Teams now need to answer the question: Why is it changing? This involves drawing the underlying system by linking variables according to their cause-and-effect relationships. Different tools are available to do this. Here, we will use a simple “connection circle”:
- List the most important variables in the story (10 to 15 variables).
- Write the variables around a large circle.
- Find a variable that directly causes another variable to increase (e.g., +consumption of single-use plastics -> +plastic waste) or decrease (e.g., +recycling -> -plastic waste).
- Draw an arrow from the cause to the effect.
- Add polarities to each arrowhead to describe the relationship: use a “+” if an increase in the first variable is accompanied by an increase in the second; use a “-” if an increase is accompanied by a decrease.
- Repeat steps 3, 4 and 5 until you can tell the story by moving from one variable to another along the arrows.
Once all arrows are drawn, students should check for feedback loops. A feedback loop occurs when a variable is both a cause and an effect in the same causal loop.
If the problem statement describes accelerated growth or decline, there should be a reinforcing feedback loop in their connection circle (e.g., + use of plastic packaging > + consumer convenience > + demand for plastic packaging > + use of plastic packaging).
If the problem is one of slowing growth or decline, balancing feedback may be at work (e.g., + use of plastic packaging -> + plastic waste -> + plastic recycling -> – plastic waste).
Tip: A causal loop has a reinforcing action when there is an even number of (or zero) “-” polarities. A causal loop has a balancing action when there is an odd number of “-” polarities.
It’s a good idea to name a loop to remember its role in the system (e.g., a plastics addiction loop).
Teams can also identify delays and levels in the system. A delay is an interruption between an action and its consequences. They can be indicated by two horizontal bars across an arrow. A level is a variable that accumulates over time (e.g., the volume of plastics in landfill). It can be labelled with a small box.
The finished model should be able to explain the behaviour as described in the problem statement (step 2). At this stage, it’s a good idea for teams to validate and correct their model with domain experts.