Did you know that a science table can show exactly the same number before and after an experiment even though the quantity actually changed? A scale that displays whole grams may show 150 g twice while the object has lost a fraction of a gram. The table is not necessarily wrong. The problem begins when we translate identical readings into the absolute claim that nothing happened.
A finding of no measured change means the recorded readings did not differ under a particular measurement method. It does not automatically prove that the underlying quantity was perfectly constant. To decide what the result supports, check the instrument’s resolution, accuracy, repeatability, conditions and the size of the change your experiment was designed to detect. Report the limitation instead of inventing an undetected value.
Why the instrument is part of the evidence
A measurement is not a direct window into an infinitely precise reality. It is a result produced by an instrument, a procedure and an observer. A kitchen scale, ruler or thermometer records within its design and the way it is used. A device with coarse steps may not distinguish two close quantities; a poorly calibrated device may also give inaccurate readings. These are different problems.
The US National Institute of Standards and Technology defines instrument resolution in terms of the ability to distinguish small changes. Its measurement-resolution guidance also cautions that the number of displayed digits alone does not prove the real instrument resolution. That distinction helps students avoid confusing a neat decimal display with trustworthy precision.
Worked example: a whole-gram scale and an invented drying experiment
Imagine a fictional classroom investigation that weighs a small wet cloth before and after a short, teacher-supervised drying interval. The scale is described as displaying mass in whole grams, and it shows 150 g at the beginning and 150 g at the end. We are not claiming these are actual pupil observations. They are deliberately invented numbers to practise measurement reasoning.
The safest descriptive statement is: ‘The scale displayed 150 g at both measurement times; it showed no change at its reported whole-gram display level.’ The unsafe statement is: ‘The cloth lost absolutely no water.’ The measurement does not justify that certainty. A small mass loss may fail to change the displayed value.
For example, if a hypothetical ideal scale rounds each perfectly measured value to the nearest gram, an initial mass of 150.2 g and a final mass of 149.8 g would both display 150 g. The real difference in this invented scenario is 150.2 − 149.8 = 0.4 g of mass loss. The observer could not recover that 0.4 g just by looking at the two whole-gram displays.
- Initial hypothetical mass: 150.2 g; displayed whole-gram reading: 150 g.
- Final hypothetical mass: 149.8 g; displayed whole-gram reading: 150 g.
- True mass difference in this constructed illustration: 0.4 g.
- Difference between the two displayed readings: 0 g.
- Defensible result from those displays alone: no detectable change at the device’s reported display level, not proof of exact physical equality.
The numbers 150.2 g and 149.8 g are explanatory possibilities, not data secretly measured by the instrument. A pupil must never insert those finer values into the results table as if they had been observed. The table should retain the two displayed 150 g readings, together with what is known about the instrument and setup.
What a rounding interval can tell us—and what it cannot
In an idealised model that rounds correctly to the nearest gram, a displayed 150 g could represent an underlying value from about 149.5 g up to, but ordinarily not including, 150.5 g, with endpoint details depending on the rounding rule. Two identical displayed values therefore do not determine the exact difference. The true readings in this model might be separated by nearly a gram while still rounding the same way.
However, this simple interval is not a validated statement of the real scale’s accuracy or measurement uncertainty. A scale can be uncalibrated, affected by its load, unstable or insufficiently repeatable. Rounding explains one possible source of missed change; the full uncertainty can be wider and must not be invented from the last displayed digit alone.
A diagnostic question that reveals the conceptual mistake
Ask the pupil: ‘If the display reads 150 g twice, which conclusion is supported?’ Offer three choices: (A) the cloth lost exactly zero mass, (B) the recorded readings were equal, or (C) the cloth definitely gained mass. Only B follows from the displayed data. Ask next whether B proves A; the correct answer is no. This is a language-and-evidence question before it becomes an arithmetic question.
Then change the context from mass to temperature or length. A digital thermometer that displays one decimal place may show the same value at two times. Can the learner explain why that does not prove an unchanging exact temperature? If they can transfer the reasoning without a remembered mass example, their understanding is becoming independent.
Resolution, accuracy and repeatability should not be collapsed into one word
- Resolution concerns how finely the device can discriminate changes under relevant conditions.
- Accuracy concerns how close a measurement is to an appropriate reference value, considering error.
- Repeatability concerns how similar readings are when measurements are repeated under specified similar conditions.
- A displayed reading can be stable but inaccurate; a finely displaying device can still have poor accuracy.
- For an honest school conclusion, record what is known about the device rather than assuming all three properties from the number of digits.
A scale with 0.01 g on its screen is not automatically better evidence than a well-used instrument with coarser digits. Instrument performance, calibration, environmental conditions and suitability to the task matter. This is why science uses documented measurement procedures rather than simply choosing the prettiest display.
How to redesign the experiment responsibly
Before repeating, estimate the approximate change the experiment is likely to create and ask whether the available method could detect a difference of that size. If the purpose is to study tiny mass changes, a whole-gram scale may not be appropriate. A supervised class could choose suitable school-approved equipment or change the planned observation duration if scientifically justified and safe.
Keep the method consistent. Use the same container, record whether the balance was zeroed correctly, avoid moving the setup between environments that differ, and record actual observation times. If a container or protective cover is used, weigh consistently and account for what belongs in the mass being compared. These choices can matter as much as the scale’s displayed increment.
Do not assume a longer experiment will necessarily produce a measurable change, and do not alter the procedure secretly after seeing disappointing results. When a revision is made, label the new run separately. A transparent record of the original limitation is itself scientific work.
The difference between ‘no evidence of change’ and ‘evidence of no change’
If the measurement method is too coarse for the question, equal readings give weak information about small changes. They might be compatible with many possible underlying outcomes. Strong evidence that a change is smaller than a practically important threshold requires a suitable design, sufficiently informative measurements and an appropriate analysis. A beginner should not use the phrase ‘proved no effect’ merely because two cells show the same number.
The same principle appears in other subjects. A pupil who does not hear a very quiet sound cannot prove that there was no sound; a camera photograph taken from too far away may not show a small detail that exists. Those analogies help with the reasoning, but they do not replace the science of each measuring method.
Progressive practice: move from visible differences to hidden ones
- Stage 1: Use a ruler with clear millimetre marks and compare two objects that visibly differ by several millimetres.
- Stage 2: Discuss two lengths that appear equal on a coarse ruler; ask what the ruler can and cannot establish.
- Stage 3: Use invented before-and-after readings rounded to whole grams and identify a possible small hidden change.
- Stage 4: Have the learner write a precise one-sentence result that does not invent extra decimals.
- Stage 5: Present a new instrument and ask what information about resolution, repeatability and conditions would be needed.
- Stage 6: Ask for a better, safe measurement plan and let the learner choose and defend it independently.
These tasks need not require buying specialised devices at home. Fictional data and a familiar classroom ruler are enough to practise honest inference. More advanced investigations should be supervised by teachers with suitable equipment and safety controls.
Independent challenge with worked reasoning
Suppose a fictional water container reads 21.0 °C before and 21.0 °C after five minutes on a digital display. Another instrument, known only to show one decimal place, is unavailable for comparison. Can a pupil conclude the water remained at exactly 21.000… °C? No. We can say only that the recorded displays were the same, while recognising possible changes below effective detection or other measurement limitations. The number of decimal places does not establish the complete uncertainty.
Now suppose a different experiment records 21.0 °C and 22.4 °C with the same device and measurement method. The displayed difference is 1.4 °C. It may indicate warming, but interpreting the cause still requires a fair design, trustworthy measurements and the experiment’s context. Detecting a numerical difference and explaining why it occurred are distinct questions.
How parents and teachers can notice progress
A tutor can ask pupils to circle words that overclaim: exactly, proved, impossible, always. Then they can revise a sentence to match the available readings. Parents can ask, ‘What is the smallest change your equipment can reliably show?’ If that is unknown, the responsible answer is ‘We need to check.’ Children should not be rewarded for inventing a precision the instrument never supplied.
Observe whether a learner preserves raw readings, distinguishes a display from an underlying quantity, and proposes a fairer measurement procedure in a fresh example. Those independent behaviours are stronger evidence of learning than a polished table with many decimal places.
Important limits
Different instruments measure different physical quantities and have distinct sources of error. Measurement resolution is not the same as an analytical detection limit, and neither alone captures all uncertainty. A simple school demonstration cannot justify a full professional uncertainty budget; this article introduces the logic rather than replacing instrument manuals or scientific supervision.
School safety matters. Do not use hazardous heaters, chemicals or unsecured electrical equipment merely to make a change larger. When a test cannot measure its intended effect safely with the available tools, redesign the task or report the limitation honestly.
Frequently asked questions
Do identical displayed numbers prove nothing changed?
No. They show that the recorded displays were equal; a small physical change may not have been resolved.
Can I write estimated hidden decimal places in the results table?
Not as measurements. Use such numbers only when clearly labelled as illustrative possibilities, never as observed data.
Does a digital instrument with more digits guarantee a better experiment?
No. Check resolution, accuracy, repeatability and whether the instrument suits the expected change.
Should we discard an experiment that shows no measured change?
Not automatically. Preserve the readings, evaluate whether the method could detect the effect of interest, and explain limitations.
What is a useful next step for a learner?
Give an unfamiliar result table and ask the child to identify the claim it supports, the claim it does not support and one safer improvement.
Continue learning
Our Core Aim discussion of missing observation times develops the same commitment to truthful records. For related methods, read practical accuracy, precision and experimental errors. Explore the eduKate Education Hub and the Clementi Mathematics teaching reference for explanation-first, diagnosis-first, independent checking.
The learning goal is neither to doubt every number nor to believe every displayed digit. It is to say exactly what the measurement can support—and to recognise what remains too small, too uncertain or too poorly measured to claim.