A Secondary 2 student in Clementi times a toy car rolling down a ramp. One classmate releases the car, another starts the stopwatch, and a third calls out when the car reaches a strip of tape. The class records 2.31 seconds. They repeat the trial and obtain 2.18 seconds, then 2.42 seconds. The car, ramp and start position appear unchanged. Why do their times vary? It is tempting to blame the car. Part of the difference may be in the people pressing the buttons.
Did you know a digital stopwatch can display hundredths of a second without making a human-operated experiment accurate to hundredths of a second? The display’s resolution is not the same as the measurement’s accuracy or uncertainty. Manual timing records more than the event. It can include the observer’s detection of a cue, decision to press a button, hand movement, device behaviour and agreement about what counts as a start or finish. In a very short investigation, those human contributions can be large compared with the duration being measured.
For students studying Science experiments, practical skills, timing measurements and fair tests, this is not a small technical detail. It is a lesson in what a measured number means. We will examine the two timing endpoints, the direction of delays, repeated trials, reliable averages, automated alternatives, reaction-time checks and the difference between a fair setup and a sound measurement. Parents and teachers can use the examples to diagnose why careful children sometimes report inconsistent practical results.
The 50-second answer
In a manually timed experiment, the measured interval depends on when the start button is pressed and when the stop button is pressed. If a human presses Start later than the actual beginning, the recorded time tends to be too short. If the same person presses Stop later than the actual end, the recorded time tends to be too long. These two delays can partly cancel, reinforce or vary from one trial to another. A stopwatch that reads to 0.01 seconds does not eliminate them.
The practical fix is to define the event clearly, use consistent cues, repeat appropriate trials, report sensible precision and, when very short intervals matter, use a suitable automated timer or video-based method if available. Repetition can help estimate variability but cannot automatically remove a consistent timing bias. A number on a screen is evidence only after you understand the method that produced it.
1. The event has a time; the observer has a response
There are at least two timelines in the ramp experiment. In the first, the car physically crosses a starting marker, travels along the ramp and crosses the finish marker. Those are the events we want to measure. In the second, a pupil sees or hears a cue, decides that the cue has occurred, moves a finger and presses a stopwatch button. The stopwatch records the pupil’s button presses, not the car’s border crossings directly.
A diagram drawn as two rows makes the difference visible. Write the physical times on top: actual start at time zero, actual finish at time 2.00 seconds. Beneath that, write hypothetical button presses: Start at 0.18 seconds, Stop at 2.30 seconds. The displayed interval is 2.30 − 0.18 = 2.12 seconds. The car still travelled for 2.00 seconds. The extra 0.12 seconds came from the difference between two response delays.
Those values are chosen for illustration, not presented as typical human reaction times. Response delay depends on the cue, anticipation, individual, task, familiarity, device and other conditions. It would be misleading to tell every class to subtract a universal reaction-time correction such as 0.2 seconds from every measured result. The correct mathematical principle is that the difference between endpoint delays matters.
This is a first-principles idea worth teaching explicitly. If a student says “the teacher pressed late, so the time must be longer,” ask “At which end?” Pressing late at the start and pressing late at the finish have opposite effects on the recorded interval. The direction of the error cannot be determined until we know the timing sequence.
2. A useful equation that does not need advanced physics
Let the actual start be S and the actual finish be F. The true interval is F − S. Suppose the observer presses Start after a delay a, so the start button time is S + a. The observer presses Stop after a delay b, making the stop button time F + b. The displayed interval is (F + b) − (S + a), or the true interval plus b − a.
That little equation explains four common situations. If a and b happen to be equal, the delays cancel in the difference, although that does not mean either button press was simultaneous with the event. If b is larger, the time appears longer than reality. If a is larger, it appears shorter. If the delays change between trials, the measured duration fluctuates even when the car’s actual journey does not.
The model is deliberately simple. It ignores clock calibration, display resolution, internal device latency, human anticipation before a cue, and ambiguity about a car touching a line. It is a helpful starting model, not a full uncertainty budget. Students should use it to understand the sign of errors rather than to pretend that every real delay has been measured exactly.
One teaching strategy is to swap the hypothetical values. Take a true two-second event. Let start delay be 0.30 seconds and stop delay 0.10 seconds. The measured value is 1.80 seconds. Now reverse the delays. It becomes 2.20 seconds. Nothing about the ramp changed. Only the timing relationship did. This is often the moment when pupils begin to see why a seemingly precise display is not the whole answer.
3. Resolution, precision, accuracy and uncertainty are not synonyms
The stopwatch says 2.31 seconds. Its smallest displayed step appears to be 0.01 seconds. That is a statement about display resolution. It does not show that the event’s true duration lies within 0.01 seconds of the reading. A human may have started or stopped the timer tens or hundreds of milliseconds after observing the event, depending on the circumstances. A device also may have its own timing uncertainty.
Precision, when used in measurement, concerns how closely repeated results agree under specified conditions. A class might record five times clustered around 2.30 seconds. That repeatability is useful information. Accuracy concerns closeness to the correct value, which may be unknown. A consistently late stop cue could give impressively repeatable but biased times. Uncertainty describes the range or distribution of plausible measurement values under a declared model and procedure; it should not be confused with a guarantee that the true value sits inside an arbitrary interval.
A classroom example makes the distinctions practical. Group A records 2.11, 2.12 and 2.12 seconds. Group B records 1.98, 2.09 and 2.23 seconds. Group A is more consistent. It is not automatically more accurate, because we have not established a trusted reference time. Perhaps Group A shares the same stopping habit on every trial. Perhaps Group B is observing a car whose release speed genuinely differs. Without further evidence, comparing spread alone cannot settle truth.
This precision vocabulary is useful beyond stopwatch work. It applies to measuring temperature, mass, length and electrical quantities. The words become clearer when anchored to an actual method instead of memorised as disconnected exam definitions.
4. Starting a stopwatch on a shouted word adds a hidden event
Imagine a pupil releases the toy car and says “Go!” at nearly the same moment. The timer operator presses Start when hearing the word. Which moment is the start? The physical release, the beginning of the spoken sound, the arrival of that sound at the listener’s ear, or the button press? For a long event, tiny differences may not matter much. For a very short one, they can matter enough to complicate interpretation.
A useful improvement is to choose one operational start cue. If the scientific question is the travel time from one fixed line to another, define the physical crossing as the start. If a manual operator cannot reliably perceive both crossings, record that limitation and consider a different method. If the activity is about classroom coordination rather than precise dynamics, a spoken cue may be adequate; the purpose of measurement determines the required accuracy.
The most dangerous solution is a hidden change in method between trials. On the first attempt, the observer begins timing at “Go.” On the second, they press when the car visibly moves. On the third, they anticipate the release. The data now combine different definitions of the interval. Repeating a procedure only helps when the repeated procedure is meaningfully the same.
Teachers can ask a pair of students to write a start rule before performing the experiment. “Press Start when the front of the car crosses the first line” is clearer than “start when it goes.” For some setups, the clearer rule may still be hard to execute; clarity of definition is necessary but not always sufficient for accurate measurement.
5. The finish line also needs a definition
A toy car is not a point. It has a front wheel, front bumper, centre, rear wheel and shadow. A student may stop the timer when the front bumper reaches the finish line. Another may wait for the back wheel. Both can sincerely say, “I stopped when it passed the line,” yet they measured different intervals. A ball falling through a frame raises a similar question: first contact, centre crossing or complete passage?
This is not a reason to demand unreasonable precision from children. It is a reason to write a simple, reproducible condition: stop when the front edge reaches the finish mark. Draw the cue on the worksheet. For an experiment with two observers, rehearse which feature is being watched. If the line is thick, decide whether the near edge, centre or far edge counts. In ordinary classroom tasks, this degree of clarity is often enough to improve consistency markedly.
Be careful, however, about attributing every variation to reaction time. If different students choose different finish events, the underlying definition has changed. That is a method error, not simply a late button press. The repair is to standardise the event rather than adding or subtracting an imagined reaction delay.
The distinction is especially important when interpreting averages. Averaging a mix of different finish definitions does not recover a single well-defined physical interval. The average could look smooth while no single child actually measured the quantity stated in the question.
6. Why short events magnify the importance of human timing
Suppose the actual duration of one motion is about two seconds, and manual endpoint errors amount to a difference of 0.10 seconds on a trial. That is around five percent of the event. For a twenty-second event, the same absolute difference is around 0.5 percent. For a fraction-of-a-second event, it may be enormous. The exact human timing error is not constant across those situations, but the ratio illustrates why short durations are difficult to measure manually.
Students may think a faster car is automatically easier to time because the event ends quickly and they can concentrate. In fact, the task can become harder because there is less opportunity to observe the start and finish separately and because endpoint uncertainty occupies a larger share of the interval. A teacher can change the ramp distance, use slower motion or choose a suitable timing apparatus, provided the change still answers the intended scientific question.
There is a trade-off. Slowing an experiment can alter the phenomenon being studied. If the question concerns the car’s speed on a particular ramp, reducing the incline to obtain longer times produces a different condition. The remedy cannot be to change the experiment arbitrarily until the stopwatch behaves. Better measurement design preserves the scientific target while improving observation.
This points to a broader principle of experimental design: choose a measurement method that is suitable for the scale of the quantity. A kitchen timer may work for an activity lasting minutes; a hand-operated stopwatch is poorly suited to measuring the duration of a single finger snap. Matching tool to time scale is a better habit than celebrating an extra decimal place.
7. A fair test does not guarantee a reliable timer reading
Students are often taught to keep conditions the same: use the same ramp, car, distance and surface. That is good design for comparing a chosen variable. But a fair physical setup can still produce unreliable data if the measuring procedure changes. Consistent materials are not a substitute for an operational definition or an appropriate instrument.
Imagine an investigation asking whether increasing ramp height changes travel time. The group carefully keeps the car mass and travel distance constant. They use the same stopwatch but change timer operators halfway through. The recorded difference across heights may now reflect both the changed ramp and the changed timing response. Even if the operator stays the same, different anticipation at different heights can affect the result.
A useful research question is: “Could something else have changed at the same time as our intended variable?” That includes not only obvious factors such as surface friction but also observation conditions, method, instrument and definition. Scientific control is a system that connects the experiment’s physical design to the evidence used to draw a conclusion.
For classroom assessment, this is an opportunity to move beyond the phrase “fair test.” Ask learners to justify how the timing method measures the variable and what would make two trials comparable. A student who can name the stopwatch yet cannot define the start and stop has identified an instrument, not yet a complete method.
8. Repeated readings: what they can rescue and what they cannot
Repeating trials is useful because the class can see whether the results are stable. If five times are 2.10, 2.18, 2.15, 2.23 and 2.14 seconds, the average is 2.16 seconds. The spread itself is information. A teacher may ask learners to report the range, consider a graph or decide whether the differences are large enough to change the comparison being made.
If trial-to-trial timing errors are mostly variable and not systematically biased, averaging multiple observations can reduce the variability of the mean. This is a statistical principle with assumptions; it is not a magic property of the word “repeat.” If the timer operator always stops late for the same reason, twenty repeats may yield a very stable wrong result. If the release point drifts, the class may be averaging different physical experiments.
Students sometimes remove any result that seems inconvenient. A much longer trial is labelled an “outlier” and deleted because it spoils the average. That can hide genuine evidence. Was the car caught on the ramp? Did the timer start twice? Was the stop signal missed? If a procedural mistake is documented, exclusion may be justified under a stated rule. If the result is merely unusual, it needs investigation, not automatic erasure.
A good school table therefore records more than times. Include trial number, operator, any obvious anomaly and a concise method note. The goal is not to create laboratory bureaucracy but to leave enough information for someone else to understand what the numbers mean.
9. Why changing the person can change the number
Two timer operators can respond differently to the same visual event. They may differ in attention, familiarity, anticipation and button technique. That does not make either person careless. Human response is part of the instrument chain when a measurement is manually triggered. NIST’s stopwatch calibration guidance explicitly considers measurement methods and uncertainty; its technical treatment of human reaction-time contributions is a useful reminder that even professional timing work has to account for operators.
For a classroom comparison, keeping one operator across conditions can remove one source of inconsistency, although it does not automatically remove the operator’s bias. An even stronger design might compare the operator’s results with an automated timing method where appropriate. If groups must share a single stopwatch, record who operated it and do not mix results casually.
A simple exercise gives everyone a turn on the same event. For example, display an unpredictable visual signal and ask each observer to respond with a button. The measured response delays will vary. Such an activity is not a direct calibration of the start and stop biases in the ramp experiment; the cue and motor actions differ. Its value is to make the general issue visible and to challenge the assumption that a person reacts at exactly zero delay.
Do not use this activity to rank students by intelligence, attentiveness or personal worth. Reaction times vary with many factors, and one classroom exercise is a poor basis for judging a person. The educational aim is better evidence reasoning, not a competition over reflexes.
10. Do smartphone timers solve the problem?
A phone may display hundredths or thousandths of a second, but tapping its screen is still a human action. The device’s user interface, app behaviour and screen updates add their own characteristics. Extra digits do not establish extra accuracy. A phone recording a video, however, may support a different approach: review the same footage and identify the frames in which a moving object crosses the starting and finishing markers.
Video is not automatically perfect. Frame rate, motion blur, camera viewpoint, rolling shutter, frame timestamps and ambiguity about exact crossing all matter. At 30 frames per second, the nominal interval between consecutive frames is about 1/30 second, or 0.033 seconds, but real recording details and event placement between frames determine the actual uncertainty. It would be unsound to claim that video always times an event to exactly 0.033 seconds.
For a classroom experiment, video review has a conceptual advantage: the start and finish can be examined repeatedly with the same visible definition. Two pupils can discuss why they chose a particular frame. If the footage clearly shows the movement, this may reduce some endpoint-response errors associated with real-time button presses. If the video is unclear, poorly lit or viewed at the wrong angle, it can simply trade one problem for another.
The decision should match the learning objective. If the lesson is about reaction time itself, manual timing may be central. If the lesson is about the car’s physical motion, a more direct timing method may support the intended inquiry.
11. A simple beam sensor—and a warning about new sources of error
A light gate or other suitable motion sensor can trigger a timer when an object breaks a beam. This reduces dependence on a pupil noticing the crossing and pressing a button, provided the sensor is correctly set up. It can be especially helpful for events too short for reliable hand timing. But automated systems still require clear start and finish definitions, appropriate calibration, sufficient resolution and correct placement.
If one gate is 20 cm higher than the intended start mark, the automation has reliably timed the wrong interval. If the object’s front edge is used at one gate and a different part at another, geometry matters. If a sensor misses thin objects or responds differently to reflected light, the result may be incomplete. Technology removes some human error channels while introducing technical ones.
Students need not master sensor electronics to appreciate the difference. Ask: “What exact event causes the timer to start?” and “What exact event causes it to stop?” The answers should describe something observable and reproducible. “The machine knows” is no more satisfactory than “The stopwatch knows.”
This view protects students from gadget worship. It encourages evidence-based choice of apparatus. A cheap manually timed method can be adequate for a long simple event. A more elaborate electronic device is justified when the desired accuracy, event duration and available resources call for it. The best instrument is not always the one with the fanciest screen.
12. The reporting trap: 2.31 seconds can imply too much
A pupil records three manual timings: 2.31, 2.42 and 2.18 seconds. The calculator reports an average of 2.303333333… seconds. Should the student report 2.303333333 seconds? No. The arithmetic can produce many digits even when the measurements do not justify them. The displayed mean’s detail must be interpreted against the data collection method and the reporting instructions.
It is reasonable to preserve the original readings in the data table as shown by the stopwatch. When reporting a calculated mean, choose a precision consistent with the measurement context and the teacher’s specifications. In the illustration, 2.30 seconds may be a clear summary, but it should not be read as guaranteeing accuracy to one-hundredth of a second. If the goal is to communicate uncertainty, the spread and method are often more informative than a longer decimal expansion.
Avoid teaching a universal rule that all stopwatch results must have a fixed number of decimal places. Scientific reporting conventions depend on the instrument, uncertainty assessment, calculation and level. The more useful rule is: do not allow the calculator’s digits to outrun the evidence.
This also clarifies how to answer an exam question about measurement. If asked for a value, report according to the expected format. If asked about reliability, explain how the timing method, observer and trial-to-trial variation affect confidence. The same number can be presented correctly in a school worksheet and still require caveats in a serious investigation.
13. A fifteen-minute Clementi-floor investigation
Place two clearly marked lines along a safe ramp and use a toy car. Give each of three students a role: release operator, timer and recorder. Write the start and finish rule together before the first trial. Record five manual times, with a short note for any visibly unusual run. Then keep the ramp and car unchanged but swap the timer operator for another five trials.
Compare the spread of each set. Do not assume different averages prove different reaction speed, because the car may have been released slightly differently and the samples are small. Instead, ask what the design allows you to say and what it does not. Record whether the start rule was followed consistently. If a safe phone video or school sensor is available, compare one set of physical crossings with the manual times while acknowledging each method’s limitations.
Next ask the group to design a better version using the equipment they actually have. Perhaps they can use a longer track, a more visible crossing marker, the same timer operator across conditions, practice trials, or a recorded video. Each proposed change should include a reason and a possible side effect. Longer track: easier timing but may change speed. Video: better frame review but may be limited by frame rate. Same operator: improves consistency but may retain personal bias.
Finish with a short written claim: “Our timing results suggest …, but because …, we cannot confidently conclude … . The most useful next improvement would be … .” This trains learners to use limitations as part of scientific reasoning, not as an apology pasted onto the last line of a report.
14. The teacher’s route: separate the concept from the measuring method
Before a practical lesson, state what students are meant to learn. If the target is speed = distance ÷ time, pupils need a meaningful distance and time and should see why the relationship makes sense. If the target is experimental design, pay closer attention to definitions, controls and uncertainty. If the target is using a stopwatch safely and correctly, a manual timing activity may be the right tool even though it is imperfect.
Avoid giving a task in which all three aims compete without explanation. A pupil may spend the whole lesson arguing over a 0.09-second difference while the teacher intended them to understand average speed. Alternatively, an exercise may request a conclusion about tiny changes in travel time even though the manual measurement cannot confidently distinguish them. The question should be answerable at the resolution and accuracy the apparatus can support.
Offer a modelled example of endpoint reasoning. Show one start delay and one stop delay and calculate the displayed interval. Then ask students to predict the direction when the delays are swapped. Only after the concept is secure should they be asked to critique an unfamiliar setup. This sequence is consistent with evidence-informed explicit teaching: clarify the principle, demonstrate it and gradually transfer the decision to the learner.
When assessing student results, distinguish a wrong formula from a sound formula applied to uncertain measurements. One pupil may calculate average speed perfectly from unreliable times; another may collect adequate times but misunderstand the arithmetic. Their next learning steps differ. A correct diagnosis saves time and makes feedback specific.
15. The parent’s route: one question is better than ten criticisms
A parent often sees a practical table with several different times and asks why the student was not consistent. That question can sound like a judgment of effort. Begin instead with curiosity: “Who decided when to press Start and Stop?” Ask the child to describe the two events and whether all trials used the same method. Listen for ambiguity rather than immediately searching for a single expected number.
If the child says, “The stopwatch has two decimal places, so it must be accurate,” demonstrate the difference with a slow visual signal. Press the phone timer when you see a finger rise, then compare with another person’s timing. The point is not to compute a professional calibration constant. It is to discover that human actions are part of the method. Try not to turn a simple discussion into more homework than the assignment needs.
When the worksheet asks for an average, check that the child understands what is being averaged and whether anomalous runs have a documented reason for exclusion. When it asks for a fair-test improvement, encourage a response tied to the problem: “Use a light gate to reduce manual button-press uncertainty” is stronger than “Use better equipment” if the equipment is appropriate and available.
The parent can help most by rewarding an explanation that respects evidence. “We cannot tell from these times alone” may be a more scientifically mature answer than an unjustified precise claim. Science education includes learning how not to overstate what a classroom result proves.
16. The learner’s route: a five-question timing audit
After recording time data, ask yourself: What physical event was the start? What event was the finish? How did the timer detect each? Did the same procedure apply across trials? Is the variation small enough for the scientific comparison I want to make? If you cannot answer one question, mark it as an uncertainty or method limitation rather than inventing a confident story.
Use a short annotation in your notebook: “Timer started on the front wheel crossing marker A and stopped on the front wheel crossing marker B; operator unchanged; five trials.” It is much more informative than “Timed with stopwatch.” If the class uses voice commands instead of visual line crossings, write that instead. Clarity does not require using the teacher’s preferred method when that is not what happened; it requires reporting what actually happened.
For a variable-testing question, check whether the pattern in the data is larger than the variation between similar trials. If two ramp setups differ in mean time by 0.01 second but manual trials within each group vary by 0.2 second, a strong claim about that tiny difference is not supported by the simple evidence. That observation does not prove the setups behave identically. It means the current method may not discriminate them.
Finally, practise explaining the delay sign without the formula. Late Start tends to shorten the recorded duration; late Stop tends to lengthen it. When you can reason that through, you are less likely to memorise a misleading statement that “reaction time always makes time too long.”
17. Common failure modes, each with a different repair
More decimal places are mistaken for accuracy. Repair by comparing display resolution with the method’s response delay and uncertainty.
A late press is assumed always to increase duration. Repair by separating the start and stop endpoints and calculating their effect on the difference.
Different finish cues are mixed in one data table. Repair by agreeing on one visible feature and recording the method.
The operator changes halfway through a variable comparison. Repair by controlling or recording the operator as a possible influence.
Repeated trials are treated as proof of correctness. Repair by explaining the difference between repeatability and bias; compare with an independent suitable method where possible.
An unusual value is deleted because it spoils the average. Repair by investigating the event and documenting the exclusion rule, if any.
A faster object is timed with the same method without considering relative error. Repair by matching the technique to the event duration while protecting the scientific target.
A sensor is assumed infallible because it is electronic. Repair by checking its trigger condition, placement and calibration.
The calculation is criticised although the data collection is the weak link. Repair by tracing the evidence chain from physical event to button press to table to inference.
These are not nine ways of saying “be careful.” They are nine different reasons a numerical result may not answer the intended question.
18. Evidence and caveats
The technical grounding for this guide includes the US National Institute of Standards and Technology’s publication on stopwatch and timer calibrations, including methods and sources of uncertainty. That guide was written for measurement professionals, not to prescribe a primary-school practical lesson. The classroom activity and example times here are original illustrations informed by the general measurement principles. They are not a reported experiment involving children in Clementi and they should not be interpreted as norms for human response delay.
Educational guidance from AERO supports explicit demonstration, manageable chunks of new information, checks for understanding and independent application as competence develops. That helps justify the teaching sequence of first showing two endpoints, then practising with examples, then analysing a real practical. It does not provide a measured effect size for this particular stopwatch lesson.
Uncertainty can be analysed with sophisticated statistical methods, but a Secondary 1 or 2 student does not need to calculate a full metrology budget to learn the core distinction. Conversely, advanced students should not mistake the simple b − a model for a complete treatment of timing uncertainty. Laboratory devices and protocols vary. When timing has safety, engineering or competition consequences, follow the relevant professional standards rather than a classroom shortcut.
The most dependable conclusion is modest: manually timed data should be interpreted as results of a particular human–instrument method, not as perfect readings of the event itself.
19. Frequently asked questions
Does a stopwatch showing 0.01 seconds measure that accurately?
Not necessarily. It displays intervals in increments that appear to be 0.01 seconds. Manual start and stop actions can introduce larger uncertainty. Readout resolution, repeatability and accuracy are different ideas.
Does pressing the stopwatch late always make the time too long?
No. A late start tends to shorten the measured interval; a late stop tends to lengthen it. The net effect depends on both delays.
Should we subtract 0.2 seconds to correct for human reaction time?
Not as a universal rule. Endpoint delays vary, may partly cancel and depend on the task. A fixed adjustment without measurement or justification can make the result worse.
Do three trials prove a Science investigation is reliable?
Three trials provide some information about variation but do not guarantee accuracy. A shared bias or inconsistent method can affect all three. Whether more trials are useful depends on the task and evidence required.
Is video always better than manual timing?
No. Video can reduce certain button-press problems and allow repeated review, but frame rate, blur, perspective and event definitions matter. Choose the method that fits the scientific question.
What if the teacher wants exactly two decimal places?
Follow the assessment’s recording instruction while understanding that the number of reported digits does not guarantee corresponding accuracy. If asked to evaluate results, discuss method and uncertainty separately.
Why do two pupils get different times for the same motion?
Possible explanations include differing start/stop responses, event definitions, release conditions, instrument operation and genuine physical variation. More evidence is needed to identify the cause.
How can I improve a school timing experiment without buying equipment?
Define the cues, mark the course clearly, rehearse the method, use a suitable duration, keep roles consistent where sensible, repeat trials, document anomalies and compare results modestly. If timing uncertainty still dominates, state that limitation honestly.
One last thought before the next practical
The stopwatch is not lying. It is faithfully reporting an interval between button events. The scientific question is whether those button events represent the physical interval we actually wanted. Asking that question once can improve an entire practical report: the start becomes clear, the finish becomes clear, the evidence gains limits, and the conclusion becomes proportionate to what was measured.
For the different issue of a sensor that takes time to settle after being moved, read Why a Thermometer Reading Changes After You Move It. For an investigation in which repeated trials repeat a shared mistake, read Why Repeating a Science Experiment Can Repeat the Same Mistake. These are separate error mechanisms, not interchangeable labels. The How Education Works hub brings measurement reasoning back to the wider question of how learners develop reliable judgment.
Sources and further reading
- National Institute of Standards and Technology, Stopwatch and Timer Calibrations (2009 edition), published 28 January 2009. Recommended practice on timing methods, calibration, traceability and uncertainty.
- National Institute of Standards and Technology, Stopwatch and Timer Calibrations, published 1 January 2004. Earlier guidance establishing the scope of the timing-calibration problem.
- Australian Education Research Organisation, Explicit instruction: Practice guide. Teaching background for demonstrating methods before independent analysis.
- Australian Education Research Organisation, Scaffold practice, updated 14 May 2026. Teaching background for guided and independent practice.
Editorial note: All toy-car timings and classroom exchanges are illustrative scenarios; no human reaction-time experiment or device calibration was conducted for this article.