Did you know that a wonderful teaching example can be completely accurate without proving that the teaching method improves results? A worked solution shows how a method operates. A cheerful student story may show what happened to one person. Neither automatically tells us what usually happens to a group of learners.
To judge educational claims fairly, separate an illustration (a hypothetical or selected example that helps explain an idea) from outcome evidence (observations of what learners could actually do, measured under stated conditions). Then ask whether the evidence supports a descriptive claim about those learners, or a stronger claim that the method itself caused improvement. The two questions need different levels of support.
The Difference Matters to Families Looking for Good Help
Parents meet bold educational claims in advertisements, videos, worksheets and AI-generated lesson plans. A demonstration may be helpful: it can show how a teacher uses a model or diagnoses a misconception. But a demonstration is not an average, a guarantee or a controlled comparison. Equally, a single test score cannot explain by itself which part of instruction produced that outcome.
This distinction protects good teaching as much as it protects families. It allows a tutor to say, “Here is the method I use, and here is how I check whether your child is learning,” without pretending that one example proves a general result.
Worked Example: One Lesson, Then a Small Class Record
The following figures are an invented teaching scenario, not eduKate student data or a published trial. A teacher demonstrates how to explain why a wet shirt dries more quickly in moving air under otherwise comparable conditions. In the demonstration, one pupil sketches the moving air and correctly explains a plausible mechanism. That is a useful illustration of the teaching move. It does not establish that every pupil can now give an independent explanation.
Now imagine that thirty pupils take a short independent concept question before the teaching sequence: nine give an adequate explanation. After several lessons, twenty-one of the same thirty give an adequate explanation to a comparable but not identical question. That is 9/30 = 30% before and 21/30 = 70% after, a difference of 40 percentage points in the proportions answering adequately.
Those hypothetical figures describe an observed change for that class, provided the scoring standard was applied consistently. They still do not, by themselves, prove the lessons caused the change. Students may have studied elsewhere, the later question may differ in difficulty, and repeated exposure to the topic may contribute. A credible claim should state what was measured, who was measured and what uncertainties remain.
Four Different Claims—Four Different Evidence Needs
- “Here is how the method works.” A clearly labelled worked example can demonstrate its steps.
- “This learner improved on this task.” We need trustworthy before-and-after work or another clearly defined observation from that learner.
- “This group improved over the period.” We need consistent measures, counts, time windows and information about who participated.
- “The method caused the improvement.” We need stronger evidence that separates the method’s effect from alternative explanations, often through a carefully designed comparison.
Each sentence has a different burden of proof. Being fair does not mean demanding a randomised trial before trying any sensible homework routine. It means using the appropriate strength of language for the evidence actually available.
A Practical Claim-Check for Students and Parents
- Find the claim: is the writer showing a process, reporting an outcome or promising an effect?
- Classify the example: is it hypothetical, anonymised from a real case, selected as a success story or drawn from a systematically measured group?
- Inspect the measure: what could students do independently, and how was that judged?
- Check the denominator: how many learners were eligible and how many were included?
- Check the comparison: are the before and after tasks comparable? Was anything else likely to change?
- Match the conclusion: keep the claim as narrow as the evidence warrants.
A selected success story can still be inspiring. It should simply be labelled for what it is. A fictional example can still be excellent teaching. It should not be presented as a real student outcome.
How a Parent Can Test the Method for Their Own Child
Instead of asking only whether a tutor’s approach sounds impressive, request a clear starting diagnosis, an explanation of what will be taught and a plan for checking progress. For instance, ask the student to solve a familiar fractional problem, explain the method aloud and then solve a different problem without a model. Note the actual assistance needed. Repeat later when appropriate. That gives the family useful information about this child, even if it is not formal research.
In a classroom, teachers can distinguish a board demonstration from evidence collected through exit questions, independent transfer tasks and longer-term assessment. If an example failed, inspect the teaching, the prerequisite and the question instead of treating the child as a failure.
What Good Research Can and Cannot Tell Us
Evidence reviews such as the Education Endowment Foundation’s metacognition and self-regulation summary bring together multiple studies and warn that implementing an approach well matters. Research evidence helps educators choose promising strategies, but an average across studies is not a promise about every individual learner, school or tutor.
In everyday decisions, we can combine that broader research with direct observation of a student’s work. It is useful to recognise that classroom evidence, causal research and a parent’s practical progress record answer related but different questions.
Common Mistakes
- Calling a fictional illustration a case study of a real child.
- Treating one exceptional success as a typical expected outcome.
- Describing a before–after comparison as proof of cause.
- Quoting a percentage without its denominator or measurement method.
- Rejecting a helpful explanation merely because it is not a formal experiment.
- Confusing an impressive-looking worksheet with evidence of independent understanding.
A Progress Check That Actually Changes Behaviour
Ask an older learner to take three promotional or educational statements and label each as an illustration, a measured observation or a causal claim. Then have the learner specify one additional piece of evidence needed to support the stronger claim. A week later, repeat with different examples from school research, fitness or technology. Success is the learner’s ability to adjust certainty and ask the right evidence question unaided, not their ability to recite the labels.
For related reading, visit how to read test scores as skill evidence, the eduKate Education Hub and the Clementi teaching benchmark, which emphasises finding the learning gap, explaining the method and checking the next independent attempt.
Frequently Asked Questions
Is a parent testimonial worthless?
No. It may honestly describe that family’s experience. It should not be treated as a representative average or a guarantee for other children.
Does a before-and-after score prove a method worked?
It shows a change only if the measure is valid and comparable. Other causes may exist, so use cautious language about causation.
Can a tutor share invented teaching examples?
Yes, if clearly labelled as illustrative. Invented examples help explain processes; they must not be disguised as real student outcomes.
The Core Aim
Education should teach students not just to admire a compelling example but to ask what it actually establishes. A useful illustration makes learning visible. Good outcome evidence shows what changed. Careful judgement keeps those achievements separate so parents, teachers and students can make better decisions without false promises.