Did you know that “roll the car down the ramp and time it” can describe several quite different experiments? One student might start the car at the top and another halfway down; one may stop the clock when the wheels cross a line, another when the front bumper crosses. Both may believe they followed the same instruction.
A repeatable science method is a written procedure detailed enough for the same investigation to be carried out under specified, consistent conditions. The immediate solution is to state the materials, quantities, set-up, action, measurement start and end points, what stays constant, what changes and how results are recorded. Ask a partner to follow it without verbal hints: any necessary question exposes a gap worth fixing.
First diagnose what is missing, not merely whether the method sounds scientific
Hand a learner a one-line instruction: “Test how fast the toy car moves.” Ask five questions. Which car? What distance? What starts the car? Where does timing begin and finish? How many attempts? If the learner cannot answer from the written method, a second investigator must invent decisions. That makes a fair comparison harder.
The central distinction is between a result (what was observed) and a method (how the observation was produced). A method is not a story written after seeing the answer. It is an operational description that lets someone independently carry out the planned steps.
Worked example: write a method for a one-metre toy-car run
Illustrative classroom exercise, not actual pupil data: A group wants to measure the time taken by one toy car to travel along a straight 1.00-metre section of a smooth track after it is released from a fixed starting position on a ramp. The group uses the same car, ramp angle, surface and timing rule in three runs.
The weak instruction is: “Put the car on the ramp, release it and time it.” That instruction does not say where to position the car, what a metre means on the track, what counts as the finish, or whether a push is allowed.
A method another learner could attempt
- Mark a starting line and a finish line 1.00 m apart along the straight measured section. Record what part of the car will cross each line—for example, its front edge.
- Place the car with its front edge at the agreed release mark on the same ramp for each run. Keep the ramp and track in the same arrangement.
- Release the car without an extra push. Start timing when the front edge passes the start line and stop when it passes the finish line.
- Record each elapsed time in seconds using the same timing method and the precision that the instrument actually supports.
- Repeat the run three times, resetting to the agreed release mark each time. Note any collision, late start, or other irregularity rather than quietly deleting a reading.
Suppose the three recorded times are 2.4 s, 2.5 s and 2.3 s. Their sum is 7.2 s; the arithmetic mean is 7.2 ÷ 3 = 2.4 s. The result belongs to these invented readings and their specified procedure. It is not a claim about all toy cars, tracks or pupils. A hand-timed measurement is also affected by human reaction time, so close results do not prove the true travel time is known exactly.
A crucial refinement: a section of track after a ramp is not the same as the ramp itself. If pupils time from release, the distance and physics being measured differ from timing only a marked straight section. Both might be reasonable questions, but they must not be silently mixed.
What the written method must make visible
- Question: What outcome is being tested or measured? Name the dependent quantity and its unit.
- Set-up: Specify relevant dimensions, positions, instruments and material conditions.
- Action: Use ordered verbs that tell a second investigator what to do, not only what to observe.
- Boundaries: Define what counts as the start, finish and a valid run.
- Controls: Say what remains unchanged, and explain any deliberate change to one variable.
- Recording: Provide a results table that separates readings from comments about anomalies.
Do not turn a method into an impossible catalogue of everything in the room. Record details that could change how the investigation works or how the answer is interpreted. In a simple toy-car run, the colour of the classroom wall usually does not matter; the start position and timing rule do.
Diagnose three common misconceptions
“If we obtained the same number, our methods must have been the same”
Similar values can arise by coincidence or because two different errors cancel. Inspect the written procedures and conditions, not just the totals.
“Repeatable means perfectly accurate”
Repeated readings may agree while a stopwatch starts late every time. NIST’s terminology guidance distinguishes repeatability under specified same conditions from accuracy and from reproducibility under changed conditions. A classroom partner check is a practical test of clarity; it is not a formal proof of measurement validity.
“An unexpected result should be replaced with the expected one”
No. Preserve the original record and describe an observed problem. An irregular run may justify a clearly labelled additional measurement, but inventing a smoother value damages the evidence. This connects with eduKateSG’s guide to keeping predictions separate from measurements.
Progress from guided practice to independent writing
Guided: Give two students the weak instruction “measure how far a paper plane flies.” Together, identify the launching line, measuring endpoint, plane design, launch rule and number of trials. The adult models only the first two clarifying questions.
Supported: The learner writes an eight-step method for the paper-plane task. A partner circles every step that needs an unspoken decision. The writer revises those steps and explains which variable each revision controls.
Independent: Give a new scenario, such as counting the number of paper clips a magnetic object can lift under a specified set-up. Without referring to the first model, the learner drafts a method, creates a result table, names two plausible confounders and asks a peer to carry it out. Use ordinary classroom safety rules and teacher-approved materials.
What should a parent or teacher look for?
Progress is observable when a learner names the measured quantity and unit, specifies a stable set-up, defines start and stop rules, anticipates a reasonable source of variation and can explain why a peer’s question led to a revision. Keep the first and revised method side by side. That evidence is more informative than saying a student has become “more scientific.”
A teacher can ask pairs to exchange procedures but not explanations. A parent can use a simple household activity such as timing a paper ball travelling along a table, without conducting unsafe experiments. Afterward, ask “Which choice did you have to guess?” instead of supplying the next sentence.
Limits: clarity helps but does not remove uncertainty
Some natural systems change between trials, and some measurements are too sensitive for classroom tools. A clearly written procedure makes limitations visible; it cannot make the object behave identically. Nor does repetition make a poorly posed scientific question meaningful. When the investigation uses people, living specimens or hazardous materials, follow the school’s safety and ethics guidance before collecting any data.
Frequently asked questions
How long should a science method be?
Long enough to make all consequential choices explicit, but short enough to follow. A labelled drawing or simple table can carry details more clearly than another paragraph.
Should the results appear in the method?
Not as if they were instructions. Put measured values in a results section. A method may specify what readings to record and how the summary will be calculated.
What if two pupils still get different times?
Check whether they used the same operational rules, then assess ordinary measurement variation. Do not claim a faulty method solely because readings differ.
Continue learning
The next useful step is to compare the method with actual records. Read how to keep predictions and observations separate, then explore the wider eduKate Education Hub. The diagnosis-first approach also appears in the Secondary 1 Mathematics Clementi reference: explain the underlying rule, practise it and check independently. A strong science method gives another learner the same chance.