A tutor can know a great deal about how a learner is performing on educational tasks. That is not the same as knowing everything about the learner.
Good educational judgement becomes stronger when its boundaries are visible.
Make the claim no larger than the evidence—and leave a route for the evidence to change it.
A tutor can observe current performance
We can see whether a learner starts independently, retrieves a fact, represents a problem, selects evidence, explains a mechanism, uses a method accurately, checks, transfers or recovers after a failed route.
Repeated observations under relevant conditions can support increasingly confident statements about those current capabilities.
A tutor can identify patterns
If the same error recurs across comparable work, or performance changes reliably when support, timing or question surface changes, that pattern is educationally useful.
Patterns help us choose what to test or teach next.
A tutor can form hypotheses about mechanisms
A response trace may suggest weak retrieval, transfer, interpretation, representation or execution.
But the mechanism should remain a working hypothesis until the evidence distinguishes it from plausible alternatives.
This is why Why We Use Working Hypotheses Instead of Labels for Children is a central parent boundary.
A tutor cannot infer human worth from academic performance
A score, reading level, writing weakness or difficulty with Mathematics is not a measure of a child’s value.
Educational evidence should remain educational.
Ordinary tuition evidence does not authorise medical or psychological diagnosis
A tutor may notice a persistent difficulty and recommend that a family seek appropriately qualified advice. That is different from diagnosing a developmental, psychological, sensory, medical or other condition from tuition behaviour.
Professional boundaries are part of competence.
Future performance must remain uncertain
We can estimate readiness and identify current trends. We cannot know exactly what mark a learner will receive months later on an unseen examination under future conditions.
That uncertainty is one reason eduKate Does Not Promise Grades.
Confidence should rise with converging evidence
One task supports a small claim. Repeated independent performance across time, task variation and realistic conditions can support a stronger one.
When evidence conflicts, the correct response is not to hide the conflict. It is to identify which condition may explain it and what next observation would be useful.
Parents should expect revision
If a tutor’s first hypothesis is never allowed to be wrong, the system is not learning from the child.
Changing the interpretation after better evidence is not inconsistency. It is what evidence-based teaching should do.
A claim ladder: how large a conclusion does the evidence support?
Educational evidence becomes safer and more useful when the size of the claim rises gradually with the strength of the observation.
| Evidence | Reasonable claim | Claim that would be too large |
|---|---|---|
| One difficult question | “This task was not completed successfully.” | “The learner cannot do this topic.” |
| The same error across several comparable tasks | “This response pattern appears to be recurring.” | “This is a permanent weakness.” |
| Performance improves when one specific support is supplied | “That support may be compensating for a missing operation.” | “We now know the complete cause.” |
| Independent performance remains stable across time and variation | “This capability appears increasingly secure under these conditions.” | “Future examination performance is guaranteed.” |
| Evidence repeatedly conflicts across school, tuition and home | “The conditions may be changing the performance; we need a better comparison.” | “One observer must be wrong.” |
The consequence test
The more a conclusion could change opportunity, subject placement, workload or the way adults speak about the child, the stronger the evidence should be before that conclusion is treated as settled.
Educational observation should stay inside educational scope
A tutor can describe persistent patterns in academic tasks and can suggest that a family seek another appropriate perspective when the pattern falls outside ordinary teaching. The tutor should not inflate classroom or tuition observations into claims that require a different professional authority.
Uncertainty should still produce a next move
“We do not know yet” is useful when it is followed by a bounded question: compare timed with untimed work, remove one prompt, use a changed-context task, or gather one more comparable school sample. Uncertainty becomes unhelpful only when it prevents any sensible low-risk action.
The trust receipt
Parents should be able to tell which statements are direct observations, which are working interpretations and which remain unknown. That distinction protects the learner from overclaiming and makes later revision feel like responsible updating rather than contradiction.
What this page does not mean
Uncertainty is not an excuse for indecision. Low-stakes, reversible teaching moves can often begin before every explanation is settled.
The larger the consequence of being wrong, however, the stronger the evidence should be before an educational model is allowed to control opportunity.
Where to go next
For how evidence is interpreted more deeply, use the Marie Curie Series. For the practical question of whether the service should continue when the learner changes, continue to When Should Tuition Reduce, Change Direction or End?. For the full set of parent decision routes, return to the Parents’ Guide.
eduKate Sengkang · Good educational judgement becomes more trustworthy when it knows where its authority stops.
eduKate Sengkang 