Learning objective
Translate a behaviour-over-time pattern into evidence-qualified reinforcing and balancing loops with explicit delays and leverage tests.
Why it matters
The same action can relieve a symptom now and strengthen its cause later. Without feedback and delay, the rebound looks surprising instead of structural.
Core concept
The first loop balances backlog in the short term. The delayed capacity pathway can reinforce deterioration.
Scenario
Support backlog falls after overtime, rebounds after two weeks and is followed by attrition months later.
Worked decision
Keep the short-term response, fatigue pathway, error rework, experienced-capacity loss and hiring delay in one model. Test protected recovery, demand reduction and error prevention with the overtime policy.
Decision rule
Intervene on a loop only when the causal links, delay, expected pattern and rejection signal are explicit.
Ethical boundary
Do not model worker capacity as an unlimited machine variable. Include fatigue, agency, workload distribution and the voices of people who carry the intervention cost.
AI lens
AI can check diagram syntax and evidence traceability; humans own causal judgement, boundaries and acceptable consequences.
Quick check
A loop contains one negative link. How is it classified?
A. Reinforcing.
B. Balancing.
C. Beneficial.
D. Invalid.
Answer: B, assuming the loop closes and the polarity is correct.
Next step
Complete the support-backlog case and identify the delay that most changes the intervention choice.