Protocol
Draw a connection circle when causes chase each other
Some problems are not chains. They are loops wearing a chain's costume.
When it fits
- Several variables influence one another and a linear root-cause list keeps producing contradictions.
When to avoid it
- A causal arrow is a model, not proof. Avoid diagrams so dense that every variable appears to cause every other variable.
Why it matters
Place the important changing variables around a circle. Draw an arrow only when you can state how a change in one variable is expected to influence another, including direction where useful. Follow the links until reinforcing or balancing loops become visible. Mark weak links as hypotheses that need evidence.
Steps
- Choose a small set of variables that can meaningfully increase, decrease or change.
- Connect variables only when you can state the proposed influence.
- Trace closed loops and note likely delays.
- Mark uncertain links for observation or testing rather than polishing the diagram.
An example
Workload, queue age, overtime, defect rate and rework can form feedback loops that a flat cause list misses.
Check your result
The map reveals at least one circular dependency or explicitly shows that the suspected variables do not form one.
Keep this limit in mind
- A causal arrow is a model, not proof. Avoid diagrams so dense that every variable appears to cause every other variable.
Connected ideas
Useful withAsk 'and then what?' before taking the first-order win
Evidence and sources
The Waters Center presents connection circles and causal connection maps as tools for identifying interdependencies and causal links in systems.
A drawn causal connection is a model to test, not proof that the relationship exists or has the assumed direction.
Tools of Systems Thinking Courses · Causal Connection Circle Mapping