Quick Read
Curie is built to make invisible learning more visible.
But visibility is not the final goal. Education is.
Once the evidence is sufficient for a safe, narrow next step, continuing to measure can waste time, increase anxiety and create an illusion that perfect certainty is required before teaching can begin.
The Stopping Rule asks when we know enough to make the next justified educational move—and should stop collecting evidence for now.
The One-Sentence Answer
Stop measuring when the remaining uncertainty is unlikely to change the next safe educational action, then teach, observe the return, and reopen the question only if later evidence requires it.
Where the Stopping Rule Sits in Curie
The Blind Spot Test makes uncertainty visible. The Sampling Window asks whether the evidence is representative enough. Evidence Convergence asks whether different signals support the same interpretation.
The Stopping Rule asks whether any remaining uncertainty still matters for the next decision.
The Diagnostic Problem: Measurement Can Become Its Own Activity
If every uncertainty triggers another worksheet, another test and another observation, measurement begins to consume the time meant for learning.
A good system therefore needs not only rules for gathering evidence, but a rule for stopping.
Curie should reduce avoidable uncertainty, not pursue impossible certainty.
Enough Depends on the Decision
The amount of evidence needed is not fixed.
| Decision | Evidence need |
|---|---|
| Give one extra practice set | Low threshold; a plausible narrow weakness may be enough |
| Change the teaching route for a topic | Moderate; locate the likely weak layer |
| Declare a capability independently secure | Higher; include repeatability, support removal and suitable transfer |
| Make a broad statement about the learner | Very high—and usually avoid broad labels altogether |
A Useful Stopping Question
If the unresolved uncertainty went one way or the other, would we choose a different next educational action?
If the answer is no, more testing may have little immediate value.
If the answer is yes—especially if the alternatives imply very different teaching—then one Discriminating Test may be justified.
English
If a learner clearly understands the passage but consistently struggles to connect evidence to written inference, we may already know enough to teach that bridge. We do not need to prove every other English capability first.
For developmental English stages, continue through the English Tutor route.
Mathematics
If the learner selects the correct Mathematics route but repeatedly loses accuracy in algebraic execution, a narrow execution repair may begin immediately. More route-selection testing adds little if it would not change that next step.
For developmental Mathematics stages, continue through the Mathematics Tutor route.
Science
If a learner’s mechanism is accurate but scientific wording is imprecise, we may already have enough evidence to teach expression without reopening the whole conceptual diagnosis.
For developmental Science stages, continue through the Science Tutor route.
Additional Mathematics
If A-Math route selection and algebra are stable but verification repeatedly fails, the next repair can target checking. It is unnecessary to re-diagnose the entire topic before beginning that repair.
Stopping Is Provisional, Not Permanent
The Stopping Rule does not close the case forever.
Education is iterative. Teach the narrow repair, observe what returns, and reopen the diagnosis if the world disagrees.
This is why the Evidence Loop matters: stopping measurement now is compatible with learning more later.
The Cost of Over-Testing
- Teaching time is displaced.
- Learners may experience repeated evaluation instead of useful instruction.
- One capability can become over-measured while other needs wait.
- Repeated exposure can create a Measurement Effect and alter the very state being examined.
- Adults can mistake additional data for additional understanding.
Repair: A Practical Stopping Sequence
- State the next educational decision.
- List the minimum evidence that decision genuinely requires.
- Check the current Sampling Window.
- Identify any unresolved uncertainty that could change the action.
- If none would change it, stop testing and act.
- If one would change it, run the smallest useful discriminating observation.
- Teach.
- Observe the return and reopen only when new evidence warrants it.
Tutorial and Tutor
A Tutorial should spend most of its energy on useful learning, not perpetual diagnosis. Once the boundary is sufficiently located, the teaching event should move into repair, practice and transfer.
The Tutor dashboard can state: “Evidence is sufficient for the next repair. Remaining uncertainty does not currently change the action.”
What Parents Can Ask
- What decision are we trying to make?
- Would another test actually change what we do next?
- Is the uncertainty important or merely interesting?
- Can a safe narrow repair begin now?
- What later evidence will tell us whether the repair worked?
Development: Measurement Should Hand Responsibility Back to Learning
The developmental purpose of Curie is not to make the learner permanently dependent on external measurement. It is to make the learning state visible enough that the learner gradually understands what evidence to trust, what remains uncertain and when to act.
Voyage: Mature Learners Know When to Stop Checking and Begin
Across the Voyage, independence includes calibrated action under uncertainty. University, work and adulthood rarely provide perfect information before every decision.
The mature learner learns to gather enough evidence, act proportionately and remain correctable.
Darwin: Adaptation Cannot Wait for Perfect Certainty
The Darwin Series asks whether an old method still fits a changed environment. Adaptation often needs to begin before every unknown is resolved. The Stopping Rule protects that movement.
Boundaries
- Do not stop before evidence is sufficient for the actual decision.
- Do not demand perfect certainty for ordinary low-risk teaching choices.
- Do not use the stopping rule to ignore meaningful contradictory evidence.
- Do not collect more evidence merely because measurement is available.
- Keep every stopping decision provisional and open to later correction.
The Core Rule of the Stopping Rule
Gather only enough trustworthy evidence to justify the next proportionate educational action, stop measuring when unresolved uncertainty would not change that action, then return attention to teaching, learning and the evidence that naturally follows.
Frequently Asked Questions
Does stopping mean we are certain?
No. It means we know enough for the next decision while preserving uncertainty that does not yet need resolution.
What if new evidence appears later?
Reopen the question. Curie conclusions remain correctable by new evidence.
Why is this an instrumentation principle?
Because a measurement system without a stopping rule can begin serving measurement itself rather than the human learning purpose it was created to support.
Marie Curie Series · Blind Spot Test · Sampling Window · Discriminating Test · Measurement Effect · Evidence Loop · Tutor
eduKate Sengkang | Small groups of up to 3 | 1.5-hour tutorials | Properly Taught Kids Shine a Bright Light Into the Future.
eduKate Sengkang 