Did you know that a neat-looking science table can hide exactly how the measurements were obtained? A student might write “average flight distance: 227 cm” but leave no record of the original three flights. That summary is useful, yet another person cannot check it or see whether one unusual result changed the answer.
Raw measurements are the values directly observed or read from an instrument, with their units and relevant conditions. Calculated results are new quantities derived from those observations, such as a difference, average, ratio or rate. The immediate rule is to preserve the original readings unchanged, label calculated columns clearly and show the operation that connects them. A calculated number is not an extra independent observation.
Quick answer: record, derive, then interpret
- Define what is actually measured: distance in centimetres, time in seconds, mass in grams or another quantity.
- Record the instrument reading for every trial, including its unit and any stated resolution; do not silently replace it with a rounded group average.
- Keep calculated results in separate named columns or a separate summary table.
- Write the formula or operation, use consistent units, and check arithmetic on the original values.
- Distinguish a missing measurement from a measured zero. Missing is not the same as zero.
- Discuss patterns and limitations only after the recorded readings and derived quantities are checked.
Why this distinction matters in school science
Science is not simply the production of a plausible final number. It is a documented path from a question and a method to evidence and a conclusion. When students merge raw and calculated values, three avoidable problems arise. First, a teacher cannot independently check the arithmetic. Second, an unusual observation may disappear inside an average. Third, later corrections become risky because nobody can tell which entries were observations and which were derived.
Raw does not necessarily mean perfect. A ruler may be read at a slight angle; a stopwatch may be started late. Keeping the recorded value preserves what happened during the activity, not a guarantee that the instrument captured the exact underlying quantity. A transparent table helps us discuss measurement uncertainty without pretending that every reading is equally reliable.
The US National Institute of Standards and Technology’s explanation of measurement uncertainty is a useful primary reference: measurement is a process, and a reported value does not erase uncertainty. For a school worksheet, students usually need the simpler operational lesson first: never pass a calculated figure off as a separately measured one.
A complete hand-checked example: paper-airplane flights
Original fictional classroom activity: Three flights of the same paper airplane are made in an indoor area under similar conditions. A metre tape marked in centimetres is used to record the horizontal distance of each flight from a fixed start line to its landing point. For this teaching example, the recorded readings are 215 cm, 240 cm and 225 cm, each to the nearest centimetre. These are invented numbers, not results from real pupils.
Write the raw table with one row per trial: Trial 1: 215 cm; Trial 2: 240 cm; Trial 3: 225 cm. The labels and units belong in the heading or beside every reading. Do not write “average 227 cm” in the same field used for Trial 3, because an average is not a fourth flight.
Calculated total: 215 cm + 240 cm + 225 cm = 680 cm. Calculated mean: 680 cm ÷ 3 flights = 226.666… cm per flight. Reported to the nearest whole centimetre for this exercise, the mean distance is 227 cm. The unrounded value is retained in the calculation; reporting a rounded mean does not alter the three original readings.
Calculated range: the largest measured distance minus the smallest measured distance is 240 cm − 215 cm = 25 cm. A range describes the spread of the measurements; it is not a fourth flight distance and not evidence that the flights followed a particular physical distribution.
Check two ways. Arithmetic: 215 + 240 + 225 = 680, and 680 ÷ 3 ≈ 226.7. Reasonableness: a mean of about 227 cm lies between the smallest 215 cm and largest 240 cm; it would be suspicious if the average were 280 cm. The range cannot be negative if calculated as maximum minus minimum. This check catches many copying and arithmetic errors.
What a well-labelled results page looks like
Put the measured data first: “Trial number” and “Horizontal flight distance (cm)”. Then write a clearly separated “Calculated summary” section containing “Mean flight distance (cm): 227” and “Range (cm): 25”, alongside the unrounded arithmetic that led to the displayed average.
If a school task requires a single combined table, use distinct headings such as “Measured distance (cm)” and “Calculated speed (cm/s)” rather than leaving both under “Results”. Each row should represent a well-defined trial. Include a note describing the method and any measurement resolution actually used; do not invent instrument accuracy after the fact.
A different example: measured starting and ending volumes
Suppose a separate fictional activity records liquid volume as 75 mL before a procedure and 68 mL after it. The two raw readings are 75 mL and 68 mL. The change, computed as 75 − 68 = 7 mL less, is derived. Recording “7 mL” as though it were a third volume measured by a cylinder misrepresents the evidence.
The calculation describes an observed decrease in recorded volume. It does not by itself prove that 7 mL evaporated; spillage, transfer losses, retained liquid or measurement differences might also explain some of the change. This is the boundary between a calculated result and a scientific causal claim.
A learner diagnostic: four entries, four labels
Ask the student to label each entry as a direct measurement, a calculation or an interpretation: “240 cm read from the tape”; “227 cm average across three flights”; “25 cm difference between longest and shortest”; “The airplane is always more stable when launched gently.” The first is a raw reading; the next two are calculated summaries; the last is an interpretation or broader claim that the three distances alone do not establish.
If a learner calls the mean “raw”, ask which instrument directly displayed that mean in the trial. If no instrument did, it was obtained by calculation. If a learner calls an unsupported claim “data”, ask what precise recorded measurement verifies it. The aim is to build an evidence trail, not to penalise vocabulary for its own sake.
Common mistakes and why they happen
Replacing the original measurement with the mean
This makes the results appear tidier but destroys information. Preserve each reading even if it differs from the expected pattern. If an entry is known to be a transcription error, document the correction against the original record instead of secretly overwriting data.
Rounding every reading again during calculation
If readings are already recorded as whole centimetres, do not needlessly round them further to the nearest ten before computing an average. Keep sufficient detail through the calculation and state the reporting rule used for the final result. Do not present a value with more decimal places as proof of more accurate measurement.
Treating a missing trial as zero
If Trial 2 was not measured, “missing” is not 0 cm. Including a fabricated zero would wrongly lower the mean. Record the absence, explain why it occurred if known, and calculate only from observations appropriate to the question. See the linked discussion about missing observations before using incomplete tables.
Claiming the average proves repeatability or success
One average does not show the spread, the number of trials or whether conditions changed. A report should include the underlying results and, where useful, a basic measure of variation. School science explanations should not imitate laboratory certainty they have not earned.
Practice progression: from guided to fully independent
Level 1: identify raw and derived
A pupil times three puzzle attempts: 32 s, 28 s and 36 s. Those numbers are raw stopwatch readings for this invented exercise. The total 96 s and mean 32 s are derived. Ask: if another student sees only “mean = 32 s”, what information about variation has been lost? The original values reveal an 8-second range.
Level 2: calculate and validate
Measured plant heights in a fictional exercise are 14 cm, 15 cm and 16 cm. The sum is 45 cm, the mean is 15 cm, and the range is 2 cm. Check that 15 lies between 14 and 16. The data might be three plants rather than repeated readings of the same plant; the question must make that distinction explicit before interpretations are offered.
Level 3: correct a mixed-up results page
A student lists “Trial 1 = 4.2 s; Trial 2 = 4.4 s; Trial 3 = 4.0 s; Trial 4 = 4.2 s (average)”. Repair this by keeping exactly three trial rows and a separate mean of (4.2 + 4.4 + 4.0) ÷ 3 = 12.6 ÷ 3 = 4.2 s. The number is correct, but its original label falsely suggests a fourth observation. This exercise separates numerical accuracy from truthful recording.
Level 4: plan a new investigation
Let a learner design a safe activity such as timing how long it takes to sort ten cards on three attempts. Ask them to decide which numbers the timer or observer will record directly, which values require arithmetic, and what conditions must be similar for the comparison to be meaningful. The teacher should review the method and safety before beginning.
Parents and teachers: useful next steps
For parents: Instead of asking “Is your answer correct?”, ask “Which numbers did you actually measure, and which did you calculate?” Let the child show one calculation from the recorded values. This is more revealing than reciting a final average. Avoid replacing an inconvenient observation with a more pleasing number.
For teachers: Give learners two versions of the same report, one with only a mean and another with the underlying trials. Ask which one can be audited. Then introduce a purposeful copying error and see whether students can detect it using the original record. Where an experiment requires more advanced uncertainty work, teach it explicitly instead of expecting pupils to infer it from the word “accurate”.
For students: Keep dates, trial IDs, units and method notes with your results, and show the formula whenever you derive a new value. Ask whether the table alone would allow a classmate to reproduce the calculation. If it would not, add the missing information before writing your conclusion.
What counts as observable progress?
On a new investigation, check whether the learner can (1) label all direct observations correctly, (2) show a calculation that can be checked from them, (3) report units honestly, and (4) avoid claiming that an average or difference is a direct measurement. A later unseen task tests transfer better than repeating the identical airplane values.
The limit matters. Preserving raw measurements does not prove that a science experiment is well designed, that the equipment was calibrated, or that the pattern is causal. Those questions need separate checks of method, controls and uncertainty. Keeping an inspectable record is the necessary foundation for those checks, not their replacement.
Frequently asked questions
Can a calculation be called a result?
Yes. A calculated mean or change is a result when it is clearly identified as derived. The problem is presenting it as though an instrument directly produced the number.
What if the instrument reports an average automatically?
Describe the instrument output accurately, including what samples went into it if known. An automatically computed average is still a derived value even when the device displays it.
Should every raw value have decimal places?
No. Record readings consistent with the measurement method and instrument. Writing extra decimals does not improve the underlying measurement.
Do we discard an unusual reading?
Not simply because it looks inconvenient. Inspect the method, check for a documented procedural or transcription problem and follow any justified exclusion rule transparently. Otherwise preserve the value and discuss its possible implications.
Related reading and a stronger habit
Continue with How to Record a Science Method Change and How to Use a Blank Sample to strengthen experimental integrity. The immutable Mathematics teaching reference models diagnosis, stepwise explanation and independent checking. Explore the Core Aim learning hub and the Education Hub for broader learning guides.
A trustworthy science report lets another reader walk backwards from its conclusion to the calculations and then to the actual measurements. Save the evidence, show the derivation and be clear about what the numbers can—and cannot—say.