here is not worth much! For all I know the
height of the water in the tank still exerts an influence on the upward
flow, since the pump is actually lifting more water when the height of
the water in the tank drops [effective length of the pipe is greater, so
more mass to lift, so it goes slower?!?].)
I _do_ want to emphasize that we don't want to teach kids (or my graduate
students) to get in the habit of using IF_THENs for model structure. They
have a role as test inputs, scenarios, policy switches at some point in
time, and so on, but generally they are inappropriate as model structure
trying to capture the aggregate action of multiple decision makers
operating independently at a decision point.
And we don't need more than a few good uses of IF-THENs to convince kids
(and my graduate students) that you ought to _always_ use an IF-THEN. We
think in IF-THENs naturally, so we easily slip into the error of excessive
use of IF-THENs in models.
So I'd like us to pick natural science modeling situations, and social
science modeling situations, that move our students toward good practice.
If IF-THEN is the only good way out of the example Ed (and Mary Ellen?)
were talking about, then I think I'd hunt for another example that
reinforces better modeling habits.
...George
Date: Fri, 25 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Ed Gallaher <gallaher@teleport.com> (by way of k-12sd)
Subject: Re: Greetings
YESSS!
I am very encouraged to see this approach being considered from the point
of view of a person involved in evaluating propsals, rather than from those
trying to obtain the funding!!
The tide seems to be changing, but within the biomedical research realm
(mostly National Institutes of Health funding) there has been a distinct
prejudice AGAINST using modeling in a grant proposal; in fact I have been
told in no uncertain terms by some very knowledgeable individuals at the
review level in NIH that this would be the kiss of death for a grant. The
correct "grantsmanship" procedure is to (a) listen to this advice, (b)
write a grant that will get funded, and (c) do what you want with the
modeling after getting the money. The grant that will get funded provides
experimental pilot data, states clear hypotheses, provides a clear
description of the lab experiments that will be done, and considers the
various results that might be obtained.
This is a practical approach, and I've been using it successfully in a
number of areas for about 20 years. At some point, however, it begins to
be counterproductive. If the goal is -specifically- to demonstrate that
System Dynamics can be a useful tool, then one must convince the study
section to accept this as part of the specific aims of the grant. Then,
and only then, will they consider one's work successful when a renewal is
submitted. Otherwise, they will consider the experimental work, and only
the experimental work to be successful, and will STILL interpret the
modeling component to be "fluff", or a nice hobby that the researcher
engages in.
Recently I have bit (bitten?) the bullet and submitted grants over and over
again with the modeling component intact. This has been frustrating and
very hard work, but it finally is beginning to pay off with grant funding
that -formally- includes modeling. (I might add that moral support from SD
colleagues, Nan, Jay, Ron, Diana, Tim, Scott, Janet, Mary, Jim Hargrove,
the Waters Foundation, and many others has had a huge impact on my
continuing enthusiasm and work in this area! Thanks to all!!)
Now I see that Dan Burke is encouraging the use of models WITHIN the
context of education reform, to be recognized and encouraged by a major and
influential funding agency.
Hooraaay!
Ed Gallaher
>Hi.
>
>My name is Dan Burke and I am a Senior Staff Associate in the
>Directorate for Education and Human Resources, National Science
>Foundation. Among my responsibilities is to work with our education
>systems reform programs, including the Urban, Statewide and Rural
>Systemic Initiatives. These are multimillion dollar awards to urban
>school districts, states, and consortia of rural districts
>respectively. As a condition of these awards we require the recepient
>to reform their K-12 math and science curriculum, student assessments,
>professional development, funding streams, policies, and community and
>other partnerships. As a biologist by training, a systems approach
>and modeling seems quite natural, but what I have seen is that while
>isolated individuals in many school districts have good ideas
>concerning the pieces of a system, few really understand what it means
>to think about the system as a whole. What I would like to do is
>involve our awardees in systems thinking and modeling as they plan and
>implement their reforms. I am developing a few simple models to
>demonstrate and I would be very interested in learning about school
>districts that use system dynamics in their planning that I could use
>as examples.
>
>Thanks,
>Dan Burke
Date: Fri, 25 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: LB <lynne@byramhills.csnet.net> (by way of k-12sd)
Subject: Re: correctness of models
In my profession (which is large scale system design), it is common to use
several
different models simultaneously to analyze a problem and define the
solutions. As
you say, no single model is perfect. By using more than one, each with a
different emphasis and "flavor", we are able to create a composite picture
that is
more complete and accurate than otherwise possible. An interesting dynamic
developes among the different models that make up the composite when they are
keyed to each other; it works almost like a prototype.
Lynne Bernstein
by way of k-12sd wrote:
> Hi,
>
> Mary Ellen Verona
> mverona@mvhs1.mbhs.edu wrote:
>
> >Here is my question - how do we determine whether to introduce this more
> >complicated formulation - which of course also neglects many factors of
> >flow? There are more questions - e.g. the idea above introduces some
> >minimal physical theory to guide curve fitting. When and how do we need
> >to do this?
>
> Don't lose sight of the fact that all models neglect some factors, be they
> known or unknown to us. Some models are more complete than others, but even
> very crude models may be useful to us. There is never a complete model or a
> "right answer." It is important to stress this to students. We tend to bring
> children up to believe that there is always a "right answer" to any question.
> This is, IMHO, a good opportunity to introduce the notion of a "good enough
> answer" to the students. They simply must understand that all models have
> shortcomings, and that the more we know about the shortcomings, the more
> useful the model becomes. From there they can move forward. Surely by the
>time
> a student reaches high school he/she can understand the concept and some of
> the implications of the square of a number, so that should not be an
> impediment. I have introduced squares of numbers (and square roots) to
> children as young as 7-8 with satisfying results. Don't expect them to
> completely grasp it at that age, but 7-8 is not too early to introduce the
> concept with some simple application. Perhaps you would like to have the
> students experiment with both the linear and sqrt models and decide for
> themselves the utility of each compared to the amount of effort expended and
> desired accuracy? How about letting them decide what the important questions
> are, such as "What range of burette volume are we interested in
>modeling?" Are
> we really interested in that last drop? It all should be put into context.
> Good luck!
> Art Seaman in Jackson, TN
Date: Fri, 25 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Ed Gallaher <gallaher@teleport.com> (by way of k-12sd)
Subject: Re: IF-THEN statements
>Are we all having fun or gaining something from this?
(I am! ejg)
>If so, Ed has asked for a response, so...
>
>On Thu, 24 Sep 1998, Ed Gallaher wrote:
>
(By the way, I know almost
>nothing about fluid dynamics, except what I observe, so my opinion on the
>accuracy of this approach here is not worth much! For all I know the
>height of the water in the tank still exerts an influence on the upward
>flow, since the pump is actually lifting more water when the height of
>the water in the tank drops [effective length of the pipe is greater, so
>more mass to lift, so it goes slower?!?].)
(GPR is probably right here, but I ignore this influence with the
simplifying assumption that the pump flow is constant. It certainly is
likely that the pump flow is actually influenced by the pressure head.)
>I _do_ want to emphasize that we don't want to teach kids (or my graduate
>students) to get in the habit of using IF_THENs for model structure. They
>have a role as test inputs, scenarios, policy switches at some point in
>time, and so on, but generally they are inappropriate as model structure
>trying to capture the aggregate action of multiple decision makers
>operating independently at a decision point.
ABSOLUTELY!
>And we don't need more than a few good uses of IF-THENs to convince kids
>(and my graduate students) that you ought to _always_ use an IF-THEN. We
>think in IF-THENs naturally, so we easily slip into the error of excessive
>use of IF-THENs in models.
>
>So I'd like us to pick natural science modeling situations, and social
>science modeling situations, that move our students toward good practice.
>If IF-THEN is the only good way out of the example Ed (and Mary Ellen?)
>were talking about, then I think I'd hunt for another example that
>reinforces better modeling habits.
I agree completely. This is why I do -not- use the example of an output
pump when teaching the rain barrel. It provides a good example of very
simple MATH (linear decline of volume), but adds a number of not-so-simple
modeling concepts.
Since my real goal at that stage is to teach very simple -modeling-, I
mention the linear output briefly in the oral discussion and then move past
it to the exponential outflow.
From a pedagogical point of view I believe George is suggesting that early
teaching should not demonstrate what we -don't- want our students to do,
followed by a bunch of reasons why they shouldn't do it. Rather we should
demonstrate what we -do- want them to emulate, and hope that they absorb
the best modeling practice as soon as possible. I agree with this.
Ed Gallaher
Portland, OR
Date: Fri, 25 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Mary Ellen Verona <mverona@mvhs1.mbhs.edu> (by way of k-12sd)
Subject: Re: correctness of models
I love the graph idea! Certainly we use graphical factors when we don't
know what else to do - but of course that's the point: looking at the
problem from the kid's point of view - .
Mary Ellen Verona
mverona@mvhs1.mbhs.edu
***** new address ******
Maryland Virtual High School
Montgomery Blair High School
51 East University Boulevard
Silver Spring, MD 20901
301-649-2880
Date: Fri, 25 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: "Janice C. Kowalczyk" <kowalcjn@ride.ri.net> (by way of k-12sd)
Subject: Re: Stories as a Means of Exploring Causality
Linda,
I share the same interest you do in Stories. I have been working on a
multi-stage Stella model of Dr. Seuss's "The Lorax" that I hope to be able
to share with others in the future. This book is an excellent book for
such discussions. I would love to see your final list of suggested books
and be happy to share my model with you at some point.
Janice Kowalczyk
>Greetings to you all --
>
>I am conducting a semester-long research study on "Stories as a Means of
>Understanding and Exploring Causality." The work is being done in
>conjunction with Harvard's Project Zero.
>
>Here's my question for you:
>
> What stories can or do educators use to help students explore causality?
>
>For example, Dr. Zeuss's The Butter Battle Book is a story of escalation
>(similar to the arms race between the US and Russia). The Giving Tree can
>be viewed as an example of the Tragedy of the Commons archetype, a
>universal pattern of exhausting limited resources (i.e. the Georgia's
>fishbanks).
>
>For each story, I will give a brief description of the story line and
>identify framings and insights relevant to causation and feedback. The
>results will then be used as part of a larger research project.
>
>Any ideas? As in the past, I'm happy to send you the results of my work -
>especially if you are able to contribute story ideas.
>
>My best, and thank you in advance,
>
>Linda
>Linda Booth Sweeney
>Harvard Graduate School of Education
>18 Fernald Drive #22
>Cambridge, MA 02138
>617-354-1390
>e-mail: Linda_Booth_Sweeney@harvard.edu
Janice C. Kowalczyk
Educational Math Specialist, Rutgers University
"We are continually faced by great opportunities brilliantly disguised as
insolvable problems"
Date: Mon, 28 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Ed Gallaher <gallaher@teleport.com> (by way of k-12sd)
Subject: Re: correctness of models
>On Thu, 24 Sep 1998 16:13:50 -0400 gallaher@teleport.com (Ed Gallaher) wrote:
>
>>The point here (and it is the ONLY point)
>>is that we want an equation, any equation, that lets us reliably
>>interpolate between the standard data points.
>
>I'm wondering if you think that in general, the mindless curve-fitting is
>OK in
>this sort of application, and would you say that the point of the
>interpolation
>is to translate between known and unknown?
Yes, I do believe it is OK in this specific instance. It bothered me when
it was first explained to me because I assumed it was just more of the same
mindless curve fitting that we often see. But yes, I like your
explanation; the point of the interpolation is to provide a smooth link
between the x- and y-axis values on the graph, and NOTHING MORE!
ejg (Ed Gallaher)
>David Gibson <gibsond@quark.vsc.edu>
>VISMT Professional Development Specialist (802) 244-8768
>Montpelier Schools Director of Curriculum (802) 225-8070
>U.S. Dept of Education WEB Project Director (802) 229-4660
Date: Mon, 28 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: Rolfe Stanley <rstanley@together.net> (by way of k-12sd)
Subject: modeling, water in a bucket etc.
I have been reading the continued discourse on "water in a bucket" and
funding of modeling grants. My 2 cents for what it is worth:
Water in a bucket. If water is coming out an opening in the base of the
bucket or at any level then the rate is controlled by the pressure of the
water above the hole. This value is the density of the fluid*acceleration
of gravity*height of the water above the based of the hole. The first 2
values remain constant during the simulation but height decreases to zero.
Thus in my mind the rate of flow out the opening decreases exponentially.
I have a simple simulation in STELLA that expresses this concept if anyone
is interested. Incidentally, the hole can be thought of as a decay
constant, while the water in the bucket can be thought of as a "parent". If
we collect the water in another bucket then this amount at any one time is
the "daughter". The simulation then becomes an expression of an
irreversible reaction and is a good way to illustrate radioactive decay.
If you are interested in the model I send it to you (binhex code Macintosh).
Funding and modeling. I advise graduate students during their research
proposal writing. During this time I try to have then develop a STELLA
simulation that is a valid representation of their research so that they
can use it as a "platform" for testing multiple working hypotheses. Thus
modeling, in my mind, becomes a very valuable effort in that it opens up a
whole range of hypotheses that can be tested.
Rolfe Stanley
Stanley Computer Center
Fletcher Extension
Date: Tue, 29 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: sthompson@foundation.panasonic.com (by way of k-12sd)
Subject: intro
Greetings -- This message is both a self-introduction and a request for
information.
I'm Scott Thompson, Assistant Director of the Panasonic Foundation. I
started my career as a high school English teacher, then went to work in
the communications division of an international church organization, and
more recently worked as the Director of Dissemination and Project
Development at the Institute for Responsive Education -- a non-profit
focused on reforming K-12 education with an emphasis on
family-school-community collaboration. I've been with the Panasonic
Foundation for two-plus years. It is a corporate philanthropy, which for
more than a decade has devoted all of its resources to whole-system reform
of K-12 education. Unlike most foundations, Panasonic does not make
grants. Rather we form long-term partnerships with school districts, and
teams of consultants with expertise in systemic educational change provide
technical assistance to those districts at foundation expense. We also
bring teams from each partner district together for a two-and-a-half day
annual conference. It is our conviction that, while much has been learned
about reform at the individual school level, the lasting and widespread
changes needed in a nation (the United States) with two and half million
teachers will never be achieved until we learn how to transform whole
systems of schools into learning communities. This means that people
throughout the system will need to learn how to let go of the bureaucratic,
factory model of educational administration and be promoters and supporters
of semi-autonomous schools that are learning communities. The dynamic
complexities of change on this scale require leaders to be system thinkers
-- thinking and acting strategically and collaboratively. (If anyone is
interested in knowing more about the foundation and its work in more than a
dozen districts across the U.S. I have printed materials that I'd be happy
to mail free of charge. I'd just need a postal address. You can also read
about us on line at http://www.panasonic.com/foundation. The on line
information is, however, fairly limited at this time.)
Part of my work at Panasonic involves editing our two newsletters. On of
these is called Strategies for School System Leaders on District-Level
Change. It is published by the foundation in collaboration with the
American Association of School Administrators and reaches their
more-than-13,000 members. Our forthcoming issue will look at the
relationship between systems thinking/organizational learning and systemic
educational change. Each issue includes a number of case studies of
districts engaged in district-level changes relating to that issue's theme.
For the last couple of months I've been doing lots of networking and
learning about school and district-level efforts along these lines. At
first nearly every example I ran into involved the use of system dynamics
at the classroom level -- aimed at helping students become system thinkers
with perhaps some effort to apply organizational learning disciplines at
the whole-school level, but with very little in the way of full-system,
district-level applications. However, as I've persisted examples of
districts applying systems thinking disciplines across the system have come
to my attention. In fact, the outlines of an interesting issue are
beginning to take shape. At the same time, there may be folks on this
listserve who know of districts that are further along or are more creative
or deep or sophisticated or whatever than the ones we've come up with. I
would welcome recommendations. If you have school districts to recommend,
could you provide a contact name and phone number?
In case you'd like to check out back issues, here's the on-line address for
Strategies: http://www.aasa.org/Pubs./strategies/contents.htm.
In case anyone is interested in our profiles of schools engaged in deep
reform efforts, I'll also provide an on-line address for the other
newsletter I edit -- P(superscript: 3 )(that's "P-cubed" for Panasonic
Partnership Program): http://thechalkboard.com/panasonic/com.
My thanks to the sponsors of this listserve for providing this avenue for
networking.
Scott Thompson
Date: Tue, 29 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: "Jean-Louis Cordonnier" <jlcord@wanadoo.fr> (by way of k-12sd)
Subject: re : correctness of models
Hi all.
My way of teaching is prompted by the french philosoph Gaston Bachelard.
He wrote :
"Donner et garder un intérêt vital à la recherche, tel est le premier devoir
de l'éducateur, à quelque stade de la formation que ce soit" (in La
formation de l'esprit scientifique 1938, p9)
Mary Ellen asks "how do we determine whether to introduce this more
complicated formulation... ?"
I would better say "is it the wright time to make them investigate (further)
?".
I had this week to teach about underground water, effects of well pumping,
springs...
During the investigating process, I asked my students (16-17 years old)
about the "rain barrel" (my barrel was a long vertical pipe ) : "Describe
exactly what will happend ?"
("On ne peut étudier que ce qu'on a d'abord rêvé" G Bachelard ; in
psychanalyse du feu p 12)
Allmost all of them thought the flow would be constant. The experimentation
showed them they were wrong ("On ne voit que parce qu'on prévoit"). The
question became : "How can we explain the decreasing flow ?".
My pipe was a one meter long plastic pipe, not transparent (only because I
had allready this one & couldn't find another). So they had to imagine the
water level in the pipe. After a discussion they changed their first
oppinion to " the flow is a decreasing function of pressure".
The crucial point is that they had to change their mind, and "climb one
step" because the "real world" dit not fit exactly to their first belief.
I would say that the wright time to introduce a more complicated model is
when it forces to investigate further.
Mary Ellen's question was a very good one because it forces us to
investigate. Everyday I read first the following of this tremendous serial
story !!!
I used then an aquarium full of coarse-grained sand, filled with water. Like
undergroundwater in the rocks. I opened the plug. And the water came out...
How ?
Jean-Louis Cordonnier
36, rue Lavisse
66000 PERPIGNAN
jlcord@wanadoo.fr
Date: Tue, 29 Sep 1998
To: k-12sd@sysdyn.mit.edu
From: RICHARD TURNOCK <Richard_Turnock@pgn.com> (by way of k-12sd)
Subject: Grant money
Oregon is having a Partner's Fair on November 9 to network people from
businesses, consultants, academics, with the local school districts and ESD
staff. Following the Partner's Fair there is a bidders conference telecast
on closed circuit TV throughout the state. The state has Goals 2000 money
and is targeting $3.75 million to professional development for teachers.
I work for PGE, an electric utility in Portland. Before I go to the
conference, I would like to communicate with educators about how to
approach school districts and ESD staff about funding system dynamics
training for teachers. I would like to have an alliance between my company
and one or more school districts (maybe ESD) to offer System Dynamics
training for teachers. Anyone in Oregon reading this and interested in
an alliance, please call me at 503-464-8503.
The grant money would fund the cost of the instructor and any expenses or
facility costs. System Dynamics course material is the same as the summer
workshop course published as part of NSF grant by Ron Zaraza. The course
would be taught for 10 weeks, 3 hours once a week in the evening. Teachers
would get continuing education credit through local university.
Thanks
Richard Turnock
Portland General Electric 1WTC0903
121 SW Salmon St.
Portland OR 97204
phone: 503-464-8503
FAX: 503-464-2223
richard_turnock@pgn.com
www.pge-edsvcs.com
End of September