British Petroleum Is Not Jackson Middle School: Different Best Modeling
Practices For Different Environments
Ron Zaraza Tim Joy Scott Guthrie
CC-SUSTAIN Project CC-SUSTAIN Project CC-SUSTAIN Project
Co-Director Core Team member Core Team Mmember
Wilson HS La Salle HS Wilson HS
1151 SW Vermont 11999 SE Fuller Ave 1151 SW Vermont
Portland OR 97219 Milwaukie OR 97222 Portland OR 97219
(503) 235-2745 (503) 659-4155 (503) 916-5280
rzaraza@ pps.k12.or.us lasalle4@ northwest.com sguthrie@ teleport.com
Educators attempting to bring the concepts of system dynamics into the classroom have
always looked to the experiences of systems dynamicists for guidance. Until recently, the
only training in system dynamics focused on its traditional policy and business uses, so
teachers followed the ideas of the leaders in the field with great faithfulness. The
importance and impact that system dynamics pioneers have had in the development of
educational uses of systems is evident in the role people such as Jay Forrester and George
Richardson have played at the K-12 Systems conferences. Their participation in the K-12
systems in education listserv (k12ed@ sysdyn.mit.edu) is further evidence of their interest
and commitment to the use of System Dynamics in the K-12 environment. Thus, it is
only appropriate that a discussion they were involved in on the list-serve provided the
motivation for this paper.
The question of “best”, “preferred”, or “most probably successful” practice for model
construction became the focus of a discussion. This was a result of a question about the
possibility of modeling a high school as a system. George Richardson observed that
“Since the early days of system dynamics it has generally been agreed that one can not
build a model of a system (like a high school), but instead one must take a problem or
interrelated set of problems as the focus of the model.”! He suggested that modeling the
system, rather than a well defined problem within the system, would be difficult, if not
impossible. This response provoked a series of exchanges lasting about a week, in which
some participants asserted that the need to focus on problems, rather than a general
system, was unnecessary, while most supported Richardson. An interesting insight came
from Jay Forrester, who, while supporting Richardson, noted that changing George’ s
assertion to “it has generally been true”? might well resolve the issue, since it is a less
absolute statement. The issue, however, was not one that participants were willing to
drop.
It quickly became clear that the underlying problem was understanding model
conceptualization. What are the necessary (or at least useful or most probably successful)
steps or components in conceptualizing and constructing a model? As the discussion
developed, Richardson shared the components of model conceptualization he uses with
his students:
1George Richardson in a message to the K-12 listserv dated 12/31/97
2Jay Forrester in a message to the K-12 listserv dated 12/31/97
Problem focus*
Problem Dynamics*
Context*
Audience*
Model Purpose*
Model Boundaries
-temporal
-conceptual
-causal
Aggregation
Reference Modes
Initial Policy Options
Model Sectors
Important processes in each sector
Important levels and associated rates in each process and sector
Apparently important feedback loops
Next steps?
The components identified with an asterisk (*) are considered to be the most crucial to
the modeling process.
Richardson's components focus attention on a problem, not on a system. Listserv
comments from other participants who are actively involved in the modeling of dynamic
systems tended to support the emphasis on problem identification and delimitation as
primary factors in successful model building. This approach to model conceptualization
has also seen wide-spread use in educational applications of system dynamics. It is the
ability of system dynamics to address interesting and exciting problems outside the reach
of conventional tools and methodology that often first attracts teachers. Among members
of the CC-STADUS/CC-SUSTAIN core team most initially saw system dynamics as a
better way to teach and solve problems in mathematics, physics, history, biology, and
other disciplines. This focus is consistent with the “best” or “common” practice of
focusing development of a model on a clearly defined problem. It is also an entirely
appropriate use of system dynamics in education.
The educational materials developed at the K-12 level usually follow this approach. The
materials are designed to address a clearly defined problem, develop a model to illustrate
it, then use the model to obtain actual numerical results. Several curriculum packages
available from the Creative Learning Exchange provide excellent examples of this
approach to using systems dynamics. The Radioactive Decay package consists of
preliminary text and models about linear and exponential growth and decay, simple
models and questions about radioactive decay, and finally, exploration of radioactive
decay sequences using both a model and questions. The entire curriculum package is
defined by focusing on the problem of radioactive decay. Activities develop both the
concepts and models, providing an alternative to traditional approaches. The final
activity, in which students construct a 3- element radioactive decay sequence, includes a
model that yields numerical results which cannot be obtained with the mathematics
3George Richardson in a message to the K-12 listserv dated 1/2/98
normally available to K-12 students. Thus, the models are used to solve increasingly
complex variations of the basic problem.
A variety of predator-prey relationships are also examples of the problem-focused use of
modeling. Because the interrelationships are impossible to describe quantitatively
without very advanced mathematics, they have frequently been subjects of dynamic
models used by teachers. They provide the only vehicle which is able to explore the
problem of how species interact and yield numerical results. They are also excellent
examples of the problem-focused modeling activities that constitute the vast majority of
models and curriculum currently in pre-college use.
Since content- specific ideas are an important part of the curriculum in every discipline,
the use of system dynamics as a problem solving tool is an appropriate and powerful use.
However, it is important to guard against system dynamics remaining only a tool. For
system dynamics to have its maximum impact on student learning, it is necessary to look
at the real potential of systems concepts. Teachers repeatedly justify the use of models
and system dynamics as a way of “getting students to ask better questions”. The
specific- problem oriented models do develop this ability to some degree. However, the
real power of system dynamics is its potential to equip students with the ability to look at
real-world systems and begin to ask questions that will build an in depth understanding
of the system. Leaming through system dynamics can develop the critical thinking and
analytical skills that will allow students to make intelligent evaluations about the complex
problems and systems they will encounter. In order to maximize the possibility for this
type of intellectual growth, modeling and systems concepts must not only focus on
narrowly defined problems. Much broader and ambiguous problems can and should be
used as a starting point, since this is the pattern students encounter in their daily lives.
Models that develop the questioning skills are often not problem- specific, that is, they do
not look at a well defined and delimited problems. Instead, they tend to be models
painted with a “broader brush”, lacking in well defined boundaries or details. Looking to
the “best” practice of system dynamicists, teachers often reject such models because they
are not “accurate” enough, because they don’t allow students to draw definite conclusions
about clearly defined questions, they don’t give enough “facts”. The models are often
described as “ambiguous”. These types of models often trigger criticism of the model
and even the idea of modeling.
An experience that a number of Portland (Oregon) area teachers have had illustrates this
type of situation. Global studies is a freshman (9th grade) course taught in many local
schools. One of the unifying ideas that teachers often use in this class is the role
population growth plays in the difficulties experienced by developing nations. To
explore this, they use a simple population model with a single stock, population, a births
inflow and a deaths outflow. The model produces simple exponential growth. When run
for a period of a hundred years, the results are dramatic and frightening. Nations like
Malawi show a growth in population of 2200% or more.
Almost inevitably, the students react to the model by talking about the hormible situation
that the people will find themselves in. Often the teacher gives them a deceptively simple
assignment: develop a policy that can be implemented to prevent disaster. The result is
unrealistic, and not satisfying. Policies show little understanding of the system, because
the system hasn’t been explored. This lack of realism may motivate teachers to step away
from the use of models. The problem is not with the model, but with the educational use
of the model. Much greater benefits can be gained if the simple model is seen as the
“doorway” to the system. Stopping at the door yields little knowledge. Going through it
and exploring is riskier, but can yield great benefits. This riskier path is also followed by
some teachers.
In most classes, one or more students react negatively to the model results. They say that
the situation will not develop as modeled. The criticisms they express vary, but
frequently include:
The birth rate won't stay that high. (for a variety of reasons)
The death rate will increase. (for a variety of reasons)
There will be massive starvation
They will add more farmland.
They will practice birth control.
They will emigrate.
They will import food.
They will impose family planning.
They’ re not that stupid!
These comments can present a teacher with an opportunity to explore the system in
greater detail, to step fully through the door. This exploration begins with questions
about the comments. What are you assuming? What do you know? What do you think?
What other information do you need? Where can you find the information? (Note the
similarity to Richardson’s components!)
These questions lead to a variety of activities. Students can do individual research about
important factors that are revealed by both teacher and student questions. They can
explore interconnectedness of the topics and questions. Further modeling is a possibility,
either by students, the teacher, or a cooperative effort. Of course, each of these activities
can initiate the cycle all over again. The choice is up to the instructor. The result,
however, is an appreciation for the inherent complexity of most problems we encounter.
Rather than being surrounded by simple cause-effect relationships, students are faced
with increasingly complex situations in which small actions can have large reactions.
This pattern leads to in depth learning, as well as initial experiences in exploring a
system. The understanding of system dynamics grows more dramatically than in
problem- focused activities. System dynamics concepts become an implicit part of the
syllabus, rather than a tool for learning content.
This approach to using system dynamics is both riskier and more time intensive, but has
the potential of greater gain. When the progress of the course is defined by the questions
students ask, where the class is going is uncertain. Following the student’ s interest may
increase student involvement, but it can also expand the time spent in a single area. What
makes the approach even more difficult is development and selection of suitable “initial”
models. There are few examples of such models to get teachers started. Systems suitable
for exploration exist, but emphasis has been in the direction of more clearly defined
problems. It is clear that major efforts must be made to identify systems and additional
questions that lend themselves to this approach before this alternative use of systems can
grow beyond a few practitioners. Once the number of such simple models and
curriculum materials grows, more teachers can begin to pursue systems- focused model
use as well as problem-focused modeling.
Clearly in education there will ultimately be two distinct “best” practices in modeling.
One, following the widely accepted practices of traditional systems modeling, focuses on
well defined problems, developing models that build a detailed understanding of the
problem. The other appears to be almost diametrically opposed. It looks at systems in
the broadest sense as an initial step, building understanding of the system over time.
There is no well defined problem. The exploration of the interconnectedness among and
within systems itself is the “problem” These two different practices reflect the fact that
the use of systems in education addresses different needs than in traditional system
dynamics. While solving and explaining defined problems is important in both education
and traditional uses of system dynamics, education has a larger and more vital task -
development of thinking skills. System models that generate more questions than they
initially answer may well be the most powerful “tool” for developing this most important
of skills. The transition to use of systems concepts in daily life is quick and obvious.
Even system dynamics novices find it difficult to watch the evening news or read a
newspaper without wondering how businesses/governments/individuals can propose and
implement simple solutions to what are obviously problems arising in complex systems.
Experiencing that same kind of situation in classes will build that realization in students.
The two “best” modeling practices actually complement each other. Systems-focused
models are inherently simple in structure, serving as “triggers” for the questioning
process. Their purpose is not to address problems, but rather to expose their presence in a
system. Problem-focused models are designed to answer well defined problems.
Frequently, the system-focused models allow students to generate more specific
problems that can be the focus of models. Even if models are not used, system concepts
are focused on the questions and problems revealed by discussion of the systems models.
Extensive work is being done in this area by Scott Guthrie and Megs Patton of Wilson
High School (Portland, OR) who are teaching a Science-Technology-Society/W orld
issues course using system dynamics. They report remarkable development of insights
into complex issues by high school juniors and seniors.
With the realization that the purposes of system dynamics in education are more varied
than in traditional system applications, it is possible to wonder whether or not it is
possible to define the components of model conceptualization for educational uses as
George Richardson has for general systems work. Returning to his components, it is
clear that he has already done the task for educators. His components for general systems
work are identical in problem- focused educational modeling. Even for those very
different system-focused models, his components work quite well. What remains is a
matter of definitions or clarifications. Looking at developing systems and thinking skills
in students, Richardson's “problem focus” becomes a broad rather than a specific
problem. Exposing and exploring the interconnectedness of a specific system rather than
one problem in the system becomes the focus of the model. “Model purpose” becomes
an emphasis on raising questions rather than answering them. All other components
transfer to the different approach with little significant change. The components of
conceptualization truly are generic, with only the context changing. They serve as a
useful guide for both “best” practices in education.
References
1. Forrester, Jay W. 1985. “ ‘The’ model versus a modeling ‘process’ ,” System
Dynamics Review. 1 (1): 133-134
2. Richardson, George P. (Editor), Modeling for Management: Simulation in Support of
Systems Thinking (International Library of Management). Dartmouth Pub. Co, 1996
CC BY-NC-SA 4.0