6.1 Facts, Assumptions, Uncertainty & Constraints

Facts, assumptions, uncertainty and constraints

Incomplete does not mean random

MFS explicitly names making decisions with incomplete information as an assessment area. The public page does not publish a proprietary decision model, so preparation should focus on general evidence discipline. Incomplete-information questions provide enough material to compare responses even though no option offers perfect certainty.

Begin with four columns:

This prevents an inference from silently becoming a fact.

Suppose a delivery has not arrived by 14:00. Known: it was scheduled for noon; the receiving desk has no delivery record; a task depends on its contents at 16:00. Inferred: the courier is delayed. Unknown: whether the item was dispatched, delivered elsewhere, or cancelled. Constraints: two hours remain and an alternative supplier needs ninety minutes. Calling the courier or sender is more useful than debating unverified causes.

Observation versus interpretation

“The indicator is flashing amber” is an observation. “The device is about to fail” is an interpretation unless the scenario defines the indicator. “A colleague has not replied” is an observation. “The colleague is ignoring the request” assigns a motive without evidence.

Translate loaded descriptions into neutral facts. This reduces confirmation bias—the tendency to seek evidence supporting an early theory—and attribution error, where behaviour is explained by character rather than circumstances.

Source quality

Weigh information by proximity and reliability:

1. direct observation or authoritative current instruction;
2. verified record or qualified report;
3. consistent independent reports;
4. single unverified account;
5. rumour or speculation.

Reliability is not absolute. A direct observation may be incomplete, and an old official record may no longer reflect current conditions. Note time, scope, and whether sources are independent.

Identify decision-relevant unknowns

Not every missing detail deserves investigation. Ask: Could this fact change which option is best?

For a room booking conflict, the organiser’s favourite colour is irrelevant. The number of participants, alternative rooms, required equipment, and start time can change the decision.

Prioritise questions by value:

• likelihood that the answer changes the decision;
• consequence of deciding incorrectly;
• time and cost to obtain it;
• reliability of the available source.

A one-minute check that could prevent a serious error has high value. A lengthy investigation into a low-impact detail may be delay disguised as diligence.

Constraints and thresholds

Some facts act as hard constraints. An option requiring authority the candidate does not have is not feasible. A response that misses a stated deadline fails even if otherwise efficient. Convert constraints into tests:

• Is it permitted?
• Can it be completed in time?
• Does it preserve required safety or confidentiality?
• Does it use available resources?
• Does it satisfy the explicit objective?

Eliminate infeasible options before ranking the remaining ones.

Act, check, or escalate

Choose among three broad responses:

Act now

Appropriate when the hazard or deadline is immediate, the action is within authority, and delay creates greater harm. Prefer reversible, proportionate steps that stabilise the situation.

Gather information

Appropriate when a quick reliable check could materially change the decision and delay is tolerable. Name the information and source; “investigate further” is too vague.

Escalate

Appropriate when the issue exceeds authority, has serious consequences, requires specialist judgement, or creates a conflict that cannot be resolved locally. Escalation should include the known facts, actions already taken, uncertainty, and required decision.

These can be combined. A strong response may first protect people or preserve evidence, then notify the responsible person and obtain missing information.

Assumption testing

For each proposed answer, ask:

1. What must be true for this to work?
2. Is that premise stated?
3. If it is false, what harm follows?
4. Can the assumption be checked cheaply?

Prefer options that rely on fewer unsupported assumptions. Do not confuse caution with inaction: refusing every decision until all uncertainty disappears is often impossible.

Update when evidence changes

Good judgement is iterative. Form a provisional view, seek a discriminating fact, and update. Do not defend the first idea because time has already been invested in it. State confidence proportionately: “The record suggests X, but Y is not confirmed.”

In practice review, write the fact that would most strongly support your choice and the fact that would make you change it. This exposes hidden assumptions and trains flexible reasoning.