Stock and flow diagram of adoption Diagram cre...

Image via Wikipedia

Teaching about System Structures is very hard to do, but vital

February 2011

System Dynamics is both worldview and skill set – a specific framework for understanding complex systems via a method that seeks the endogenous sources for behaviors of interest to us.  At the outset of this particular query strand, I posited Donella Meadows’ Thinking in Systems kernel idea – systems have three parts: elements, behaviors, and structure.  It is relatively easy to identify the elements in a system, a bit more difficult to identify behaviors, and strikingly difficult to describe the structure generating the said behaviors.  Nonetheless, in the systems tradition of Jay Forrester/Barry Richmond/Donella Meadows, defining the structure of a system as the interconnections of stocks, flows and feedback loops are the fundamental feature of system dynamics.

To that end, let’s take a closer look at Meadows’ assertion and our own response to how we teach structure.

Meadows contends that behaviors tell us a system’s purpose, something quite different from a stated goal.   For example, a high school’s stated goal may be to prepare students for college, but drop out rates in the 30 – 40 % range belie that goal; the school’s purpose or function – what is actually occurring – may more accurately be described as to separate, or sift the college ready from the unqualified masses.  This sample helps us know that system behavior is what we are seeking to understand and, thereby, affect to some end.

Diana Fisher (  comments that mathematics is “always trying to associate behavior and structure, except that the structure we traditionally use is a function formula.”

STELLA model using stock/flow and feedback structure to denote exponential growth

Using generic stock/flow and feedback diagrams to represent traditional mathematical behaviors is very helpful indeed since “students seem to be able to remember the diagram better than the equation.”

There are tremendous benefits here, when using the generic structures.  Complex and, therefore, highly unpredictable behaviors ensue when one or more structures combine.  Students heretofore successful in traditional modes find these exercises daunting.

Sarah Boyer ( states that “when we are methodically replicating someone else’s model our own ideas come about regarding structural experiments we would like to run.”  How true.  The reconstruction of proven models helps young students underscore what she called “structural systems thinking.”

Anne Lavigne ( ) reminds us of Donella Meadows famous use of the Slinky© to demonstrate how we are often hood-winked by event-behaviors into utterly misreading what’s a causal agent (see Thinking in Systems, pp1-2).

Paul Newton ( focuses on two structures: stock/flow structure and feedback structures.  Starting with water pouring demonstrations that lead to stock/flow models of the water pouring helps adult learners see a correlation between something real and a model, and between the structure of something and what it actually does.  With water pouring especially, the coupling of the physical (water poured from one glass to another) with the metaphysical (when to stop pouring) helps students understand the feedback structure not readily apparent : tell them to pour with eyes shut!

Finally, Niall Palfreyman ( stated that “the structure of an SD model is pretty fixed by comparison with the model’s behavior.”  This is a wickedly true statement that often confounds students:  how can a single model structure generate such variety in behavior?   He further noted that “modifying [model] structure . . . is difficult” as one may fear getting it right.  How do you know? He describes this task of altering as “daunting.”  We need to encourage students to engage in “this second-order kind of thinking” which focuses on changing the “structure underlying the behavior.”

While it’s clear that we have models and demonstrations that help students understand a few things about structure, this remains a difficult domain of our instruction.  In the same way we might find few college graduates who adequately explain exponential change, we might find few of our system dynamics students who could adequately explain structure and how it drives behavior.

I was personally struck by Niall’s “nagging sense of worry” in this area of structure. Is this understanding of system dynamics a product that comes only after years of practice?  Is it an idea that we can teach our students? Are there inherently demonstrable and knowable differences between two-stock structures and three- or –four-stock structures?

Perhaps more questions come to you.  I hope so.

Questions/concerns that emerged:

  • How do we overcome our own fear of altering a model’s structure for classroom use?
  • Why is this such a difficult lesson to teach?
  • Should we, perhaps, NOT teach structure, but focus on elements and behaviors?  Is structure an idea that comes LATE in the learning of systems?

Models/behaviors referred to in comments (see Creative Learning Exchange []  for lessons and samples

  • Linear change
  • Exponential growth and decay
  • Parabolic behavior
  • Sinusoidal behavior (oscillations)
  • Logistic behavior (S-shaped)
  • Convergent behavior (goal-seeking)
  • Drug Model
  • Population Mode

What Teachers Should Know About Writing and Modeling

Click on this typewriter - go to Creative Learning Exchange for Tim's "Writing and Modeling" monograph

The intellectual step of distinguishing a stock from a flow to constructing system dynamics models is awkward and many stumble, but using a notebook in this early stage can help students (and teachers) make these strides with some ease and purpose. Teachers will find the necessary guidelines for using a systems notebook as well as a long list of writing and mapping exercises that integrates the writing process with system dynamics instruction. These guidelines and exercises are especially helpful for middle and high school students who are building some early confidence in basic stock and flow mapping as well as those students ready for model building and testing.

The writing process offers not only a good metaphor for model building; it also offers a means to composing the models themselves—a clearly told story will help the model builder, and a well built model will help the story teller.

Students discover they cannot write accurately and effectively without a sound understanding of the feedback, nor can they trace the model feature terribly well without having appropriate and consistent language. The more fluent students become at telling the story, the more facile they are at building the model.

Every page of a systems notebook reveals the mental model of its creator. These are what we wish to expose, to lay bare, to test, to evaluate, and ultimately to improve—we want students (and ourselves) to have better mental models.

What System Dynamics Teachers Should Be Teaching

Open the school to the neighborhood – the inner city, the creek in the suburbs, the industry moving next door, the new 165 house subdivision coming in down the street.  Look around to what is happening in your school’s neighborhood, and study that.  Your school, and its amazingly curious students, whether they be elementary or high school, should be a beacon of research and local knowledge to your community.

The whole point of System Dynamics is to understand the closed loops that regulate the physical world around us.  If we can do this, we will live better, be happier, provide for the longevity of our community, and find those Jeffersonian things: life, liberty and that other thing.

Give students the basic tenants of system dynamics and let them learn.  It is not events but systems that tell the primary story of our lives.  It is not cause and effect, but that stock values control flows – everywhere, all the time.  Larger systems take a long time to correct – consider the difference between steering a bicycle and steering an ocean liner. Look for patterns.

A student body can build a considerable cache of knowledge and understanding about a place – seniors pass along knowledge to underclassmen all the time, mostly without adults knowing about it.  Suppose we had plan.  Imagine a community of 300 learners, aged 14 to 18, working intermittently on a the water quality of a neighborhood’s watershed.  What a tremendous service that would be, what a boon to the civic-mindedness of our youth!

Open the doors.  Look at what needs doing and needs learning locally. And then go learn!

What Teachers Should Read in System Dynamics

These titles and annotations were generated from a question by Niall Palfreyman: Can anyone give me a “Top 3” list of books which every teacher using SD should know about, and on the basis of which three books that teacher can get up and running on using SD in the classroom?

de Rosnay, J.: Macroscope – For introducing systems, my favorite book. Full text is available at A veritable treasure trove

Douglas, J., Morecroft, W. & Sterman, J.D.: Modeling for learning organizations.

Ford, A.: Modeling the Environment – Great for starting to apply SD to messy, real-world systems.

Forrester, J.: World Dynamics specifically chapter one and chapter seven (The Epilogue).  I read this, really read it, the equations, too, after working with system dynamics for a few years.

Kaufman, D.: “Systems One: An Introduction to Systems Thinking (Identifying the Larger Pattern of Interconnections)”, The Innovative Learning Series, Minnesota. – Excellent ~40 page introduction to systems thinking

Meadows, Donella.  The Global Citizen. Island Press, 1991.  This wondrous collection of articles from Meadows’ syndicated column of the same name is the seminal work of professional scientist reaching out to a lay audience. See especially the opening section called “System Dynamics Meets the Press.” For how the everyday is re-understood from a systems point of view, this is without parallel.

– – – – -.  Thinking in Systems.  Sustainability Institute, 2008.  Meadows’ posthumous book underscores the lifework of one of systems’ best communicators.  A must.

Quaden, R., Ticotsky, A. & Lyneis, D.: The Shape of Change, CLE, Acton, Mass.

Richmond, B.: An Introduction to Systems Thinking – presents the elements of modeling as analogous to the language we use to tell stories. May be downloaded from isee site.

Sweeney, L.B.: When a Butterfly Sneezes – a clear primer on the nature of systems.

Sweeney, L.B. & Meadows, D.: The Systems Thinking Playbook,

2 comments on “SystemsThings

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s