ISE548 Chap.10 Knowledge Gaps and Risk Communication
Knowledge Gaps and Risk Communication
Different gaps require different actions
Uncertainty can sit in hazards, estimates, qualitative language, incomplete knowledge, competing values or communication. This chapter maps known knowns, known unknowns, unknown knowns and unknown unknowns to operation, analysis, elicitation and adaptation.
It shows why not every gap should be forced into a probability and why a useful message states the evidence basis, consequence of error and update rule.
An assumption register complements the risk register by recording what must remain true for the analysis to hold. Each important assumption receives an evidence status, owner and invalidation signal.
Communication then translates the estimate and its limits into an observable trigger, authorised response and next review point for the intended audience.
What this chapter covers
- 01
Uncertainty in assessment
- 02
Uncertainty in management
- 03
Knowledge-gap quadrants
- 04
Qualitative probability language
- 05
Competing values
- 06
Communication triggers and feedback
Worked example · free
Route three port uncertainties
- 3Classify the three gaps.
- 3Choose an information action.
- 3Design a communication trigger.
Key terms
- Known Unknown
- A recognised gap whose outcome or value remains uncertain and can be analysed through ranges or scenarios.
- Unknown Known
- Relevant knowledge present in people or records but absent from the active decision.
- Unknown Unknown
- A condition not yet recognised, managed through resilience, monitoring and adaptive capacity.
- Update Rule
- The observation, date or event that requires an assessment or decision to be revised.
Knowledge Gaps and Risk Communication FAQ
Should every uncertainty receive a probability?
No. A recognised variable may support a range, hidden organisational knowledge calls for elicitation, and an unforeseen condition calls for resilience and monitoring. A convenient number can conceal the actual information problem.
How can qualitative words be made useful?
Define labels such as likely with probability or frequency bands, state the evidence and test whether the decision changes across the range. Shared words without shared thresholds often create false agreement.
Can more evidence resolve conflicting values?
Evidence may narrow likelihood or consequence, but it cannot decide how much safety, cost, speed or fairness should matter. Record the trade-off and decision authority instead of asking analysis to make a governance choice.
How should a known unknown be handled?
Define the unresolved quantity or outcome, build credible ranges or scenarios and test whether the decision changes across them. Identify the observation that would narrow the range and the deadline by which it matters. A single precise estimate can hide the gap. The goal is not to eliminate uncertainty but to show which decision remains robust and which evidence is worth collecting.
How can hidden organisational knowledge be surfaced?
Interview operators and suppliers, review work orders and exception records, observe the task and rehearse the proposed response. Compare formal procedure with actual work without treating informal adaptation as automatically wrong. Record who knows the workaround, when it applies, its limits and the controls it bypasses. Converting experience into shared evidence reduces dependence on one person.
What is a useful response to an unknown unknown?
Do not invent a probability for an event that is not yet characterised. Build detection, spare capacity, modular choices, escalation, recovery and learning so the organisation can adapt when an unfamiliar signal appears. Define general stress indicators and decision authority. Resilience does not predict the surprise; it limits how long the system remains unable to recognise and respond to it.
How should uncertainty be communicated to different audiences?
Operators need the observable trigger and action, managers need the range, consequence and authority choice, and technical reviewers need assumptions and evidence. State what is known, estimated and assumed, then explain the consequence of being wrong and the update rule. Ask for feedback or rehearsal because message delivery does not prove comprehension or ability to act.
What belongs in an assumption register?
Record the assumption, why it matters, current evidence, owner, review date and observable invalidation signal. Link it to the decision or control that depends on it. Separate assumptions from confirmed facts and unresolved questions so a reader can see where confidence comes from. When a signal invalidates an assumption, reopen the affected assessment rather than quietly editing the wording.
Exam move
For one uncertain project, list what is known, recognised but unresolved, known by others and not yet foreseeable. Assign a different action to each. Then rewrite the project update so it states the estimate, basis, consequence of error, next signal and person authorised to respond.
Working through Knowledge Gaps and Risk Communication in ISE548? Sia is AskSia’s AI Management tutor — ask any ISE548 Knowledge Gaps and Risk Communication question and get a clear, step-by-step explanation grounded in how ISE548 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.