eduKate Singapore

How to Categorise Assumptions | Source, Necessity, Evidence, Sensitivity, Scope and Risk

An assumption is something treated as true, fixed or acceptable so that reasoning, planning, modelling or action can proceed.

Every model, plan, forecast and decision contains assumptions. Some are explicit and tested. Others are hidden, inherited or merely convenient. Good categorisation makes those differences visible before they become failure points.

Quick answer: how should assumptions be categorised?

  • Source: evidence, convention, expert judgement, default, inherited rule?
  • Necessity: essential, simplifying, optional?
  • Evidence: strong, moderate, weak, absent?
  • Scope: local, system-wide, temporary, universal?
  • Sensitivity: how much does the conclusion change if the assumption changes?
  • Uncertainty: how doubtful is the assumption?
  • Dependency: what model, decision or task relies on it?
  • Reversibility: how costly is it to revise later?
  • Risk: what happens if it is wrong?

This article extends How to Categorise Models by treating assumptions as first-class knowledge objects rather than invisible background conditions.

1. Separate assumption from fact

A fact is supported as true within a stated context. An assumption is accepted provisionally so reasoning can proceed.

2. Separate assumption from hypothesis

A hypothesis is proposed for testing. An assumption may be accepted without being the main object of the test.

3. Separate assumption from constraint

A constraint limits possible action; an assumption describes what is believed or held fixed about the situation.

4. Explicit assumptions are easier to govern

Written assumptions can be reviewed, tested, versioned and challenged.

5. Hidden assumptions create surprise

Many failures occur because a background belief was never stated and therefore never tested.

6. Evidence-backed assumptions have support

Historical data, experiments or repeated observation may justify treating a condition as stable for a bounded purpose.

7. Convention-based assumptions depend on shared practice

Units, defaults and operating conventions can be reasonable without being natural laws.

8. Expert assumptions need provenance

Record who made the judgement, under which domain and on what basis.

9. Default assumptions reduce friction

Defaults help systems act without asking every question repeatedly, but they should be easy to override when context differs.

10. Simplifying assumptions reduce complexity

Models often hold some variables constant to make reasoning tractable.

11. Simplification is useful when omitted detail does not change the answer

The test is not whether an assumption is unrealistic, but whether it is harmless for the intended purpose.

12. Essential assumptions define the model

If changing the assumption creates a fundamentally different model or problem, it is structural rather than incidental.

13. Local assumptions apply narrowly

A classroom plan may assume one group size without implying the same assumption should govern every school.

14. Global assumptions affect the whole system

An assumption embedded in a core model can propagate into every dependent output.

15. Time-bounded assumptions need expiry

Prices, policies, demand and technology can remain stable only for a limited period.

16. Jurisdiction-bounded assumptions need location context

Legal and educational assumptions often fail when moved across countries or systems.

17. Sensitivity measures consequence

If a small assumption change produces a large output change, the assumption deserves strong scrutiny.

18. Low-sensitivity assumptions can be simpler

Where conclusions barely change, precision may not be worth the cost.

19. High-sensitivity assumptions require evidence or scenarios

Do not hide decisive uncertainty behind a single baseline value.

20. Assumptions can be correlated

Several assumptions may depend on the same hidden factor and fail together.

21. Assumption chains create dependency risk

An upstream assumption can silently influence many downstream decisions.

22. Contradictory assumptions cannot coexist silently

If one model assumes stable demand and another assumes rapid growth, the difference should be surfaced explicitly.

23. Assumption confidence should be separate from assumption importance

A highly important assumption can have low confidence; that combination is a major risk signal.

24. Assumption risk combines sensitivity and uncertainty

A doubtful assumption matters most when the system is also highly sensitive to it.

25. Reversible assumptions are cheap to revise

Early exploratory work can tolerate provisional assumptions when later correction is easy.

26. Irreversible commitments need stronger assumptions

Major capital, legal or safety decisions should not rest on weak untested assumptions.

27. Assumptions can be tested indirectly

Stress tests, sensitivity analysis and alternative scenarios can reveal whether a conclusion survives when assumptions move.

28. Assumption registers improve governance

Projects and models benefit from keeping a structured list of material assumptions, owners and review dates.

29. Assumption owners should monitor change

A named owner is more likely to notice when an assumption stops being true.

30. Assumptions can mature into evidence-backed facts

Repeated validation may strengthen status, though context and time limits still matter.

31. Assumptions can also become obsolete

External change can invalidate previously reasonable beliefs.

32. AI systems inherit assumptions from data and prompts

Training distributions, defaults and prompt framing can embed assumptions that are not visible in the final answer.

33. AI-generated assumptions should be surfaced

When a model fills missing context, inferred assumptions should remain distinguishable from user-provided facts.

34. A practical assumption record

  • assumption ID;
  • statement;
  • source;
  • scope;
  • necessity;
  • evidence;
  • confidence;
  • sensitivity;
  • dependencies;
  • owner;
  • valid time;
  • review date;
  • risk if false;
  • version.

35. Assumption categories should improve challenge

A useful scheme helps identify which assumptions can be ignored, which should be monitored and which must be tested immediately.

36. The deeper idea

Assumptions are the invisible scaffolding of reasoning. Classification makes the scaffolding inspectable.

To categorise an assumption well is to know why it is being accepted, how fragile it is, what depends on it and what breaks if it is wrong.

Final answer

Categorise assumptions by source, necessity, evidence, scope, sensitivity, uncertainty, dependency, reversibility and risk. Keep assumptions separate from facts, hypotheses and constraints, and monitor high-sensitivity assumptions because small errors there can create large downstream failures.

Continue through the series

Explore the connected learning guides

Choose the question that brought you here. Open one useful guide, try a small task, and stop when you have what you need.

Take one question further

The same learning habit can travel across subjects, while each subject keeps its own methods. These routes help you notice a difficulty, understand one part of it, and return to something you can do.

A word is familiar, but using it is difficult.

Move from recognising a word to retrieving it in a new context. Understand vocabulary plateaus.

Try it without the guide: Choose one word you already know. Close the guide and use it in a new sentence. Explain why it fits; try another context tomorrow.

A piece of writing has ideas, but the reader loses the thread.

Make the order of events and the links between sentences clear. Explore composition writing.

Try it without the guide: Choose one short paragraph. Read the relevant explanation, close it, and revise the paragraph. Ask someone to tell you what happened and why.

The Mathematics seems familiar, but marks still disappear.

Find the first point where the working stops being reliable. Find Secondary 4 A-Math mark leakage.

Try it without the guide: For a Secondary 4 A-Math question you have attempted, locate the first uncertain line. Repair that step, then try a comparable question without the worked answer.

A Science fact is remembered, but the explanation is incomplete.

Connect the evidence to a scientific idea and the resulting change. Follow the Primary Science learning route.

Try it without the guide: Choose a familiar Primary Science example. Explain the evidence, the idea and the result without notes. Then change one condition and explain your prediction.

Two accounts of the world seem to disagree.

Check the question, source, date and evidence before combining claims. Explore the World Knowledge research library.

Try it without the guide: Take one claim. Find the source best placed to support it, note its date, and state what remains uncertain. Return to your original question.

There is plenty of help, but independence is hard to see.

Check what the learner can understand and do after support is removed. Understand how education works.

Try it without the guide: Choose one small task the child has practised. Agree on a calm, brief attempt without prompts. Use what happens to choose one next step, then stop.

For the structure behind these connections, read the eduKateSingapore runtime manifest and the eduKate ecosystem boot contract. The reader map describes public navigation; those manifests preserve the wider ownership and return rules.