Quick Read
The moment a tutor gives a hint, the task changes.
The moment a learner is told that an answer is wrong, the next attempt happens in a different state. Repeated questioning can reveal understanding, but it can also guide the learner toward the answer. Even encouragement can change persistence, confidence and willingness to attempt.
The Measurement Effect asks whether the act of observing or testing has changed the capability we are trying to observe.
The One-Sentence Answer
Educational evidence is easier to interpret when we record what support, feedback and repeated exposure occurred before the learner produced the observed performance.
Where the Measurement Effect Sits in Curie
Prompt Dependence asks whether the learner can recreate an operation after support is removed.
The Measurement Effect asks an earlier integrity question: did the support change what we are now treating as evidence?
They are related but not identical. Prompt Dependence describes learner performance under support. Measurement Effect protects the interpretation of the observation itself.
Support Can Be Hidden Inside the Observation
| What the adult does | How the learner state may change | What should be recorded |
|---|---|---|
| Names the topic | Route selection becomes easier | Method context was supplied |
| Rereads a key phrase with emphasis | Attention is redirected | Meaning cue was present |
| Says “check your sign” | Error location is narrowed | Checking was prompted |
| Asks three leading questions | Explanation may be co-constructed | Reasoning was guided |
| Shows the correct answer first | Later explanation may contain hindsight | Outcome was known before explanation |
The Measurement Effect Is Not a Reason to Stop Teaching
Teaching is supposed to change the learner.
Hints, modelling, feedback and correction are useful educational tools. The problem begins only when evidence gathered after those interventions is described as though the interventions never happened.
Support is not contamination. Unrecorded support can become interpretive confusion.
Separate Before-Support and After-Support Evidence
A simple educational sequence can preserve both teaching and measurement:
- Allow an independent first attempt where appropriate.
- Observe what the learner does before intervention.
- Teach or support the boundary.
- Observe the supported performance.
- Return later after support has faded.
- Compare the three states rather than collapsing them into one.
This creates a much richer story: what the learner could do, what changed with help, and what survived afterwards.
English: A Leading Question Can Become Part of the Answer
If a tutor asks, “Do you think the writer is being sarcastic here?” and the learner then explains sarcasm well, the explanation still matters—but the route to the idea was cued.
A later uncued passage is needed before concluding that the learner independently recognises the same feature.
For developmental English stages, continue through the English Tutor route.
Mathematics: Naming the Method Changes the Task
“Use simultaneous equations” may be excellent teaching at one moment. But after that sentence, the learner is no longer demonstrating independent route selection. They are demonstrating whether they can execute the supplied route.
Both observations are useful. They simply answer different questions.
For developmental Mathematics stages, continue through the Mathematics Tutor route.
Science: Showing the Mechanism First Changes the Evidence
If the mechanism has just been drawn for the learner, a correct explanation may show that the learner can follow and express the model. It does not yet show that the learner can reconstruct the mechanism from evidence independently.
For developmental Science stages, continue through the Science Tutor route.
Additional Mathematics: One Hint Can Remove the Hardest Layer
In A-Math, one small hint—“differentiate first,” “substitute here,” “think identity”—can remove the method-selection problem while leaving the rest of the solution intact.
The tutor should therefore preserve the useful distinction: route was prompted; execution after the prompt was independent.
Feedback Creates a New State
After feedback, the learner is not the same learner facing the same task for the first time.
They now know something about correctness, location of error, expected direction or teacher intention.
This is why Self-Correction distinguishes an error the learner detects independently from an error located by the adult.
Repair: Preserve the Sequence of Support
- Observe before support where reasonable.
- Record the smallest support given.
- Notice what changed immediately.
- Do not claim independence from supported performance alone.
- Fade the support.
- Return under a fresh condition.
- Update the interpretation using the whole sequence.
Tutorial Has Two Jobs: Change Learning and Produce Receipts
A Tutorial is not a sterile measurement chamber. It is a place where teaching changes the learner.
The solution is not to separate teaching from evidence completely. It is to preserve the sequence well enough that later readers know what the learner did before, during and after support.
What Tutor Should Say
The Tutor dashboard should describe conditions honestly:
“You completed the algebra accurately after the method was named. We still need an uncued mixed-question check for independent route selection.”
Or:
“You first needed a cue to locate the evidence, but after the lesson you identified it independently in a new passage.”
What Parents Can Ask
- What could my child do before anyone helped?
- What exact hint was given?
- What did the hint unlock?
- Can the child now do the same operation without the hint?
- Was the answer known before the explanation was given?
Voyage: The Share of Adult Support Should Become More Visible as It Transfers Inward
Across the Voyage, more responsibility moves from adult to learner. That handover is easier to see when supported performance and independent performance are not blurred together.
Darwin: Support Must Adapt Without Hiding the New State
The Darwin Series asks when support should change because the learner’s environment or capability has changed. The Measurement Effect ensures we can still tell what the learner owns after the support evolves.
Boundaries
- Do not withhold useful teaching merely to preserve a clean observation.
- Do not treat every interaction as invalidating evidence.
- Do not shame learners for needing prompts; record the support instead.
- Do not describe supported success as independent success.
- Do not turn observation into surveillance.
The Core Rule of the Measurement Effect
Teach freely, but preserve enough of the before–support–after sequence to know which parts of the observed performance belonged to the learner, which were temporarily supplied, and what remained after the intervention faded.
Frequently Asked Questions
Does a hint make the result useless?
No. It changes what the result means. The performance now tells us what the learner can do with that support present.
Should tutors always wait before helping?
No. Preserve an independent attempt only when it is educationally appropriate. Teaching takes priority over creating artificial struggle.
What is the simplest protection?
State what support was present. “Independent,” “after one cue,” and “after modelling” are different evidence conditions.
Marie Curie Series · Prompt Dependence · Independent Start · Self-Correction · Tutorial · 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 