October 1999
Date: Thu, 30 Sep 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: Ed Gallaher <gallaher@mail.teleport.com>
Subject: Re: Teaching systems
>From: G P Richardson <gpr@csc.albany.edu>
>I'm worried about the word "illustrate" here.
As usual, GPR picked up on a "trivial" point, and presented some important
and articulate recommendations for the use of actual computer simulations
in the classroom. Thanks, George!
Teresa and Larry offer some additional comments.
I do believe that stock-and-flow diagrams per se provide us with a largely
unrecognized "language" which can allow us to talk intelligently about the
structure of systems. There are several caveats, however.
1. If I am discussing a dynamic problem with colleagues who I know are
well-versed in modeling AND simulation, then I can (reasonably safely)
assume that they will not 0ver- or under-interpret what I am trying to
express. This assumption can be very misleading if this colleague has not
actually built, tested, simulated, revised (etc.) their own original model
structures and simulations.
2. I believe I could lead an informative discussion with a group of Boy
Scouts, in which we discuss mammalian thermoregulation, energy sources,
heat production and loss, insulation, etc., with the goal of understanding
and preventing (potentially lethal) hypothermia. I believe I could provide
the simplest description of stocks and flows, and then interactively
develop a reasonable diagram which illustrates the underlying biology and
physics.
HOWEVER, this does NOT mean they could do the same thing in a small group
discussion, without the facilitation of a skilled system dynamicist (if I
can modestly put myself in that category . . . ).
This raises a critical issue. To what extent could a well-meaning biology
or health teacher use stock-and-flow diagrams -correctly-, in the absence
of some significant system dynamics training? As we all know, this can
become a slippery slope. The -discussion- may be meaningful, but the S&F
diagram that ends up on the blackboard may be meaningless. This leaves the
students, and maybe the teacher, with the impression that attention to
detail is not that important.
A major goal of SD is to encourage crystal clear, rigorous, accurate
thinking about the structure and dynamic behavior of systems. I suggest
that we need to ardently defend this goal!
3. I fully agree with George (and Jay of course) and others, that real
understanding comes from running simulations. With any complexity at all,
they rarely perform as expected. At this point, the development of insight
-begins-.
4. George suggests that some very simple models be used to illustrate early
in the process that (a) we often observe unexpected dynamics, and (b) yes,
we really do need to run simulations to uncover these surprises.
Here us an example from Roadmaps that hit me between the eyes: (a) A town
has a forest area that is harvested modestly on an annual basis. (b) The
school has an ecology program that plants new trees to replenish those that
are harvested. (c) A finite time is required for tree maturation (I seem to
remember 50 years). The harvesting, replanting, and maturation is in
balance so the forested area remains stable.
NEXT, there is an increased need for harvesting. The number of trees
removed increases, so the school plants more trees to replenish them.
Therefore, the number of trees remaining in the forest stays the same for
the children and grandchildren to enjoy, right? WRONG!!
This is a -very- simple model. It carries with it a surprising and very
important message. It -cannot- be predicted intuitively, but is immediately
obvious when simulations are conducted. And in the right setting it should
provide beginning students with a clear understanding that SIMULATIONS are
very important.
Ed Gallaher
Date: Thu, 30 Sep 1999
Subject: Re: Teaching systems
From: "Lawrence Weathers" <weathers@massed.net>
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
George,Sorry if I wasn't clear. I agree that models eventually should not
be black boxes to kids ( or anyone else) because they can be opened up.
But, I think, until the confidence is built up about them (and the
skepticism as well) they ARE black boxes for a while. The details of the
stock/flow and causal structure are not necessarily apparent to the
beginner, especially the young student beginning to understand causality
and consequence in other aspects of his or her life. Even though the
person acquainted with them may know how to look inside them, the newcomer
is likely to see only the parts, not necessarily the whole. Like a car's
engine that someone peers into and tries to make meaning of, the intuitive
confidence (and accompanying skepticism) is not fully developed for a
while. Not until the parts are tinkered with, and the predictability of
behavior is established, that one can start extrapolating to untested
scenarios. But, yes, I agree that this is more true for some kids than
others. As in "neo-piagetism", there are surely developmental stages, but
different kids go through them at different times. And I neglected to add,
but you picked up on, the fact that a little experience with "bad" models,
"illustrating" things we know to be untruths, is a healthy step in
developing the skepticism. I absolutely agree with you about the balance
of too much and too little illustrating. Sounds like a delicate shift of
loop dominance at some point. A teacher has to be watchful. And, yes, more
input is needed from those having experiences with young kids. I agree with
Dan, too, that a short list of simple counterintuitive models, suitable for
use with kids, would be a great tool. Thanks for the insightful comments -
thought provoking and healthy to ponder.
Larry ----------
From George Richardson:
k-12sd wrote: Thanks for the replies to my little note about
"illustrating." Two comments on the latest message (from Larry), the first
of which may be another possible case of my focusing too narrowing on the
words we use. Larry said (and I agree)... > there is benefit in at least
some short period of illustration, where > models can parallel the thinking
we go through in resolving issues that are > simple enough to reason
through without models. Then he said... > The benefit lies in building a
"confidence" in what appears to be a > black box, a model, and what it can
tell us. Larry's wording (there I go again) suggests the words "model" and
"black box" are synonyms. I earnestly hope not! (But again, maybe he
didn't mean it that way.) A "black box" model to me is a model we can't
"open up," to see the details of stock-and-flow/causal structure inside.
System dynamics models aren't "black box" models unless we wrap them in an
interface that hides what's going on inside (which I hope we don't do much,
when we're trying to help kids UNDERSTAND something). But maybe Larry DOES
mean that a simple STELLA model with its stocks and flows and causal links
and feedback loops all out there for God and everybody to see is still a
"black box" to kids (some kids?). If that's what experienced K-12 teachers
think is true for kids, I'd like to know more about that. What mysteries
remain in kids' brains when we first develop with them a simple model, with
all its assumptions (presumably) laid out for them to see, and we simulate
it? What's still "black"? Finally Larry said... > If we dwell too long on
only illustrating, we may build only the > "confidence" and not the
parallel skepticism necessary in order to use the > models to shape our
"new learning" to its fullest. Presumably, that good skepticism comes after
playing and learning with models for a while, seeing good models and bad
ones. But the earlier posting to the list that started all this suggested
if we stick to "illustrating" with simulation we may not build the
necessary motivation in kids to keep at this systems stuff 'til it bears
fruit for them. Sounds like nasty competing forces at work here -- too much
illustrating yields low or no motivation, not enough illustrating yields
low or no understanding, and low motivation (from lots of illustrating)
also yields little understanding (from boredom). I hope you all have
figured out how to balance these. What cues from the kids do you use?
...GPR
-----------------------------------------------------------------------
George P. Richardson G.P.Richardson@Albany.edu
Chair of Public Administration and Policy http://www.albany.edu/~gpr
Rockefeller College of Public Affairs and Policy Phone: 518-442-5258
University at Albany - SUNY, Albany, NY 12222 Fax: 518-442-5298
From: j.seward@sbu.ac.uk
Date: Fri, 1 Oct 1999
To: k-12sd@sysdyn.mit.edu
Subject: Introduction
My name is John Seward and I have recently joined the team of Dr. Dembe
Williams at South Bank University. I am here as a tutor in Management and
Business Systems. My background is in the car industry and then in the
computer industry, mainly US based companies, me operating in a European
Marketing position.
I am stimulated by the word "surprise." Much of the management education
and training
is about the predictable, the correct procedures. This turns inspired
engineers into
clerks. As in the mission statement, people will diligently follow a system
etc, but it takes the genius
of you good people to see that it may just be the system that has a tiny
defect or two, like my
typing! Ask any communist, member of the CIA or even our beloved
Conservative Party.
The skill element is to be prepared for the unexpected. A tautology in itself.
Any thouights thoughts, anyone? someone?
John Seward
Date: Fri, 1 Oct 1999
From: G P Richardson <gpr@csc.albany.edu>
To: k-12 listserve <k-12sd@sysdyn.mit.edu>
Subject: Simple models with unexpected or puzzling dynamics
Here is the beginnings of a list of simple models that produce
what we might call "counterintuitive" behavior -- behavior over time that
is unexpected, puzzling even when you see it displayed in a simulation, or
difficult or impossible to understand without simulation (maybe even hard
to understand WITH simulation).
A sad admission first: These little models and insights may not appeal to
little kids, or seem surprising or insightful. (Maybe my graduate
students and I are more easily surprised.) But I think most of these work
with high school kids, maybe middle school kids.
But here's a start:
1) URBAN1: 3-level model (23 equations) exhibiting growth, stagnation,
and decline of a typical city (Alfeld & Graham). Question: Why does
model experience urban decline, and what does understanding that tell you
about real cities? (Most people's first answer to this is wrong, and some
people can spend hours working with this small model trying to figure it
out how to get the model not to decay. Hard to get robust real-system
insights from 23 equations, but this model can do it.)
2) Lotka-Volterra model: two-level predator/prey model, often derided or
dismissed because it's far too oversimplified. But add a "harvesting
rate" to both populations (e.g., fishing nets that capture both food fish
and predator fish, or a pesticide that harms both gypsy moths and their
wasp predators). One gets the insight known as Volterra's principle
(which usually holds up in more realistic predator/prey/ecosystem models
but is more difficult to see). No one can anticipate the result that
harvesting some fraction of both populations actually gets wilder
oscillations and, on average, a higher prey populations (more gypsy
moths!). It's even hard to explain when you see the structure and
dynamics.
3) WORLD2 (Forrester): OK, not really small, but vividly clear, about 50
equations, five stocks, showing growth paths of population, capital
investment and pollution in a finite world. Very interesting to high
school kids, maybe middle school -- what could be more important than
world dynamics? Lots of initially surprising behavior, made
understandable by the model's feedback structure. A good one to start
with is simulating Zero Population Growth starting world-wide in 1970.
Why doesn't it work?
4) Epidemic model. (There are lots of these out there,; anyone could
build his/her own.) Two stocks (Susceptibles and Infected), an infection
rate and a cure rate, and two constants suffice to see a somewhat
surprising insight known as the "Threshold Theorem" of epidemiology: For
Susceptible populations below a certain threshold (the ratio of the
constants governing the spread of infection and the rapidity of cures) no
epidemic takes place. For initial Susceptible populations slightly above
this threshold, simulations show that lots of Susceptibles never get the
disease -- the epidemic comes to and end not because it runs out of
susceptibles (as we commonly think, looking at the causal-loop diagram of
this system) but because it runs out of Infectious people.
5) Resonance. Take a simple oscillating model, like the one for a mass
on a spring. [Stocks could be position and momentum, velocity is the
rate-of-change of position and force is the rate of change of mementum.
Velocity is momentum divided by mass. Force is proportional to
displacement (equilibrium position minus position).] Start the model in
equilibrium (position = equilibrium position, momentum = 0), and drive the
model with an oscillating "jiggle," maybe in the equilibrium position, as
if somebody is holding the spring and moving her hand up and down in
smooth oscillations. What happens to the oscillating spring?
(I put a sine wave in the equilibirum position with a period I could set.
Be sure to try a period for the disturbance roughly equal to the natural
periodicity of the spring, as well as less than and greater than that
natural period.) If kids can guess these dynamics without simulation,
then we should be giving them harder stuff! Even kids who know what a
sine wave is... You can illustrate what the simulations show with a
slinky, or even by stirring your coffee (have some towels ready).
6) Worker burnout (Homer, SDR 1,1). 19 equations, 4 levels (Energy
level, hours worked per week, perceived accomplishments per week, expected
accomplishments per week). OK, again not really simple, and probably a lot
more interesting to teachers than kids, but maybe there are some
overworked kids out there (soccer, flute, dancing lessons, fife-and drum
corps, homework, ...) who would be surprised to see how this model reveals
what personal work habits conspire to create great swings in
accomplishments. Fascinating and deeply instructive to learn that we can
actually accomplish more by limiting our work hours per week.
...which suggests I should stop working on this list and do something
else.
Please add to this list. The more we have of these motivating,
"surprising" dynamic structures, the more folk (kids, adults, clients,...)
we might intrigue and educate.
...George
-----------------------------------------------------------------------
George P. Richardson G.P.Richardson@Albany.edu
Chair of Public Administration and Policy http://www.albany.edu/~gpr
Rockefeller College of Public Affairs and Policy Phone: 518-442-5258
University at Albany - SUNY, Albany, NY 12222 Fax: 518-442-5298
-----------------------------------------------------------------------
Date: Sat, 2 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: Ed Gallaher <gallaher@mail.teleport.com>
Subject: Even Simpler models with unexpected or puzzling dynamics
From George Richardson:
>1) URBAN1
>2) Lotka-Volterra model: two-level predator/prey model
>3) WORLD2 (Forrester):
>4) Epidemic model.
>5) Resonance.
>6) Worker burnout (Homer, SDR 1,1).
This is turning into a very practical and important discussion. I am
getting a number of ideas that I can begin to incorporate in my thinking
and teaching.
I like George's examples, but as he suggested, these are not -simple-
simple. They are certainly within reach, and would be enormously
informative to anyone who worked through these examples.
However, I would like to provide a short list of -really- simple examples,
which may contain only a single stock and some very limited, but still
surprising results.
1. Forest sustainability - (details provided in my recent message). A small
forest (stock) is logged at a low fixed rate. A given time is required for
tree maturation. A fixed number of trees are planted annually to replenish
the modes harvest. The system is in balance. Now, the harvest steps up to
a new, slightly higher, but still fixed rate. Planting is increased to
compensate. Intuition tells us that the forest will remain the same size as
on ongoing community resource, but this is not what happens. Why? What
would we need to do to maintain the resource indefinitely? Is this even
possible? Will there be a temporary decrease in trees, followed by a
recovery? If so, how long will the recovery take?
All accomplished with -one- stock.
2. Compound interest. (Someone suggested that this is somewhat "old hat".
True to some extent, but still a powerful lesson to newcomers! Don't I wish
I -really- understood this when I was 25?!)
(a) Starting a new job, Sally looks into the future and investes $2K into
an IRA every year for five years, beginning at age 20, and stopping at age
30 (10 total years; $20K). She never invests another dime, but allows the
fund to accumulate via compounding until she retires at 65. John starts
working at a similar job at the same age, but has expenses (car, boat,
stereo) that prevent (?) him from investing immediately. At age 30,
however, he begins to see the light, and invests $2K per year in his IRA
-every year- from age 25-60. (Can be simulated with various interest
rates.)
Guess who comes out ahead?
(b) A doting grandfather creates six $1,000 mutual fund accounts for his
new grandson Billy Bob. Each fund is expected to yeild 6-12% annually
depending on the market. Every ten years from age 10-60 Billy Bob can close
one account and use the proceeds for whatever he would like. How much will
be available to him each ten years? Does he need to work at all for the
rest of his life? What happens if the rate is 4%? 15%?
3. "Rain Barrel" (or bathtub, or drug model; in biology known as the
plateau principle). Linear or pulse input with simulataneous exponential
output.
Starting with a fixed amount, the time required to drain the stock is
(obviously) dependent upon the drain size.
With a linear input, the stock rises to a plateau. The height of the
plateau depends upon the relative input and output rates.
How long does it take to reach the plateau? Surprisingly, the _time_
required to reach the plateau depends not upon the input rate, but _only_
upon the output rate(!).
4. Population models.
With simple, unchanging birth and death rates:
If the birth rate slightly exceeds the death, the population will exhibit
compounding growth forever, although the rate will be slowed by the deaths
occurring along the way. (Looks like compound interest.)
If the death rate slightly exceeds the birth rate, the population will
decrease to zero (looks like exponential decay in #3 above), although the
decay will be slower due to the births.
ONLY if something (lack of food, crowding, disease) intervenes to (i)
decrease births, or (ii) increase deaths, will we observe S-shaped growth.
This is a classic, clear case of shifting loop dominance.
The process of building a normalized lookup table (or graphical function)
is a real eye-opener.
What are the "normal" values? To what extent does population density
influence birth and/or death rates? How does the shape of the lookup table
influence the final population (carrying capacity)? How does the shape of
the lookup table influence the time required to reach the stable
population? When does the shift in loop dominance occur?
Whew!
Like George, it's time to stop. Gotta clean out the garage so I can work on
the motorcycle.
Ed Gallaher
Date: Sat, 02 Oct 1999
From: Gerardo Valente Lozano Morales <vlozano@gauss.logicnet.com.mx>
Reply-To: vlozano@gauss.logicnet.com.mx
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: Introduction
Hello
i'm Gerardo lozano from mexico, i'm studying my master degree.
The modeling system is one off my class, and is very interesting read
your mail's
thank you.
Gerardo Lozano
Date: Fri, 1 Oct 1999
From: Mary Ellen Verona <mverona@mvhs1.mbhs.edu>
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: Re: Simple models with unexpected or puzzling dynamics
We had Simpson's paradox discussed by our Superintendent recently - you
have stratified populations of kids. Each population increases in test
scores but the total score goes down because there is an increasing
percentage of lower socioeconomic kids coming into the system. Certainly,
this can be done simpley without heavy duty dynamics, and its not really
that puzzling. Is there any point in creating a dynamic model? What kind
of feedback is possible?
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
From: KCStarguy@aol.com
Date: Sat, 2 Oct 1999
Subject: Re: ST and children's development
To: k-12sd@sysdyn.mit.edu, jldutton@iac.net
Some interesting premises.
I think there is a need to infuse and integrate the system models (stella and
other sims) into trying to related various suject, processes and focus more
on the learning side then using the technology side.
You start with subjects and then try to weave interdisciplinary thinking into
it.
I think the leverage point is to make it more visual oriented and not so much
logical- mathetmatical. System thinking can be more then learning logical
thinking. More toughts coming.
Dr. Eric Flescher (KCStarguy@aol.com)
Project S.I.M. (Simulations, Interdisciplinary internet and Metacognitive
activities
In a message dated 9/28/99 1:22:20 PM, k-12sd@sysdyn.mit.edu writes:
<<
When do you begin with each of the disciplines? How do you start, and is
there also a pattern of complexity? Are there some obvious leverage points
developmentally? >>
From: dajoy@hoy.net (Ajoy Victor)
To: k-12sd@sysdyn.mit.edu, system-dynamics@world.std.com
Date: Sat, 2 Oct 1999 23:04:48 -0500
Subject: How "too close is too dangerous"?
I was reading a book about technology and education. The
author says:
"Critics have pointed to the influence of the allegedly
mechanized thought processes of computers on how people
think.
...
The critic is afraid that children will adopt the computer
as model and eventually come to think mechanically themselves."
"MindStorms" by Seymour Papert.
But then he argues that:
"By deliberately learning to imitate mechanical thinking, the
learner becomes able to articulate what mechanical thinking is
and what it is not. The exercise can lead to greater confidence
about the ability to choose a cognitive style that suits the
problem.
...
Analysis of mechanical thinking and how it is different from
other kinds and practice with problem analysis can result in a
new degree of intellectual sophistication".
Both models seem probable and surely one or the other thing
can happen depending on certain variables (strength of feedback
loops). My question is: what are the important things to consider
in deciding to be on "the side of the critics" or on "Papert's side".
I would like to be able to apply the above to, say: In what
conditions can I "get curious" about sex and not become a pervert?;
or why can't I "get curious" about cocaine?
Daniel
From: "Jean-Louis Cordonnier" <jlcord@wanadoo.fr>
To: <k-12sd@sysdyn.mit.edu>
Subject: URBAN1 & WORLD2
Date: Sun, 3 Oct 1999
Where can I find those 2 mdls for Stella : URBAN1 & WORLD2 ?
>Subject: Simple models with unexpected or puzzling dynamics< Cordialement
> Jean-Louis ============================================Jean-Louis
>Cordonnier36, rue Lavisse66000 PERPIGNAN (FRANCE)jlcord@wanadoo.fr
Date: Mon, 4 Oct 1999
From: Mary Ellen Verona <mverona@mvhs1.mbhs.edu>
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: Re: Even Simpler models with unexpected or puzzling dynamics
On Mon, 4 Oct 1999, k-12sd wrote:
This model is an eyeopener when put into the context of teacher
hire/availability and retirement. What happens to overall quality when
unqualified teachers are hired to fill additional slots?
> 1. Forest sustainability - (details provided in my recent message). A
> small forest (stock) is logged at a low fixed rate. A given time is
> required for tree maturation. A fixed number of trees are planted
> annually to replenish the modes harvest. The system is in balance.
> Now, the harvest steps up to a new, slightly higher, but still fixed
> rate. Planting is increased to compensate. Intuition tells us that the
> forest will remain the same size as on ongoing community resource, but
> this is not what happens. Why? What would we need to do to maintain
> the resource indefinitely? Is this even possible? Will there be a
> temporary decrease in trees, followed by a recovery? If so, how long
> will the recovery take?
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: Mon, 4 Oct 1999
From: G P Richardson <gpr@csc.albany.edu>
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: Re: Simple models with unexpected or puzzling dynamics
On Mon, 4 Oct 1999, k-12sd wrote:
> We had Simpson's paradox discussed by our Superintendent recently -
> you have stratified populations of kids. Each population increases in
> test scores but the total score goes down because there is an
> increasing percentage of lower socioeconomic kids coming into the
> system. Certainly, this can be done simpley without heavy duty
> dynamics, and its not really that puzzling. Is there any point in
> creating a dynamic model? What kind of feedback is possible?
Simpson's paradox is a reversal of the direction of an association when
data from several groups are combined into a single group. In its essence
it is a static (not dynamic) idea: one looks at, say, the means on some
measure (like percent on-time arrivals of airlines) in detail and in the
aggregate (e.g., at different airports and in all the airports combined).
One can get the odd result that one airline can be better than another at
every single airport, but have a worse aggregate overall on-time
percentage.
Or in Mary Ellen's example, give two tests to three schools, but have the
populations of the schools change between the time of the first test and
the second. You can get the odd result that at each school the average
test score rises from Test 1 to 2, but the overall average declines.
But the insight is not a dynamic insight, nor a feedback insight, I
believe. (I know, you're shocked that I (especially I?) would say that,
but there are moments when I'm actually able to recognize that not all the
cool insights in the world are feedback insights!)
Whether there is a dynamic phenomenon that has Simpson's paradox at its
heart, I don't know.
...George
From: Andy Ford <forda@mail.wsu.edu>
To: "'k-12sd'" <k-12sd@sysdyn.mit.edu>
Subject: RE: Even Simpler models with unexpected or puzzling dynamics
Date: Mon, 4 Oct 1999
This is Andy Ford joining in on the interesting discussion of simple models
that deliver unexpected or surprising results. I have one observation and
one question for the group.
Observation:
The many model suggestions and the enthusiasm of the discussion is
encouraging. I have used some of the models suggested by George Richardson
and Ed Gallaher and found them to be rewarding in one classroom setting and
disappointing in a slightly different setting. My observation (taken from
College SD classes at the sophomore level) is that he key to the success
of the learning exercise seems to be allowing time for the students to
describe their expectations for the dynamic behavior IN ADVANCE of building
and running the model. When students invest an important fraction of the
class time to provide a pencil sketch of the behavior of the key system
variable over time, some of them discover that they don't really have a
working impression of the dynamic behavior. (For this group, almost any
projection from the SD model is "unexpected." Most of the other students
will be able to prepare their own "pencil forecast" of the dynamic
behavior, but will be surprised to learn that their classmates prepare a
qualitatively different forecast. If there is enough class time to discuss
the differences in their forecasts (and their underlying models), their
receptiveness to learn from a formal, system dynamics model seems to
improve.
Question:
I find that many students are reluctant to prepare a "pencil forecast" of
the key variable in advance of a modeling exercise. Some react like the
exercise is a silly distraction prior to getting on with the "real work."
I would love to hear what others have done to inspire students to put more
effort into this opening step of a classroom modeling exercise.
Andy Ford
Program in Environmental Science and Regional Planning
Washington State University
Pullman, WA 99164-4430
USA
Phone: 509 335 7846
Email: FordA@mail.wsu.edu
Website: http://www.wsu.edu/~forda
From: LORRRAND@aol.com
Date: Mon, 4 Oct 1999
Subject: Re: Simple models with unexpected or puzzling dynamics
To: k-12sd@sysdyn.mit.edu
As an elementary teacher, I would think teacher experience with special
populations, knowledge about the needs of the lower SEO students, (such as
meaningful integrated curriculum, and real life experiences) would all
figure in somehow.
Lorraine Randazzo
Date: Tue, 5 Oct 1999
From: Mary Ellen Verona <mverona@mvhs1.mbhs.edu>
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: moving from money to economy
I wonder if this model would allow for a transition to economic ideas. I
don't think that it fits into the "simple, counterintuitive models" But
its a nice concise descripiton that many high school level kids might be
able to attemp to model.
>From the Washington Post, Thursday, September 23, 1999 (the Post has done
a series of articles on each decade of the past centure during the last
couple of weeks)
Herbert Hoover himself believes that "many persons left their jobs for the
more profitable one of selling apples."
The apple story is enough to make you think the Reds are right.
In 1930, right after the Crash, Washington State has a bumper crop of
apples. Too many to sell. So instead of dumping them, they give them to
vendors on credit.
Next thing, men are lined up in Wall Street, wearing homburgs and selling
apples, 5 cents apiece. There are so many of them they
start cutting prices on each other. At the same time, the growers get
greedy--raise the prices and don't cull the rotten ones. Pretty soon, you
can't make any money in the apple business, and it's all over.
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, 5 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: Ed Gallaher <gallaher@mail.teleport.com>
Subject: RE: Even Simpler models with unexpected or puzzling dynamics
>From: Andy Ford <forda@mail.wsu.edu>
>I find that many students are reluctant to prepare a "pencil forecast" of
>the key variable in advance of a modeling exercise. Some react like the
>exercise is a silly distraction prior to getting on with the "real work."
>I would love to hear what others have done to inspire students to put more
>effort into this opening step of a classroom modeling exercise.
EVERYONE is reluctant to put pencil to paper, from students to Ph.D.s!
We are all so conditioned against sticking our necks out and giving the
"wrong" answer. But boy, a huge amount of learning takes places between
paper/pencil and then pushing the "Run" button!
The inevitable consequence of avoiding the paper/pencil step and skipping
ahead to the Run is to look at the output and, more often than not, say
"Yeah, that makes sense," when in retrospect there is perhaps a 25% chance
that the paper/pencil would have agreed with the computer!
Ed Gallaher
Date: Tue, 05 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: teresa@northwest.com
Subject: RE: Even Simpler models with unexpected or puzzling dynamics
I couldn't agree more that asking students to "take a risk", "make a
commitment on paper about their thinking is how you first can begin to
engage them in a truly active (mentally) learning process. Building the
support within the classroom that it is okay to have a different answer
when you are just beginning to explore something is a key factor. Students
still want to have the "one right answer" the first time. We learn more by
seeing how each other approaches a problem. Exposing all of the gaps in
understanding collaboratively helps us to fill in those gaps constructively.
I also agree that these exercises play out differently with different
groups of students. Some are more receptive to and patient with the mental
challenges than others. That is pretty much what I was eluding to with my
earlier comments about students' reactions to probing their thinking with
simple, but robust models (like the bathtub). It takes time AND patience
for both the teacher AND the students.
I have begun to use "writing journals" in my classes for these activities.
Whenever possible I like to use a physical demo first that relates in some
way to the abstract model that we will study.
I ask them to write their observations about what they see, then write a
short explanation. Then I ask them to sketch a BOTG of some phenomena then
move on to building the model and again ask for another sketch before I run
the model.
This is running on a bit long but to share a very recent example...I
punched four holes vertically up the sides of a 2 Liter soda pop bottle,
asked a student to plug the holes with his/her fingers, filled up the
bottle with water and let it go. The students were to make observations in
their journals. I then asked them to think about how this was a simple
system and to explain what they thought produced their observations. I
then proceeded to draw the "bathtub" on the chalkboard and gave them some
parameters for that and asked them to sketch a BOTG of the water level over
time. I asked them to think of any properties in common between these two
systems. I have made no judgements about their responses NOR given them
any "right answers" thus far. The only model that they had worked with up
to this time was a basic population model.
I ask them to think back about the structure of that model and could we
begin to build a model of the bathtub with those same structures (Stock,
Flow, Infow, Outflow, etc.) We proceed to do this using the
computer/projector system. I only build into the model what they think is
important and as I predicted they would just consider what I gave them
literally on the chalkboard
(a bathtub resevoir, an inflow and outflow with defined rates).
I then asked them to think about any hidden aspects of the model and some
classes brought up their thoughts about water pressure and other classes
did not. They still did not make any suggestions about how it might fit
into the model or even if it should, mostly because this is still a very
new experience for them. When I run the model without a feedback loop and
they see the graph, at first they are happy to see that in many cases the
linear graph matches their own. I then proceed to give them some
information that this graph is accurate for this model but I ask them if
this model is accurate for the situation? We go on to decide together that
something about water pressure is changing the outflow over time and that
this would have an impact on their predicted graphs.
In attempting to get them to tell me where to add the loop I introduce the
idea of looking at the "physics of the model". They have previously
discussed the concept of feedback so I ask them to tell me where the
feedback comes from and what is it influencing?
Was it Barry Richmond who coined this phrase "physics of the model"? We
figure all of that out together in discussion and then I ask them to sketch
yet another BOTG graph in their journals BEFORE I rerun the model. A way to
REALLY get them to commit their thinking about the BOTGs is to have them
sketch these on small whiteboards (that I have in my classroom) and hold
them up. I can quickly assess the thinking of the entire class when they
use these. I ask them to NOT look at anyone elses board until I say so.
It is a good experience for them to see the variety of graphs.
This does take a bit of patience with high school students who are used to
being told the "one right answer" often with little wait time once a
question is proposed. This whole cycle of events, using the physical demo
first parallels the approach taken by physics instructors, David Sokoloff
of the U of O and Jim Minstrel of Gig Harbor, WA if anyone reading this is
familiar with their teaching approaches. Wow that was way too....long.
Teresa
At 10:59 AM 10/5/99 -0400, you wrote:
>From: Andy Ford <forda@mail.wsu.edu>
>To: "'k-12sd'" <k-12sd@sysdyn.mit.edu>
>Subject: RE: Even Simpler models with unexpected or puzzling dynamics
>Date: Mon, 4 Oct 1999 12:00:50 -0700
>
>This is Andy Ford joining in on the interesting discussion of simple models
>that deliver unexpected or surprising results. I have one observation and
>one question for the group.
>
>Observation:
out differently with differently than with others. groups of students>
>The many model suggestions and the enthusiasm of the discussion is
>encouraging. I have used some of the models suggested by George Richardson
>and Ed Gallaher and found them to be rewarding in one classroom setting and
>disappointing in a slightly different setting. My observation (taken from
>College SD classes at the sophomore level) is that he key to the success
>of the learning exercise seems to be allowing time for the students to
>describe their expectations for the dynamic behavior IN ADVANCE of building
>and running the model. When students invest an important fraction of the
>class time to provide a pencil sketch of the behavior of the key system
>variable over time, some of them discover that they don't really have a
>working impression of the dynamic behavior. (For this group, almost any
>projection from the SD model is "unexpected." Most of the other students
>will be able to prepare their own "pencil forecast" of the dynamic
>behavior, but will be surprised to learn that their classmates prepare a
>qualitatively different forecast. If there is enough class time to discuss
>the differences in their forecasts (and their underlying models), their
>receptiveness to learn from a formal, system dynamics model seems to
>improve.
>
>Question:
>
>I find that many students are reluctant to prepare a "pencil forecast" of
>the key variable in advance of a modeling exercise. Some react like the
>exercise is a silly distraction prior to getting on with the "real work."
>I would love to hear what others have done to inspire students to put more
>effort into this opening step of a classroom modeling exercise.
>
>Andy Ford
>Program in Environmental Science and Regional Planning
>Washington State University
>Pullman, WA 99164-4430
>USA
>
>Phone: 509 335 7846
>Email: FordA@mail.wsu.edu
>Website: http://www.wsu.edu/~forda
>
Date: Mon, 4 Oct 1999
From: G P Richardson <gpr@csc.albany.edu>
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: Re: Even Simpler models with unexpected or puzzling dynamics
On Mon, 4 Oct 1999, k-12sd wrote:
Ed's list is moving in the right direction -- really simple models.
The first three
> 1. Forest sustainability
> 2. Compound interest, investing at different starting times.
> 3. "Rain Barrel"
have the required property that one probably can't see the results without
simulation.
But the fourth, does not (I believe) have this property. Ed says:
> 4. Population models.
> If the birth rate slightly exceeds the death, the population will exhibit
> compounding growth forever, although the rate will be slowed by the deaths
> occurring along the way. (Looks like compound interest.)
>
> If the death rate slightly exceeds the birth rate, the population will
> decrease to zero (looks like exponential decay in #3 above), although the
> decay will be slower due to the births.
>
> ONLY if something (lack of food, crowding, disease) intervenes to (i)
> decrease births, or (ii) increase deaths, will we observe S-shaped growth.
> This is a classic, clear case of shifting loop dominance.
But the "curving upward" nature of the Population/Births loop can be
argued straight from the causal loop diagram -- as population increases,
the births per year increases, so what is added to population each year
increases, so the population graph must be curving upward. (We trace that
argument a lot, right?)
And the S-shaped pattern from a constraint is often argued right from the
causal-loop diagram (e.g., the Fifth Disclpline and lots of other places
with the "limits to growth" or "limits to success" archetype).
So I think a skeptical kid could be unimpressed with this population
simulation, thinking he could see the dynamics without the simulation. So
we'd just be "illustrating" with simulation, not motivating the need for
it as a crucial cognitive aid.
...George
From: "John Gunkler" <jgunkler@sprintmail.com>
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
Subject: RE: Even Simpler models with unexpected or puzzling dynamics
Date: Wed, 6 Oct 1999
Andy,
When I have studied SD models (on my own) I have several times found myself
feeling like your students, whom you describe this way:
>>I find that many students are reluctant to prepare a "pencil forecast" of
the key variable in advance of a modeling exercise. Some react like the
exercise is a silly distraction prior to getting on with the "real work."<<
(Over)generalizing from my personal reactions, I would suggest that the
"silly distraction" excuse may only be a socially correct response masking
the real reasons for the reluctance. My own real reasons, as I've been able
to analyze them in these situations, are actually something else. In spite
of the fact that I'm only studying for my own benefit (and I have specific
learning objectives in mind: I really want to learn about the model or about
techniques used in it or about how to understand SD models, etc.), I feel
embarrassed when I cannot confidently forecast model behavior -- especially
when I thought I "understood" how the model was put together. This
embarrassment turns into frustration when I subsequently realize that I
don't even know a good way to attack the problem, so I could feel more
confident in my understanding (without running the simulation.) It becomes
an approach/avoidance conflict -- I want to test my understanding by drawing
a pencil forecast before running the simulation, but I'm not at all
confident in my understanding and don't want to make a fool of myself (even
to myself! kind of "sick", huh?), and the only way I can figure out how to
improve my understanding is to run the simulation. So I'm stuck with
conflicting needs to not run and to run the simulation -- and it drives me
crazy.
Where does this leave you, as mentor? I think what I'd like from you (if I
were in your classroom) is some help with techniques for understanding model
behavior without running the simulation. That is, help me walk through the
model, identifying the causal loops, classifying them as positive or
negative, looking at the non-linearities (table/graph functions), thinking
through what behavior the simple loops would therefore generate (if they
were the only thing in the model), then [the really tough part!] thinking
about how the loops interact to generate the behavior of the key variable
you're tracking. I also might need some help from you to get "out of the
trees and see the forest" -- that is, to forget about loops and table
functions and think about the "real world" being modeled, think about my
mental model of that real world, and think about how the key variable
behaves in my mental model.
Date: Wed, 6 Oct 1999
From: G P Richardson <gpr@csc.albany.edu>
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: RE: Even Simpler models with unexpected or puzzling dynamics
Concerning enticing students to hazard guesses...
I've had some success in getting people to draw their guesses about system
behavior before a run (at least in class) by being really up front about
why we should do it. Students respond pretty well to the observation that
without committing ourselves and our expectations we rob ourselves of
seeing a "discrepant event" -- a result that conflicts with our
expectations -- and thus rob ourselves of a learning opportunity.
What my students do outside of class is anybody's guess...
...George
-----------------------------------------------------------------------
George P. Richardson G.P.Richardson@Albany.edu
Chair of Public Administration and Policy http://www.albany.edu/~gpr
Rockefeller College of Public Affairs and Policy Phone: 518-442-5258
University at Albany - SUNY, Albany, NY 12222 Fax: 518-442-5298
-----------------------------------------------------------------------
Date: Wed, 06 Oct 1999
Subject: Re(2): Even Simpler models with unexpected or puzzling dynamics
To: k-12sd@sysdyn.mit.edu
From: mikes@fc.cfsd.k12.az.us (FH Michael Slootmaker)
(In reply to George Richardson and Ed Gallaher)
I think younger kids do benefit from the population simulation. I've seen
3rd, 4th, and 5th graders come up with some major "ah-has" based on the
population simulation.
Mike Slootmaker
Systems Dynamics/System Thinking mentor
Catalina Foothills School District
Date: Wed, 6 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: Ed Gallaher <gallaher@mail.teleport.com>
Subject: Re: Even Simpler models with unexpected or puzzling dynamics
I had this feeling when I wrote the earlier message, and fully agree with
George. A model with simple (unchanging) birth and death rates (natality
and mortality, respectively) is not all that difficult or exciting. They
do lay the foundation for adding area and population density, and then a
table function which provides the loop that ultimately produces S-shaped
growth.
NOW, we are at the point where we can create three structures like this,
one for commercial buildings, one for residential housing, and one for
population, and link them together. These now give us the structure
required to examine urban growth and decay as George suggested in his
earlier message.
Wow, this has been a great discussion. I think that collectively we are
beginning to assemble a set of basic concepts and the models which
illustrate them.
Of course this all has to be put into practice in the 7-12 classroom. I
feel great about pontificating from my ivory tower medical school office.
;-) Making this work in the trenches is where the rest of you earn your
keep!!
Ed
Date: Wed, 06 Oct 1999
Subject: Two more simple models
From: "Timothy Joy" <tjoy@pps.k12.or.us>
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
These last two weeks have found this English teacher in strange lands:
Biology-- instructing sophomores on the vagaries of population, and
Physics-- playing ("instructing" is simply too generous a term in this
context) with velocity and distance. Both cases used simple models, but
brought a few keen observations that likely would not have occurred without
the simulations.
In biology, students first built population models while they also grew
actual yeast in flasks, hoping eventually to check the real against the
model. As in many experiments, sloppy methods meant many experiments were
very difficult to decipher, but the modeling seemed to help. Once we talked
about factors that increased yeast death and included a food supply,
interesting things followed. Many predicted, understandably, that once
population grew and, thereby, food declined, the death rate would increase.
BUT, none figured the unusual spikes and valleys that occur depending on the
AMOUNT a yeast would consume. In fact, they found that quickest way to kill
them off was to feed them a lot: the population grew so fast, the
inevitable crash came more quickly. What actual yeasts do is part of the
discussion that followed: what were some of the limitations of that model.
In physics, we "raced" a moving truck against a car accelerating from 0.
Once students ran a single base line test( Truck moves at 30; car
accelerates at 5; both with 0 as initial distance), they had a reference.
And they thought they knew things. Asking them to predict the outcomes of a
series of tests met with wild variations: a) truck at 30, car accelerates
at 10; b) truck at 45; car accel at 7.5; c) truck 15; car accel at 5 and
others. About one half got "a" but only one-quarter or so were right about
"b". These are VERY simple models--stocks and flows only, with a single
connector. Follow up discussions, with an actual physics in the room, were
enlightening: stocks, flows, test results, and Newton's equations all
worked through side by side.
Jay Forrester said once that a sound instructional model need not be more
than ten equations deep. That' not very big. In a whole class situation,
there is enough going on in such a model that discussions and surprises are
fruitful. Students wishing to explore more (why, for instance, a 1999 Ford
F-10 might accelerate more rapidly than a 1980 Toyota Corolla) would at
least have velocity-distance kernel in their spinning heads.
This is a GREAT discussion.
---------------------
Date: Fri, 08 Oct 1999
From: Niall Palfreyman <Niall.Palfreyman@assyst-intl.com>
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: Simpler models and pencil forecasts
John Gunkler wrote:
> don't want to make a fool of myself ...
> Where does this leave you, as mentor? ...
> That is, help me walk through the model ...
> thinking about how the loops interact ...
> to get "out of the trees and see the forest"
You know, this discussion reminds me a little of when I used to teach
object-oriented programming. The feelings of embarrassment you talk
about here, John, are very much the kind of feelings which often seemed
to get in the way there. In particular, students were reluctant to use
the compiler as a playground in which they could experiment with ways of
using the language they were studying. They were embarrassed to make
"mistakes" in the language, forcing themselves to get everything exactly
right beforehand, and then chastising themselves for getting it "wrong"
when the compiler raised a syntax "error".
Since I feel similar myself, I could well relate to their fears, and I
tried to allay them by reframing the syntax messages as fun and
interesting learning feedback. As I recall, I did three different
things:
i. I told a humorous story about my son learning to walk, falling over
and laughing with the enjoyment of the experience.
ii. I got them all to write in big, colourful letters in their notes:
"Syntax messages are the most wonderful thing which can happen to me".
iii. I played a quiz game with them, in which I gave them syntactically
incorrect statements and then got them to invent their own syntax
messages.
Could this be applied to the "pencil forecast" issue? One thing which
immediately strikes me is that it would be well to avoid the idea of the
pencil forecast as "getting it wrong", emphasising rather the fun of the
learning involved in moving from the forecast to the simulation. Maybe
one could start with the words: "You know, I was thinking last night
about this situation, and wondered how it would work in real life. Let's
see if we can make some guesses together about how it might behave."
Here the forecast becomes a focus for curiosity (always a great learning
motivator!) in which the teacher also takes part.
Best wishes,
Niall.
PS: I agree with Tim Joy - this _is_ a great discussion!
--
We have only the world that we can bring forth
with others, and only love helps us bring it forth.
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, 08 Oct 1999
From: "RICHARD TURNOCK" <Richard_Turnock@pgn.com>
To: k-12sd@sysdyn.mit.edu
Subject: Simple models become complex
>>>>>>>>>.
Pressure causes outflow rate to increase, so they are faced with a
discrepent event. Can anyone offer something analogous in a separat
context?
Mary Ellen Verona
>>>>>>>>>>>>>>
Predict this graph ahead of time.
Global warming model. Start with linear inflow and compound decay.
Sun is linear inflow to
Stock #1 is Earth Energy
Converter #1 is temperature of earth to the fourth power (link stock #1 to
Converter #1)
Outflow # 1 is Infrared Radiation to atmosphere (link Converter #1 to
Outflow #1)
Outflow # 2 is infrared to space (link converter #1 to outflow #2)
Connect outflow #1 to Stock #2 = Atmosphere
Add a converter for temperature of atmosphere and two outflows from stock
#2 connected in same pattern as stock #1. Except one outflow from
atmosphere is inflow to stock #1 (Feedback loop!)
Initial values: Sun is 1, stocks zero. (don't try other values yet!)
Predict graph of stocks.
(maybe you don't want to have to explain temperature to the fourth power
>from the stefan boltzman equation - have prebuilt model for the students
to exercise or the teacher uses projector to lead class discussion.)
Predict graphs ahead of time:
Have students build the model without using any multiplication of the
temperature. Run the model. Then try it with temperature squared, then
raise it to the third power and then the fourth power. You could have them
build four copies of the basic model, each with one of these equations and
run two comparative graphs one with all the earth energy stocks and one
with all the atmosphere stocks.
If the earth energy system feedback loop is based on temperature to the
fourth power what would you predict the graph would look like if CO2 is
added to the atmosphere and increased the ability of the atmosphere to hold
more energy?
Now predict what will happen if you use a value greater than one for the
value of the sun inflow. Try values from two through 10.
Richard
Date: Fri, 08 Oct 1999
From: jan mons <jmons@glynn.k12.ga.us>
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: Re: Even Simpler models with unexpected or puzzling dynamics
Since it has been awhile since I have written anything, I will
reintroduce myself as a mentor for a Waters Grant Program in Brunswick,
Georgia. Like many of you I am enjoying this discussion. Several points
that have been brought out have hit home with me as I have recently
started working with elementary students.
As a former eighth grade math teacher I constantly struggled with
trying to get my students to solve application problems (I refuse to use
the phrase „word problems‰). The difficulty was that their understanding
of the basic skills in math was so weak that I spent more time on those
than on the great applications problems.
The discussion about whether the population model is a good Œsimple‚
model strikes me as the same thing. An application problem that is
simple for a calculus student would be deadly for a fifth grader. Unless
of course the fifth grader had the same skills and understanding of the
concepts of mathematics as a calculus student. Ed‚s reference to
speaking differently to someone with SD understanding that to someone
who doesn‚t is the same reality that we as mentors face. I am actually
doing some of the same SD activities with second graders that I am also
doing with sixth graders. Both are new to the skills and concepts.
No matter what „simple‰ model I use, my question is what skill,
concept, and basic understanding of System Dynamics will the
model/activity build deeper understanding in for myself and my students.
In addition, of course, to an increased understanding of the particular
system we are discussing.
Last fall Tim Joy facilitated a great discussion on this list serve
about where to start and five basic concepts. In Durham at the CLE
conference a group of us met to try to design a (I hate to say it)
„scope and sequence chart‰. I am going to have to go through my notes
to find that information, but this current discussion certainly hits
home and will help put all those pieces together for me.
Jan Mons
GIST Project Mentor
A Waters Grant Project
Glynn County Georgia
------------------------------
Date: Thu, 7 Oct 1999
From: Mary Ellen Verona <mverona@mvhs1.mbhs.edu>
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: Re: Even Simpler models with unexpected or puzzling dynamics
Don't feel bad about pontificating, Ed. I'm much closer to the action
most of the time - but not close enought to refrain from pontificating. So
after being admonished by colleagues, I am asking for something very, very
simple:
A model analogous to population (which teachers have seen over and over).
But one in which at first glance, a constant inflow/outflow seems to work.
On closer thought, though, it is found that over the long haul, the flow
depends on the stock. An example that we do is a burette with slower
outflow than inflow. Students are asked to predict when it will reach a
particular volume. Pressure causes outflow rate to increase, so they are
faced with a discrepent event.
Can anyone offer something analogous in a separat context?
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
From: "Parminder S Raparia" <praparia@mahindrabt.com>
To: <k-12sd@sysdyn.mit.edu>
Subject: Introduction
Date: Thu, 7 Oct 1999
Greetings to All,
I am Parminder Singh Raparia from India. I working with Mahindra-British Telecom Ltd, Pune (near Mumbai). I am working on the project "To build MBT as a Learning Community" based on the principles given by Jay Forrester, Peter Senge, M Scott Peck, Fritjof Capra and Ancient Indian wisdom. We have been working on this project for almost two years now. We are the team of five people. To know more about our work I am attaching the CEL map*----approach paper of our center. I will be starting my own school based on the principles and philosophy of Systems thinking and System Dynamics, Community living, Nature, Ancient Indian wisdom, etc. I want to have a whole structure of education and learning based on the framework of system thinking and system dynamics. Looking forward for exploring and sharing ideas together......
Regards, Parminder Singh Raparia Center of Excellence for Learning MBT, Pune Ph: 91-20-774740/60 Ext-3015 Email: praparia@mahindrabt.com
*Please e-mail Mr. Raparia directly if you wish to receive this attachment. The "k-12sd" list moderator
Whatever you can do or dream you can do-----Begin it. face="Comic Sans MS">Boldness has genius, power and magic in it----Begin it now. --Goethe
From: "Steve Seaford" <sseaford@spusd.k12.ca.us>
To: <k-12sd@sysdyn.mit.edu>
Subject: "Joining"
Date: Thu, 7 Oct 1999
I'm new to the discussion group. My name is Steve Seaford. I am
the principal at South Pasadena Middle School in South Pasadena
California. I continue to teach, mostly what is known as a Socratic
Seminar class. As an administrator, my interest is to facilitate the
development of a "learning community" here at SPMS. My initiation in the
organizationl development arena began with my work for an organization
called the Coalition of Essential Schools.
Since then, I've attempted to integrate the "5 disciplines" into my work
with reform efforts in Los Angeles and here in South Pasadena. I
welcome all insights, ideas, suggestions, and/or questions from others that
might lead me to a deeper understanding of systems dynamics and learning.
Looking forward to the conversation,
Steve Seaford
South Pasadena Middle School
South Pasadena, CA 91030 626-441-5761
sseaford@spusd.k12.ca.us
Date: Thu, 7 Oct 1999
From: G P Richardson <gpr@csc.albany.edu>
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: Re: Two more simple models (and another)
On Thu, 7 Oct 1999, k-12sd wrote:
> In biology, students first built population models while they also grew
> actual yeast in flasks, hoping eventually to check the real against the
> model.
[...]
> BUT, none figured the unusual spikes and valleys that occur depending on the
> AMOUNT a yeast would consume. In fact, they found that quickest way to kill
> them off was to feed them a lot: the population grew so fast, the
> inevitable crash came more quickly.
Nice stories. Sounds like great teaching.
The yeast example reminds me of another small model (20 equations) a young
college student of mine did in the 1970s. He was able to replicate the
remarkable dynamics of an odd but famous experiment by A.J. Nicholson in
the 1950s. Nicholson observed populations of blow flies living on
"bullock's brain" (this may not be pretty, but it's science). He found
dramatic oscillations in the population of blowflies when he limited the
food (brain food?) to the larvae. When he limited the food to the adults,
the population ROSE to a high sustained level without the dramatic
oscillations.
That's counterintuitive enough to make a little model of this phenomenon
worthwhile. My student built a model with a simple aging chain (eggs,
larvae, pupae, immature adults, adults) and a few equations representing
the food available to larvae and adults and its effects on their death
rates and on fecundity. Voila -- he replicated exactly Nicholson's
observations.
It's fun to note that in those days, my students and I worked with a
DYNAMO-emulator that I wrote in BASIC and simulated our models on a DEC
PDP-8e, a table-top-sized computer with 8K of core memory (that's 8K, not
800K or 8 megs). The input/output device was a teletype, working at 10
characters a second. Each run was a treasure...
[I have the model in Vensim now and could put it on a web site if
people want it. But it might be better to try to build the aain chain
model yourself after consulting the reports of Nicholson's work contained
in E. G. Kormondy's Readings in Ecology (Englewood Cliffs, NJ:
Prentice-Hall, 1965: 109-112). Maybe there's a modern source somewhere,
as this set of experiments were important in the theory building about
"context competition" and "scramble competition" in ecology.]
...GPR
-----------------------------------------------------------------------
George P. Richardson G.P.Richardson@Albany.edu
Chair of Public Administration and Policy http://www.albany.edu/~gpr
Rockefeller College of Public Affairs and Policy Phone: 518-442-5258
University at Albany - SUNY, Albany, NY 12222 Fax: 518-442-5298
-----------------------------------------------------------------------
Date: Fri, 08 Oct 1999
Subject: Re: Even Simpler models with unexpected or puzzling dynamics
From: "Lawrence Weathers" <weathers@massed.net>
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
Mary Ellen
How about using the transferability of that same simple model you just
mentioned and change the names of the flow and stock to apply to some social
or economic system that saturates, reaches a limit, like total computer
sales in the US or some other situation?
Larry
----------
From: Mary Ellen Verona <mverona@mvhs1.mbhs.edu>
(snip)
A model analogous to population (which teachers have seen over and over).
But one in which at first glance, a constant inflow/outflow seems to work.
On closer thought, though, it is found that over the long haul, the flow
depends on the stock. An example that we do is a burette with slower
outflow than inflow. Students are asked to predict when it will reach a
particular volume. Pressure causes outflow rate to increase, so they are
faced with a discrepent event.
(snip)
Date: Fri, 08 Oct 1999
Subject: Simple Models
From: "Timothy Joy" <tjoy@pps.k12.or.us>
To: K12 <k-12sd@sysdyn.mit.edu>
While considering all these simple models, I recalled some reading from long ago, en route to the Concord conference in 1994, reading "Complexity" by Waltham or Walthrop or something like that. It was a survey of the state of thinking in complex theory. OK, it's an odd book for an English teacher to read, but ultimately of the same order as, say, Bleak House. Here's the quotation: "The most surprising lesson we have learned from simulating complex physical systems on computers is that complex behavior need not have complex roots(italics his). Indeed, tremendously interesting and beguiling complex behavior can emerge from collections of extremely simple components." The speaker is Christopher Langton, whose work in artificial intelligence has gained some notoriety, so I'm told. Probably a few of you know of him. All I can say is, ain't it so? Tim Joy
From: Quaden@aol.com
Date: Fri, 8 Oct 1999
Subject: Re: RE: Even Simpler models with unexpected or puzzling dynamics
To: k-12sd@sysdyn.mit.edu
Teresa:
Your story is NOT too long at all. It is a clear explanation: I can't wait to
try it with my middle school students.
Rob Quaden
Systems Mentor
Carlisle Public Schools
Date: Fri, 8 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: Ed Gallaher <gallaher@mail.teleport.com>
Subject: Re: Even Simpler models with unexpected or puzzling dynamics
In reply to Mary Ellen Verona:
I can create a simple scenario in which a very similar model is used for
intravenous drug infusion. Although the model is similar, the context would
be entirely different. It could be discussed as part of health, biology, or
modeling classes.
Let me know if this fits the bill.
Ed
From: clee@imp.spusd.net
Date: Sat, 9 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
Subject: Introduction
Greetings,
My name is Curtis Lee and I'm currently the Director of Instructional
Technology for South Pasadena USD in South Pasadena, California. My
interest in System Dynamics comes out of current work and previous
experience. I spent six years in Atlanta GA teaching at a
'Summerhill"-type alternative high school for adolescents who had been
traumatized by life, school, drugs, etc. The emphasis was on creating a
humane environment that was sufficiently senstive to the individual. My
interest in technology came about originally as a alternative strategy for
dyslexics/ADD students who were difficult to reach with traditional
approaches. This quickly evolved into the realization that information
technologies can play all sorts of powerful roles (good & bad) in learning
organizations.
I'm currentlly interested in the strategic introduction of
information/communication technologies into learning organizations using
system dynamics as the lens. I think there are real opportunities to use
inclusive technologies to personalize the student experience in these
complex systems and make inroads against the pervasive sense of anonymity
that often occurs in large public schools.
Looking for articles/books looking a the systems role of email/web on
learning organizations.
Thanks,
Curtis Lee
clee@spusd.k12.ca.us
626.665.9870
To: k-12sd <k-12sd@sysdyn.mit.edu>
From: Linda Booth Sweeney <Linda_Booth_Sweeney@harvard.edu>
Subject: Re: Even Simpler models with unexpected or puzzling dynamics
Date: Mon, 11 Oct 1999
Thanks for your contribution here Jan. I think your point about
understanding of some SD concepts not being directly tied to age is right
on. If you happen to find your SD "scope and sequence chart" notes, I know
I for one, would be eager to hear what you came up with.
Linda Booth Sweeney
>Date: Fri, 08 Oct 1999 08:44:49 -0500
>From: jan mons <jmons@glynn.k12.ga.us>
>To: k-12sd <k-12sd@sysdyn.mit.edu>
>Subject: Re: Even Simpler models with unexpected or puzzling dynamics
>
>Since it has been awhile since I have written anything, I will
>reintroduce myself as a mentor for a Waters Grant Program in Brunswick,
>Georgia. Like many of you I am enjoying this discussion. Several points
>that have been brought out have hit home with me as I have recently
>started working with elementary students.
> As a former eighth grade math teacher I constantly struggled with
>trying to get my students to solve application problems (I refuse to use
>the phrase „word problems‰). The difficulty was that their understanding
>of the basic skills in math was so weak that I spent more time on those
>than on the great applications problems.
> The discussion about whether the population model is a good Œsimple‚
>model strikes me as the same thing. An application problem that is
>simple for a calculus student would be deadly for a fifth grader. Unless
>of course the fifth grader had the same skills and understanding of the
>concepts of mathematics as a calculus student. Ed‚s reference to
>speaking differently to someone with SD understanding that to someone
>who doesn‚t is the same reality that we as mentors face. I am actually
>doing some of the same SD activities with second graders that I am also
>doing with sixth graders. Both are new to the skills and concepts.
> No matter what „simple‰ model I use, my question is what skill,
>concept, and basic understanding of System Dynamics will the
>model/activity build deeper understanding in for myself and my students.
>In addition, of course, to an increased understanding of the particular
>system we are discussing.
> Last fall Tim Joy facilitated a great discussion on this list serve
>about where to start and five basic concepts. In Durham at the CLE
>conference a group of us met to try to design a (I hate to say it)
>„scope and sequence chart‰. I am going to have to go through my notes
>to find that information, but this current discussion certainly hits
>home and will help put all those pieces together for me.
>
>Jan Mons
>GIST Project Mentor
>A Waters Grant Project
>Glynn County Georgia
Linda Booth Sweeney
e-mail: Linda_Booth_Sweeney@harvard.edu
Date: Mon, 11 Oct 1999
From: William Costello <WILL@cvumail.cvu.cssd.k12.vt.us>
To: k-12sd@sysdyn.mit.edu
Subject: Re: Two more simple models (and another) -Reply
The study by Nicholson is indeed a fascinating one...and Nicholson
himself was an interesting individual. You can see examples of his
work and thinking (especially the influence of feedback upon natural
selection), if you go to
<http://www.asap.unimelb.edu.au/bsparcs/aasmemoirs/nicholso.htm>. I
have seen some of his works archived on the internet but have not yet
encountered the sheep blowfly study.
-----------------------------
Date: Wed, 13 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: teresa@northwest.com
Subject: Re: Even Simpler models with unexpected or puzzling dynamics
Jan
I'm not quite sure what you are asking here about the population model, but
if I am interpreting your question correctly, I think it applies to some of
our Oregon State Standards and Benchmarks. These are stated as follows:
Content Standard: Use concepts and processes of change, constancy, and
measurement.
Benchmark for this standard for Grade 10: Describe the relationship
between constancy and change within systems.
Eligible Content (what the state could ask kids in their assessment test
for this standard and benchmark): Recognize that large scale constancy
sometimes is explained by opposing small scale changes. For example, a
population may remain constant even though individuals are added and removed.
This is exactly the wording for one of our Science standards. I am hearing
"balancing feedback loops" when I read this. I also ask myself, how can
students really understand the dynamics of this situation if they don't
have an opportunity to use something like a STELLA model. Could they even
begin to really understand this by just looking at a textbook graph and
tables (which is all that we've had to use until now). STELLA can make
this an interactive, living experience. You can then look at the numbers
along with the structure of the population model and explain to kids how
the math works if you want, depending upon the level of the kids.
The list of standards, benchmarks, eligible content, goes on and on
implying ST/SD as the optimum learning approach.
An interesting anecdote....One of my students questioned whether I was
aware of the curriculum for our Global Science course, the expectations for
the state assessment she'll take this spring AND
she was concerned about "knowing" enough information to be able to pass the
test. I proceeded to explain my purpose and the benefits she'll gain from
experience with ST/SD and she proceeded to drop my class. Fortunately she
did go into Biology, however, without the benefit of anymore ST/SD unless
that teacher accepts my invitation to work with his kids. Wow, will she
ever be suprised if I show up there sometime soon packing my ST/SD along
with me!!
Maybe I'm opening up another question for reflection here, but,
are there other anecdotes out there similar to this and how have you
responded? TERESA
At 12:07 PM 10/8/99 -0400, you wrote:
>Date: Fri, 08 Oct 1999 08:44:49 -0500
>From: jan mons <jmons@glynn.k12.ga.us>
>To: k-12sd <k-12sd@sysdyn.mit.edu>
>Subject: Re: Even Simpler models with unexpected or puzzling dynamics
>
>Since it has been awhile since I have written anything, I will
>reintroduce myself as a mentor for a Waters Grant Program in Brunswick,
>Georgia. Like many of you I am enjoying this discussion. Several points
>that have been brought out have hit home with me as I have recently
>started working with elementary students.
> As a former eighth grade math teacher I constantly struggled with
>trying to get my students to solve application problems (I refuse to use
>the phrase ìword problemsî). The difficulty was that their understanding
>of the basic skills in math was so weak that I spent more time on those
>than on the great applications problems.
> The discussion about whether the population model is a good ësimpleí
>model strikes me as the same thing. An application problem that is
>simple for a calculus student would be deadly for a fifth grader. Unless
>of course the fifth grader had the same skills and understanding of the
>concepts of mathematics as a calculus student. Edís reference to
>speaking differently to someone with SD understanding that to someone
>who doesnít is the same reality that we as mentors face. I am actually
>doing some of the same SD activities with second graders that I am also
>doing with sixth graders. Both are new to the skills and concepts.
> No matter what ìsimpleî model I use, my question is what skill,
>concept, and basic understanding of System Dynamics will the
>model/activity build deeper understanding in for myself and my students.
>In addition, of course, to an increased understanding of the particular
>system we are discussing.
> Last fall Tim Joy facilitated a great discussion on this list serve
>about where to start and five basic concepts. In Durham at the CLE
>conference a group of us met to try to design a (I hate to say it)
>ìscope and sequence chartî. I am going to have to go through my notes
>to find that information, but this current discussion certainly hits
>home and will help put all those pieces together for me.
>
>Jan Mons
>GIST Project Mentor
>A Waters Grant Project
>Glynn County Georgia
From: "Jean-Louis Cordonnier" <jlcord@wanadoo.fr>
To: <k-12sd@sysdyn.mit.edu>
Subject: Re: Simple models become complex
Date: Wed, 13 Oct 1999
Add : If the earth temperature is colder, the eath is less dark => the
reflected light is greater => the temperature decreases (loop).
5 stella (5.1.1) models + explanations available (in french)
Cordialement
Jean-Louis CORDONNIER
36, rue Lavisse
PERPIGNAN (FRANCE)
jlcord@wanadoo.fr
----- Original Message -----
From: k-12sd <k-12sd@sysdyn.mit.edu>
To: <k-12sd@sysdyn.mit.edu>
Sent: Tuesday, October 12, 1999 9:22 PM
Subject: Simple models become complex
> Date: Fri, 08 Oct 1999
> From: "RICHARD TURNOCK" <Richard_Turnock@pgn.com>
> To: k-12sd@sysdyn.mit.edu
> Subject: Simple models become complex
>
> >>>>>>>>>.
> Pressure causes outflow rate to increase, so they are faced with a
> discrepent event. Can anyone offer something analogous in a separat
> context?
> Mary Ellen Verona
> >>>>>>>>>>>>>>
>
> Predict this graph ahead of time.
> Global warming model. Start with linear inflow and compound decay.
> Sun is linear inflow to
> Stock #1 is Earth Energy
> Converter #1 is temperature of earth to the fourth power (link stock #1 to
> Converter #1)
> Outflow # 1 is Infrared Radiation to atmosphere (link Converter #1 to
> Outflow #1)
> Outflow # 2 is infrared to space (link converter #1 to outflow #2)
> Connect outflow #1 to Stock #2 = Atmosphere
> Add a converter for temperature of atmosphere and two outflows from stock
> #2 connected in same pattern as stock #1. Except one outflow from
> atmosphere is inflow to stock #1 (Feedback loop!)
> Initial values: Sun is 1, stocks zero. (don't try other values yet!)
> Predict graph of stocks.
>
> (maybe you don't want to have to explain temperature to the fourth power
> >from the stefan boltzman equation - have prebuilt model for the students
> to exercise or the teacher uses projector to lead class discussion.)
>
> Predict graphs ahead of time:
> Have students build the model without using any multiplication of the
> temperature. Run the model. Then try it with temperature squared, then
> raise it to the third power and then the fourth power. You could have
them
> build four copies of the basic model, each with one of these equations and
> run two comparative graphs one with all the earth energy stocks and one
> with all the atmosphere stocks.
>
> If the earth energy system feedback loop is based on temperature to the
> fourth power what would you predict the graph would look like if CO2 is
> added to the atmosphere and increased the ability of the atmosphere to
hold
> more energy?
>
> Now predict what will happen if you use a value greater than one for the
> value of the sun inflow. Try values from two through 10.
>
> Richard
Date: Wed, 13 Oct 1999
From: Mary Ellen Verona <mverona@mvhs1.mbhs.edu>
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: Re: Simple models become complex
I absolutely love this model - we use it with earth science classes.
I have not considered exposing all teachers to it. Thanks for the great
ideas on using it.
Here is a new question. In working with teachers, how important is it to
stick to their primary subject area?
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, 13 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: Rolfe Stanley <rstanley@together.net>
Subject: Re: Simple models become complex
I have followed this discussion a bit. My first concern with the model is
the label "Earth
Energy"..You are talking about global warming which affects the stored
atmospheric heat of the earth. Much of the earth's energy comes from
internal heat (radioactive & original heat from deep in the earth's
interior) We do not want to mislead students in thinking that the earth's
energy comes only from the sun...Solar energy drives atmospheric
circulation and is responsible for surface processes. It takes a very long
time for solar heat to penetrate the earth.. Thus cold cellars are used to
store veg since the temperature remains fairly constant year rouond. Keep
up the good work!
Rolfe Stanley
Stanley Computer Center
Fletcher Extension
From: "Martin Clough" <M.L.Clough-95@student.lboro.ac.uk>
To: <k-12sd@sysdyn.mit.edu>
Subject: Education of Systems Thinking
Date: Thu, 21 Oct 1999
To interested k12-list users,
I am currently conducting research as part of my Masters for Loughborough
University, England into how to teach ST to individuals at undergraduate
level.
Specifically, experiments into the effect of education on systems thinking
and how this may be used to teach a philosophy or used in management to
promote systems thinking in everyday problems.
If anybody has any experience with this area or can offer advice about
proceedings and previous related research, I would greatly appreciate
the help.
Many thanks,
Martin Clough.
Dear K-12sd List,
In reply to John Leith on biological effects of ionizing radiation. John,
you may have two models in this lesson, not one.
From the cold war days, my group at MIT modeled "the arms race" (U.S. and
Russia?) which most likely would be suitable for Freshmen. I'd be happy to
send hard copies of some of that work to anyone who will e-mail me directly
with their LAND address and mention "arms race".
What about suggestions for the radiation model?
All the best,
Nan Lux
System Dynamics Group
nlux@mit.edu
Date: Tue, 19 Oct 1999
To: k-12sd@sysdyn.mit.edu
From: John Leith <John_Leith@Brown.edu>
Subject: Pakistan-India modeling
Dear k-12sd@sysdyn.mit.edu'ers.
I am looking for help. I teach a freshman class in the biological
effects of ionizing radiation, and given the recent coup in Pakistan,
I am looking for help in trying to model the relationships between
these two countries (and their allies/oppenents such as China, etc.)
that go into whether or not they may someday into a conflict in which
nuclear weapons might be used. Clearly, major factors include
religion, land, and population pressure, but I am not sure as to how
to try and set this up using STELLA.
Any advice would be greatly appreciated.
Cheers
John Leith
From: GBHirsch@aol.com
Date: Tue, 19 Oct 1999
Subject: Re: drug addiction
To: k-12sd@sysdyn.mit.edu
Dear Jean-Louis,
My colleagues Gil Levin and Ed Roberts and I published a book called The
Persistent Poppy: A Computer-Aided Guide to Heroin Policy about 25 years ago.
The book describes a model of an urban community ( a section of New York
City) developing a heroin problem. The model is an interesting one because
it goes beyond the "obvious" aspects of the heroin problem (addicts, dealers,
crime, treatment programs, police) and examines the role of the community's
response in making the problem more persistent. Unfortunately, the presence
of a significant heroin problem in many communities today may suggest that
these structures are still at work.
The book may be hard to find because it is out of print. The authors are
Levin, Hirsch, and Roberts and it was published by Ballinger (no longer in
business) in 1975. A shorter description of the work was originally
published in the American Journal of Public Health in June, 1972 and
reprinted in Ed Roberts book Managerial Applications of System Dynamics
(available from Pegasus Publishing (www.pegasuscom.com)). The article
describes the model structure, but doesn't have the model's equations and
simulation results as the book does.
You should also look at Jack Homer's work on modeling cocaine policy which
may also address some of the issues you are examining. There should be some
specific references to this work in the System Dynamics bibliography*.
Gary Hirsch
GBHirsch@aol.com
*Try <http://www.albany.edu/cpr/sds/>
The Moderator
Date: Wed, 13 Oct 1999
Subject: Re: Even Simpler models with unexpected or puzzling dynamics
From: "Timothy Joy" <tjoy@pps.k12.or.us>
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
Most of the teachers in SD are risk-takers, at least somewhere on that
arcing trace of taking risks. We have pretty much scorned assessment,
knowing full well that we need it, but nonetheless put it aside, banking
instead on our personal experience--felt experience, student's eyes,
eye-popping understanding. All these things are OUTSIDE traditional school
realms, and so we need not be surprised that students, parents and teachers
are not ready to embrace SD. If the point of school is to get into another
school, then SD will not likely engender support.
Teresa's student who dropped her class because of her fear of not passing a
test reminds us that our students must also be willing to risk. We are
taking them for an intellectual journey that ensures grand vistas, but whose
ultimate pay off we may not yet fully see ourselves. I do remember that I
am mostly fumbling along; thank God students are around to catch me and set
us all aright.
Some students, like some teachers, are not ready. And if that crazy lady
with those funny diagrams shows up in Biology, Teresa, maybe this young
student will realize that systems really are ubiquitous.
Tim
----------
>From: "k-12sd" <k-12sd@sysdyn.mit.edu>
>To: k-12sd@sysdyn.mit.edu
>Subject: Re: Even Simpler models with unexpected or puzzling dynamics
>Date: Wed, Oct 13, 1999
>
> Date: Wed, 13 Oct 1999
> To: "k-12sd" <k-12sd@sysdyn.mit.edu>
> From: teresa@northwest.com
> Subject: Re: Even Simpler models with unexpected or puzzling dynamics
>
> Jan
> I'm not quite sure what you are asking here about the population model, but
> if I am interpreting your question correctly, I think it applies to some of
> our Oregon State Standards and Benchmarks. These are stated as follows:
> Content Standard: Use concepts and processes of change, constancy, and
> measurement.
> Benchmark for this standard for Grade 10: Describe the relationship
> between constancy and change within systems.
> Eligible Content (what the state could ask kids in their assessment test
> for this standard and benchmark): Recognize that large scale constancy
> sometimes is explained by opposing small scale changes. For example, a
> population may remain constant even though individuals are added and removed.
> This is exactly the wording for one of our Science standards. I am hearing
> "balancing feedback loops" when I read this. I also ask myself, how can
> students really understand the dynamics of this situation if they don't
> have an opportunity to use something like a STELLA model. Could they even
> begin to really understand this by just looking at a textbook graph and
> tables (which is all that we've had to use until now). STELLA can make
> this an interactive, living experience. You can then look at the numbers
> along with the structure of the population model and explain to kids how
> the math works if you want, depending upon the level of the kids.
> The list of standards, benchmarks, eligible content, goes on and on
> implying ST/SD as the optimum learning approach.
>
> An interesting anecdote....One of my students questioned whether I was
> aware of the curriculum for our Global Science course, the expectations for
> the state assessment she'll take this spring AND
> she was concerned about "knowing" enough information to be able to pass the
> test. I proceeded to explain my purpose and the benefits she'll gain from
> experience with ST/SD and she proceeded to drop my class. Fortunately she
> did go into Biology, however, without the benefit of anymore ST/SD unless
> that teacher accepts my invitation to work with his kids. Wow, will she
> ever be suprised if I show up there sometime soon packing my ST/SD along
> with me!!
> Maybe I'm opening up another question for reflection here, but,
> are there other anecdotes out there similar to this and how have you
> responded? TERESA
>
> At 12:07 PM 10/8/99 -0400, you wrote:
>>Date: Fri, 08 Oct 1999 08:44:49 -0500
>>From: jan mons <jmons@glynn.k12.ga.us>
>>To: k-12sd <k-12sd@sysdyn.mit.edu>
>>Subject: Re: Even Simpler models with unexpected or puzzling dynamics
>>
>>Since it has been awhile since I have written anything, I will
>>reintroduce myself as a mentor for a Waters Grant Program in Brunswick,
>>Georgia. Like many of you I am enjoying this discussion. Several points
>>that have been brought out have hit home with me as I have recently
>>started working with elementary students.
>> As a former eighth grade math teacher I constantly struggled with
>>trying to get my students to solve application problems (I refuse to use
>>the phrase ìword problemsî). The difficulty was that their understanding
>>of the basic skills in math was so weak that I spent more time on those
>>than on the great applications problems.
>> The discussion about whether the population model is a good ësimpleí
>>model strikes me as the same thing. An application problem that is
>>simple for a calculus student would be deadly for a fifth grader. Unless
>>of course the fifth grader had the same skills and understanding of the
>>concepts of mathematics as a calculus student. Edís reference to
>>speaking differently to someone with SD understanding that to someone
>>who doesnít is the same reality that we as mentors face. I am actually
>>doing some of the same SD activities with second graders that I am also
>>doing with sixth graders. Both are new to the skills and concepts.
>> No matter what ìsimpleî model I use, my question is what skill,
>>concept, and basic understanding of System Dynamics will the
>>model/activity build deeper understanding in for myself and my students.
>>In addition, of course, to an increased understanding of the particular
>>system we are discussing.
>> Last fall Tim Joy facilitated a great discussion on this list serve
>>about where to start and five basic concepts. In Durham at the CLE
>>conference a group of us met to try to design a (I hate to say it)
>>ìscope and sequence chartî. I am going to have to go through my notes
>>to find that information, but this current discussion certainly hits
>>home and will help put all those pieces together for me.
>>
>>Jan Mons
>>GIST Project Mentor
>>A Waters Grant Project
>>Glynn County Georgia
Date: Thu, 14 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: "Peter Kahn" <peterk@mail.7oaks.org>
Subject: Re: Introduction
Curtis,
I read your note and found your idea re pervasive anonymity in complex
organizations like schools interesting. I have not thought about the use of
communication technologies as a means of addressing the sense of aloneness
that many students must feel in schools. Having a decent face to face
relational based social life myself, I see relating through e-mail or net
as bonus, enrichment kind of thing.
I did meet with a lad who was in trouble lately(grade 5) and suggested that
we spend some time on the net together, perhaps looking for a way to posit
some questions about lack of engagement, kids who don't like school, kids
who are "cool", kids who don't like their teacher, etc. We talked about
this in terms of using the net as a tool to increase knowledge. If he were
to be able to talk with kids who had some of the same challenges as he has,
they might be able to help each other and thereby contribute to the general
welfare of kids. Because I am not aware of this ever happening before, he
and I would enter into the endeavour as an adventure in learning. This is
the kind of thing that I find fascinating. I'm not sure how system dynamics
enters into the discussion. If you or others have thoughts about this or
the way that I am entering into this process with this young lad, I would
welcome thoughts.
PeterKrahn (Vice Principal-Riverbend Community School- in Winnipeg, Canada)
-----------------------
Date: Thu, 14 Oct 1999
From: "RICHARD TURNOCK" <Richard_Turnock@pgn.com>
To: k-12sd@sysdyn.mit.edu
Subject: Earth Atmosphere model
Polar ice caps? Molten earth core?
Before extending the boundaries of a model,
or changing the assumptions, I'm learning how
to understand the behavior of the feedback loops
using test inputs (Step & Pulse),
the proper Time Specs and changing
constants * temperature ^ 4.
Starting with one as the inflow and zero
in the stocks, I find the steady state values.
I want to run hundreds of experiments to understand
this basic model so I can show others.
---------------------
Date: Mon, 18 Oct 1999
From: Niall Palfreyman <Niall.Palfreyman@assyst-intl.com>
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: Re: Even Simpler models with unexpected or puzzling dynamics
Mary Ellen wrote:
> A model analogous to population (which teachers have seen over and
> over). But one in which at first glance, a constant inflow/outflow
> seems to work.
> On closer thought, though, it is found that over the long haul, the
> flow depends on the stock. An example that we do is a burette with
> slower outflow than inflow. Students are asked to predict when it
> will reach a particular volume. Pressure causes outflow rate to
> increase, so they are faced with a discrepent event.
>
> Can anyone offer something analogous in a separat context?
I may have misunderstood this example, but it sounds to me like any
situation where inflow > outflow leads to a buildup, and where buildup
increases the outflow. I'm thinking of exponential systems, of which the
following are examples from various contexts:
1. A leaky capacitor plate. More charge flows in than flows off, but
outflow increases with increasing charge on plate.
2. Migration of a lethal gene (maybe sickle-cell?) into a population.
Decrease of the gene frequency in the population is exponential in time,
but if there is an influx of the gene through migration, then the
frequency tends towards a nonzero value.
3. Depending on what behaviour you're trying to illustrate, you might
also use the actual sickle-cell situation, in which there are three
possible genotypes: AA, AS and SS. SS is lethal, so frequency of S would
tend to zero. However AA is also suboptimal in malaria zones, so that AS
is the preferred genotype, leading to a constant influx of S due not to
migration, but to replication.
4. For a more general public you could replace your burette with a leaky
bucket into which a tap is dripping.
Any use?
Niall.
--
We have only the world that we can bring forth
with others, and only love helps us bring it forth.
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
From: clee@spusd.k12.ca.us
Sender: "Curtis Lee" <clee@spusd.k12.ca.us>
To: "'k-12sd'" <k-12sd@sysdyn.mit.edu>
Subject: Re: Introduction
Date: Fri, 15 Oct 1999
Peter,
Thanks for the interest. I'm not sure of the pure systems applicability of
your mentoring situation, other than it being indicative of the types of
novel human arrangements that these technologies facilitate(i.e. this
email). I think Sherry Turkle's work on when identity meets technology is
one aspect. Another is Kevin Kelly's analysis of how networked ubiquity
changes the way patterns occur in organizations. One of the main impediments
to change in large schools has to do with these legacy "discourse" patterns
and their underlying hierarchies. You might explore how email tends to
flatten these hierarchies and allows an employee/student to "approach" the
CEO/principal in a much less intimidating way.
Thanks,
CL
--------------------------
Date: Thu, 14 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: "Peter Kahn" <peterk@mail.7oaks.org>
Subject: Re: Introduction
Curtis,
I read your note and found your idea re pervasive anonymity in complex
organizations like schools interesting. I have not thought about the use of
communication technologies as a means of addressing the sense of aloneness
that many students must feel in schools. Having a decent face to face
relational based social life myself, I see relating through e-mail or net
as bonus, enrichment kind of thing.
I did meet with a lad who was in trouble lately(grade 5) and suggested that
we spend some time on the net together, perhaps looking for a way to posit
some questions about lack of engagement, kids who don't like school, kids
who are "cool", kids who don't like their teacher, etc. We talked about
this in terms of using the net as a tool to increase knowledge. If he were
to be able to talk with kids who had some of the same challenges as he has,
they might be able to help each other and thereby contribute to the general
welfare of kids. Because I am not aware of this ever happening before, he
and I would enter into the endeavour as an adventure in learning. This is
the kind of thing that I find fascinating. I'm not sure how system dynamics
enters into the discussion. If you or others have thoughts about this or
the way that I am entering into this process with this young lad, I would
welcome thoughts.
PeterKrahn (Vice Principal-Riverbend Community School- in Winnipeg, Canada)
Date: Mon, 18 Oct 1999
From: Joe Rimback <rimback@erols.com>
Reply-To: rimback@erols.com
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: Re: Comment
Peter and/or Curtis
I have had very limited exposure to a software package called
GroupSystems from Ventana Corporation in Tucson (ventana.com) - no
relationship to Vensim by Ventana Systems in Harvard, Mass.
GroupSystem is a brainstorming package for corporate meetings which
allows everyone to 'talk' anonymously to a common display. Not cheap as
I recall. It needs an experienced facilitator to keep things moving in
a positive direction.
It would be interesting to see what would happen in a pre-college
environment. Check out their website for more information.
Joe Rimback
Gaithersburg, Maryland
rimback@erols.com
To: k-12sd@sysdyn.mit.edu
From: Lees Stuntz <stuntzln@tiac.net>
Subject: Exposition of SD work Grades 5-12
Date: Mon, 18 Oct 1999
DynamiQueST
An exposition of student and teacher work in System Dynamics and Systems Thinking, Grades 5-12
Join us for an exciting, first time ever, fun-filled event.
May 19-20, 2000
Trinity College, Burlington, Vermont
This exposition will be a forum for students to
o Showcase their work
o Share their experiences
o Educate others about their work in Systems Thinking/System Dynamics
The exposition is open to any 5-12 grade student. Students may enter any ST/SD related materials-utilizing any or all of the tools-for evaluation with the appropriate rubrics. Each student who achieves a standard of work in any of the five areas (Behavior over Time Graphs, Causal Loops, Stock-Flow maps, Computer Simulation models and Overall Understanding) will be recognized for meeting the standard. Judges will be working from rubrics that will be available to all teachers and students in the packet of information available in November.
There will also be the opportunity for teachers to share their current work utilizing System Dynamics and Systems Thinking to further understanding.
DynamiQueST will kick off with an evening of getting to know each other and participating in interesting, challenging group activities. Saturday morning, May 20th, there will be a showcase of student and teacher work, followed by optional trips and activities in the Burlington area. Meals and lodging will be provided for nominal cost at Trinity College.
This is a unique learning opportunity for students, parents and teachers. For more information (details, rubrics, etc.),
e-mail Lees Stuntz
stuntzln@tiac.netor visit <www.trinityvt.edu/waters/dynamiquest.html> or <sysdyn.mit.edu/cle/> after November 15, 1999.
DynamiQueST Committee:
Dan Barcan and Sue Jamback, Chelmsford Public Charter School
Alan Ticotsky and Rob Quaden, Carlisle Public Schools
Larry Weathers and Dick Maki, Harvard Public Schools
Will Costello, Waters Grant Project and Chittenden South School District
Steven Roderick, Lincoln-Sudbury Regional School District
Lees Stuntz and Deb Lyneis, Creative Learning Exchange
Lees N. Stuntz
Creative Learning Exchange Phone- 978-287-0070
1 Keefe Road Fax- 978-287-0080
Acton, MA 01720 e-mail- stuntzln@tiac.net
http://sysdyn.mit.edu/cle/
From: "Jean-Louis Cordonnier" <jlcord@wanadoo.fr>
To: <k-12sd@sysdyn.mit.edu>
Subject: drug addiction
Date: Tue, 19 Oct 1999
I want to build a model of drug addiction : a short trip to sky followed by
a long stay in hell.
Does somebody know a previous work about that topic ?
Looking for datas about heroin, morphin, naloxone, methadone half-life
duration.
How drug addiction affects the receptor density and is the cause of a
neurotransmitter depletion... ?
Which equation represents dependence & tolerance ?
Thanks for your help.
Cordialement
Jean-Louis CORDONNIER
36, rue Lavisse
PERPIGNAN (FRANCE)
jlcord@wanadoo.fr
Date: Wed, 20 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: Melinda Salazar <msalazar@cisunix.unh.edu>
Subject: Re: women, poverty and systems
I'm working on a project that looks at a systems approach to women, poverty
and income in the Third World (with indigenous communities). The approach
challenges the very popular Gremeen microlending method on the grounds it
is short-term and a potential 'fix that fails.' In the long-term, there
are potential deleterious impacts on family: pitting women against men for
limited resources, weakens link to family, reduces the already tenuous role
of men in the family and increases women's burden. Has anyone suggestions
for further inquiry/exploration?
Melinda Salazar
Date: Wed, 20 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: Ed Gallaher <gallaher@mail.teleport.com>
Subject: Re: drug addiction
>I want to build a model of drug addiction : a short trip to sky followed by
>a long stay in hell.
>Does somebody know a previous work about that topic ?
(snip)
>Jean-Louis CORDONNIER
Jean-Louis, and Gary Hirsch:
The Persistent Poppy, and Homer's work on cocaine policy, are addressed
primarily at the social, legal, political, and economic aspects of
addiction. These aspects are _very_ important. However, they are quite
different from the questions being asked above, which are primarily
biological, physiological, and pharmacological in nature.
Jean-Louis is asking very specific questions about the response of the body
to these drugs, and the biological mechanisms which contribute to these
responses.
I have been studying drug tolerance and physical dependence in mice and
rats for about 25 years now. Mostly alcohol and related sedative hypnotics,
but I'm also interested in the concepts and theories related to a wide
variety of other drugs such as opiates, stimulants, nicotine, caffeine, etc.
I have long had the intuitive feeling that the study of "adaptive systems"
would be relevant here, and I spent years looking for the appropriate
tools. This led me to investigate differential equations, Laplace
transforms, Bode plots, Nyquist analysis, etc. etc. etc. It was
interesting, and challenging, but I was aware of the fact that if I took
the considerable effort required to really use these tools, my colleagues
in biomedical and clinical research would not have a clue what I was
talking about.
This ultimately led me to Wayne Wakeland's computer simulation courses here
in Portland, where I was introduced to DYNAMO, and ultimately STELLA, along
with a great many other simulation tools.
Eureka!
That was 13 years ago. I have learned much, much more about SD modeling
from many of the best, including Forrester, Richardson, etc. On the other
hand, I have had a _very_ difficult time (until recently) gaining
acceptance for this powerful approach from my colleagues, granting
agencies, etc.
I am leading a seminar course this quarter for our graduate students in
Behavioral Neuroscience, in a department which specializes in these areas.
To be perfectly honest, assembling the appropriate materials is hurting my
brain . . . Not only has no one done this work, but the work that has
been done is very confusing, and in my opinion often misses the mark. I
hope to contribute to more meaningful work using SD, but it is very slow
going.
For example, addressing the last question on changes in receptor density:
On Monday morning I heard a short presentation from one of our faculty
members regarding changes in the level of GABA-A beta-2 receptor subunits.
Mice were exposed to alcohol for 3 days and then receptor levels were
measured in the cerebellum at two time points. Today at noon I heard an
almost identical experiment looking at the effects of alcohol on the kappa
opiate receptor.
As many of you realize, this makes about as much sense as trying to analyze
your shock absorbers by taking one or two single snapshots after driving
over a bump.
* * It is the TRANSIENT BEHAVIOR of the system that provides all the
information! * *
BUT, if no one does the experiments with this in mind, then how does one
find the data to build a realistic model?
The title of the journal club is "Transient Analysis of Acquired Drug
Tolerance". It starts in two weeks and runs for six sessions. Wish me luck!
Ed Gallaher, Ph.D.
Assoc. Professor
Behavioral Neuroscience and Physiology-Pharmacology
Oregon Health Sciences University
Date: Wed, 20 Oct 1999
From: "RICHARD TURNOCK" <Richard_Turnock@pgn.com>
To: John_Leith@Brown.edu
Cc: k-12sd@sysdyn.mit.edu
Subject: Pakistan
I saw your email on the K12 systems listserv.
Book: System Effects, by Robert Jervis
subtitle: Complexity in Political and Social Life
1997, Princeton University Press
The book doesn't show a stock and flow model.
It discusses the feedback loops and interactions of countries.
There are several references to the nuclear issues with Pakistan.
Maybe you could borrow it from the library
and copy the pages listed in the index at the back of the book.
Richard
Date: Thu, 21 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: "Jay W. Forrester" <jforestr@MIT.EDU>
Subject: Re: Pakistan-India modeling
You may find useful information below.
>Date: Tue, 19 Oct 1999 19:18:58 -0700
>To: k-12sd@sysdyn.mit.edu
>From: John Leith <John_Leith@Brown.edu>
>Subject: Pakistan-India modeling
>
>Dear k-12sd@sysdyn.mit.edu'ers.
>
>I am looking for help. I teach a freshman class in the biological
>effects of ionizing radiation, and given the recent coup in Pakistan,
>I am looking for help in trying to model the relationships between
>these two countries (and their allies/oppenents such as China, etc.)
>that go into whether or not they may someday into a conflict in which
>nuclear weapons might be used. Clearly, major factors include
>religion, land, and population pressure, but I am not sure as to how
>to try and set this up using STELLA.
>
>Any advice would be greatly appreciated.
>
>Cheers
>John Leith
Date: Wed, 20 Oct 1999
From: G P Richardson <gpr@csc.albany.edu>
To: k-12sd <k-12sd@sysdyn.mit.edu>
Subject: Re: Pakistan-India modeling
On Wed, 20 Oct 1999, k-12sd wrote:
[...]
> I am looking for help in trying to model the relationships between
> these two countries (and their allies/oppenents such as China, etc.)
> that go into whether or not they may someday into a conflict in which
> nuclear weapons might be used. Clearly, major factors include
> religion, land, and population pressure, but I am not sure as to how
> to try and set this up using STELLA.
This is an enormously complex problem. STELLA won't help until we have a
pretty good conceptualization of the intricate and complex structure of
resources and pressures acting in this charged socio/politcal/economic
system. I doubt this is a good choice for a freshman class, even at
Brown. It might be a good PhD dissertation.
But one way to proceed might be to ask groups within the class to focus on
some part of the mess, e.g., the dynamics of religious differences in the
area, the dynamics of population groups, the dynamics of economic change,
the dynamics of arms investments and alternatives to armed conflict, the
dynamics of national pride, the dynamics (stability/instability) of
political factions, and so on.
The tools for focusing would be graphs over time, causal maps, and a view
of stocks as resources and traditions (as well as simple accumulations
like population groups). The goal of each group would be extreme clarity
in a relatively small area so that when all the views are combined a rich
complex web of interacting forces gets laid out. It would be like having
an economist's view interwoven with a sociologist's view, and those with
an ecologist's view or political scientist's view. No one perspective is
going to capture this mess.
I would usually advocated starting simply and getting a simulation model
going quickly and build to complexity, carrying understanding with you at
every step. But here it would be so easy to be superficial -- "it's an
arms race, and that's a positive loop, so we're in trouble..." And I'm
afraid that premature quantification in STELLA or Vensim or whatever would
force the thinking to be pretty superficial. Almost seems that something
this difficult cries out first for gathering the complexity so that all
see just how impossibly complex it is and all can begin to agree on where
the dynamic uncertainties are.
Then the reason for moving to STELLA would be clearer -- we would claim to
know a lot about the system, but be unable to mentally simulate it, so
simulation modeling would help us. Maybe.
Of course, we could make a mess of the causal mapping if we allow
ourselves to think glibly -- "religion," for example, is crucial in this
problem, but it is not a quantity exerting pressures in the system. We'd
have to tease out what about the various religious tendencies create
social, economic, and political pressures in this mess. This would be
tough but instructive. So students would have to be careful guided to
turn ideas into forceful maps with definite (if uncertain) dynamic
implications. (Some nouns, like "religion", just don't belong in a causal
map; we'd replace such concepts with things like the pressures that
produce growth or decline in the attractiveness of jihad in the Moslem
population, or evangelical tendencies in Hindus or Christians, or
whatever.)
I think the dynamics of ozone holes would be a lot easier.
...GPR
-----------------------------------------------------------------------
George P. Richardson G.P.Richardson@Albany.edu
Chair of Public Administration and Policy http://www.albany.edu/~gpr
Rockefeller College of Public Affairs and Policy Phone: 518-442-5258
University at Albany - SUNY, Albany, NY 12222 Fax: 518-442-5298
-----------------------------------------------------------------------
---------------------------
To: k-12sd <k-12sd@sysdyn.mit.edu>
From: Linda Booth Sweeney <Linda_Booth_Sweeney@harvard.edu>
Subject: Re: Education of Systems Thinking
Date: Fri, 22 Oct 1999
Martin,
i have some experience in this area and would be happy to share information
with you. First, it would help me to know what you mean by Systems
Thinking? What concepts and/or tools are you trying to teach?
Regards,
Linda Booth Sweeney
Harvard Graduate School of Education
75 Reservoir Street
Cambridge, MA 02138
Tel: 617-354-1390
Fax: 617-491-3496
e-mail: Linda_Booth_Sweeney@harvard.edu
>From: "Martin Clough" <M.L.Clough-95@student.lboro.ac.uk>
>To: <k-12sd@sysdyn.mit.edu>
>Subject: Education of Systems Thinking
>Date: Thu, 21 Oct 1999
>
>To interested k12-list users,
>
>I am currently conducting research as part of my Masters for Loughborough
>University, England into how to teach ST to individuals at undergraduate
>level.
>Specifically, experiments into the effect of education on systems thinking
>and how this may be used to teach a philosophy or used in management to
>promote systems thinking in everyday problems.
>
>If anybody has any experience with this area or can offer advice about
>proceedings and previous related research, I would greatly appreciate
>the help.
>
>Many thanks,
>
>Martin Clough.
Date: Thu, 21 Oct 1999
To: "k-12sd" <k-12sd@sysdyn.mit.edu>
From: Stuart Kermes <skermes@eastconn.org>
Subject: Re: women, poverty and systems
Melinda,
You may want to look at some of the development work of the Heifer
Project. They donate livestock to Third World families so that they can
have an ongoing way to produce food. Part of the arrangement is that
receipients must donate the first offspring of their animal to a neighbor.
I believe that in Tibet they ran into the same issues of empowering women
and stressing the traditional