Approve a method before trusting a conclusion
The strongest English, Mathematics and Science learner can do more than correct a finished poster. They can examine a proposed study before it begins, find the conditions needed for meaningful measurements and decide which later claims would be justified. This volume uses an original breathing-rate learning activity to practise exactly that skill. The output is a protocol approval note, not a health assessment.
After Vol 0573’s English situational email, the question changes: what would a careful student committee need to check before collecting observations? This volume owns that upstream design decision; Vol 0570 addresses review of an existing respiratory evidence hearing, while Vol 0572 owns distribution comparisons.
Quick summary for the learner
State the purpose of the activity, identify a comparable pair of observations, keep the timing and counting method consistent, distinguish volunteers from complete records, preserve units and anonymity, explain the biological mechanism at the correct level, and approve only statements that follow from the measured data. Keep the proposed activity voluntary, gentle and school-supervised; do not turn it into a medical test.
For the 2027 examinations, use the official SEAB G3 syllabus directory. English is K300, Mathematics is K310 and pure Biology is K325. Combined Science offerings carry separate G3 codes and their own syllabuses. The work below is a cross-subject teaching exercise, not a new SEAB examination paper.
How the three disciplines divide the work
Biology explains gas exchange, breathing mechanics and the relationship between respiration and the body’s demand for energy. It does not turn one school measurement into a diagnosis. Mathematics specifies the numerator, denominator, measurement interval, paired observations, units, distribution and graph. English identifies exactly what the evidence can support in a proposal, label or short report.
Confusing these jobs creates errors. A mathematically accurate mean cannot rescue an invalid comparison of unlike groups. A correct sentence about alveoli does not validate an uncontrolled experiment. A persuasive poster does not improve the reliability of its data. Think of the three subjects as cooperating checks on one decision.
The invented classroom protocol
A committee proposes to invite twenty volunteers to a short, supervised educational activity. Each student who chooses to participate may count breaths at rest, undertake a brief gentle walk and count again at a specified point afterwards. The committee plans to record anonymous codes, time windows and observations. The hypothetical completed dataset below contains eight full pairs; its numbers are invented for teaching.
No pupil should be pressured to participate, continue or report a personal result. The task is to decide whether the learning protocol can collect comparable classroom observations, not to prescribe exercise, screen illness or compare pupils’ health. Schools and supervising staff should determine what activities are appropriate in the actual setting.
Begin with the research question
A defensible classroom question is: ‘Among volunteers with complete paired observations, what differences in recorded breathing rate appear between a resting observation and a count shortly after gentle walking?’ It defines the group, the measured variable and the two conditions. It does not ask the data to establish a long-term health outcome.
A weak question is: ‘Does walking make everyone healthy?’ That would require a different kind of evidence and cannot be answered by two simple classroom counts. Rewriting the question before data collection is often the most important design repair. If the question becomes narrower, the eventual answer can become much more reliable.
Agree on what a rate means
A breathing rate is the number of breaths expressed per unit time, commonly breaths per minute. If students count for 30 seconds, doubling the count converts it into breaths per minute. If one uses 20 seconds and another uses 30, the raw counts cannot be compared directly. Record the actual observation window rather than relying on memory.
The transformation must be visible. Nine breaths observed in 30 seconds correspond to 18 breaths per minute, assuming the count was accurate. Nine breaths observed in 20 seconds correspond to 27 breaths per minute. The same count number can represent two different rates. A source table without the interval is not ready for a claim.
Make the before-and-after pairing real
A paired design records both observations for the same coded participant. If the second observation is missing, that row is not a complete pair. The student’s identity need not be public, but the two observations need a stable code so they can be matched correctly. Do not merge one person’s resting observation with another person’s post-walk count.
A paired design helps describe within-person differences under two defined conditions. It does not, by itself, settle every cause of the difference; movement, timing, individual variation and measurement error may still matter. A school study should report what was controlled and what remains uncertain instead of claiming the design removed all limitations.
Distinguish timing from elapsed time
‘After walking’ is incomplete unless the protocol identifies when the count begins. A count begun immediately after stopping and one begun several minutes later may describe different parts of a response and recovery. The committee must choose a practicable observation point and use it consistently, or record elapsed time as part of the dataset.
A graph that connects measurements at different time points can be interesting, but it answers a time-course question rather than a single before-and-after comparison. Decide the question first. A student’s choice to stop or skip a reading must be respected; incomplete observations should not be silently relabelled as zero.
A completed eight-pair example
Consider the following invented rest-to-after readings, in breaths per minute: A 14 to 20; B 16 to 22; C 15 to 19; D 13 to 18; E 18 to 24; F 16 to 23; G 17 to 22; H 15 to 21. Each letter represents one matched observation pair. These eight records are sufficient for a teaching calculation, not for generalising to a whole population.
The resting total is 124, giving a mean of 15.5 breaths per minute. The after-walk total is 169, giving a mean of 21.125. The difference between means is 5.625 breaths per minute. Individual paired differences are 6, 6, 4, 5, 6, 7, 5 and 6, which also sum to 45 and average 5.625. These equalities rely on comparing the same eight matched records.
Explain the difference of means carefully
For the same complete pairs with equal weighting, mean(after minus before) equals mean(after) minus mean(before). That algebraic relationship is useful for checking arithmetic. It does not hold unchanged if the before and after averages are calculated on different sets of pupils, or if one side silently omits records.
Keep the qualification near the number. A valid line is: ‘Across these eight matched records, the mean post-walk rate exceeded the mean resting rate by 5.625 breaths per minute.’ A misleading line would be: ‘Walking increased every student’s health by 5.625.’ The units and interpretation would both be wrong.
A small mean hides variation
The individual differences in the invented set range from four to seven breaths per minute. Their mean does not describe every participant’s precise response. It is a useful centre, not a replacement for the eight rows. A graph or table can show spread, but you must choose a form that keeps the question clear.
Revisit Vol 0572 for median, quartiles and spread. In this volume, the goal is not to redo that lesson; it is to decide whether the protocol records enough detail to make distribution-based statements defensible. A committee should collect observations that can be checked rather than only a headline average.
Identify missingness before writing the headline
Suppose twenty students were invited, fifteen volunteered, twelve attended and eight produced complete pairs. All four counts can appear in a participation flow, but only eight belong in the paired calculation. If missing readings are related to withdrawal or measurement difficulty, their absence may also affect interpretation. Do not call them zeros and do not guess the missing results.
A responsible report can state the number of complete pairs and acknowledge that results represent a limited self-selected group. The report should not imply that the people who did not participate had the same response. This is part of a fair communication method, not a weakness to hide.
Biology explains the mechanism, not the value of a person
Breathing moves air to and from the lungs; gas exchange occurs across alveolar surfaces; blood transport and cellular respiration are connected but distinct processes. An observed breathing-rate change does not directly measure the rate of diffusion, the amount of oxygen carried in blood or how healthy a particular student is.
Use Vol 0571’s alveolar gas-exchange explanation for the biological mechanism. When a question asks why breathing becomes more rapid after physical exertion, use syllabus-level processes appropriate to the stated conditions. In a school poster, avoid importing a physiological diagnosis or a claim about future health that the classroom activity never investigated.
Write observable variables separately from interpretations
A table can record anonymous participant code, observation interval, rest count, post-activity count, calculated rate and the time point. It should not contain a column labelled ‘health quality’ computed from breathing rate. That would disguise an opinion as a measurement.
Distinguish raw count from converted rate, and measured result from explanation. For example, a recorded ten breaths in a 30-second window is a raw count. Twenty breaths per minute is a conversion. ‘This pupil’s lungs improved’ is an untested conclusion. The first two entries belong in a measurement worksheet; the third does not.
Ask what would make the data non-comparable
Different counting windows, inconsistent timing, missing resting periods, counting while talking and switching from self-counts to observer-counts can all undermine comparison. A fair educational protocol should explain which method will be used and when a record will be treated as incomplete. A committee that names these problems early can improve its exercise before anyone writes a poster.
Do not overengineer the study into a medical investigation. The value of this exercise is teaching control of variables and honest interpretation under safe school supervision. The protocol should be simple enough for staff to review and students to understand.
Choose graphs for the actual claim
For matched before-and-after observations, a table or paired display can make each relationship visible. A bar chart of two means can give a simple overview but hides individual variation. A time-series graph is suitable only if readings are actually taken at named times. Axes must say what the numbers are and what units they use.
A misleading vertical scale can exaggerate a small change. An unlabelled graph can imply universality. Decide whether the reader needs the comparison of two group means, the spread among individual responses or the sequence over time. Those are different questions, and one chart need not answer them all.
The English approval note must match the evidence
A cautious public sentence could say: ‘In eight complete matched records from a voluntary classroom demonstration, the mean breathing rate was higher shortly after gentle walking than at rest.’ An honest qualification follows: ‘These observations are a learning example and cannot assess individual health or long-term benefits.’
A claim should be rejected if it says everyone became healthier, if it reports twelve paired results when only eight are complete, or if it describes a diagnostic exercise. English does not merely decorate a correct calculation; it controls how that calculation is communicated to another person.
Make a protocol approval decision
Approve the learning activity only if the named school supervisor agrees the proposed activity and observation method are suitable, participation is voluntary, recording is appropriately managed, the interval and timing are specified, and the final display will state the sample and its limitations. If any critical element is missing, ask for revision before collection rather than pretending the conditions are already met.
This is not a legal checklist or a guarantee of physical safety. It is a classroom reasoning model showing how a student can make a bounded decision with evidence and clear conditions. If school rules or actual health circumstances require more stringent review, the teacher and responsible adults determine that process.
Fit the skill to real 2027 examinations
K300 English tests source use, audience, writing and comprehension. K310 Mathematics includes data handling, line graphs, statistics and reasoning in real-world contexts; its two papers last 2 hours 15 minutes each and carry equal 50% weight. K325 Biology assesses knowledge, application and experimental skills. Combined Science students must read the syllabus for the particular subject combination they take.
See the K300 English, K310 Mathematics and K325 Biology official syllabuses. EMS is this site’s interdisciplinary study route, not the name of an additional SEAB examination paper. Each subject remains assessed under its own official scheme.
Protocol clinic 1: The count was not a rate
The situation: One student records eight breaths in twenty seconds while another records eight in forty seconds. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: A report says their breathing rates were equal because both raw counts are eight. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Convert each count to a per-minute rate using its own time window: 24 and 12 breaths per minute respectively. Try this check: Ask what metadata must be recorded so that another learner can reproduce the conversion. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 2: The pairing is broken
The situation: One resting reading is linked to a different pupil’s after-activity result. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The spreadsheet calculates a change for a person who never provided both observations. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Use anonymous stable codes and count only true matched pairs in paired-change analysis. Try this check: Check how many complete pairs remain after the mismatched row is excluded. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 3: A high reading is called unhealthy
The situation: An after-activity number is higher than its resting partner. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: A display labels the student unhealthy or unfit. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Report the observed difference without assigning a medical judgement; the measurement alone does not support it. Try this check: Replace health language with an objective statement of method, rate and time. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 4: Missing means zero
The situation: A volunteer stops before the second reading. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The committee enters zero so that the row remains in the graph. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Mark the second observation missing and do not include a fabricated change in the paired analysis. Try this check: Distinguish a genuine measured rate of zero from an absent observation in a general data-handling problem. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 5: The group has changed
The situation: Resting means are based on twelve pupils but post-activity means on eight. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The writer subtracts the means and calls the result an average within-person change. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Restrict the comparison to the same matched pupils if studying individual changes, or explain that two groups were used. Try this check: State the assumption required for the difference-of-means equality applied in this guide. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 6: A timer is started late
The situation: Half the volunteers begin counting immediately and others begin several minutes later. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The poster combines all readings as if they represented the same time after walking. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Specify the delay and make timing comparable; if timing differs, record and report it rather than concealing it. Try this check: Describe how the scientific question changes when the aim becomes studying recovery over time. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 7: The mean is the only visible result
The situation: A report displays 5.625 as the mean difference for eight people. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: Readers are led to think every pupil changed by exactly 5.625 breaths per minute. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Show the individual differences or their spread alongside the mean where relevant. Try this check: Use the range from four to seven to explain what the mean leaves out. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 8: Unclear diagram units
The situation: A chart axis reads ‘rate’ with ticks 10, 20 and 30. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: A reader cannot tell whether numbers are breaths per minute or counts per observation window. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Label the axis and ensure the graph values match the recorded unit and conversion. Try this check: Write a title and axis labels suitable for a paired-before-after classroom chart. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 9: A poster turns learning into treatment
The situation: The proposed programme is presented as a therapy for lung problems. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: A short observational activity is incorrectly given a clinical purpose. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: State the educational aim and refer genuine health questions to appropriate professional and school procedures. Try this check: Draft a one-sentence public disclaimer that is clear without alarming students. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 10: A volunteer is publicly ranked
The situation: The display includes individual names beside higher or lower readings. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: Other pupils infer a health ranking from the ordering. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Use anonymous codes and focus on measurement variation rather than ranking individuals. Try this check: Decide whether a grouped display would communicate the learning point with less exposure. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 11: The measurement method changes
The situation: Resting counts are self-reported while after counts are observer-counted. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The report assumes both methods are perfectly interchangeable. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Record methods and keep the approach consistent where practical; acknowledge method changes if unavoidable. Try this check: Identify the potential measurement-bias question before interpreting any apparent difference. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 12: A teacher’s approval is invented
The situation: The activity has not been reviewed by staff. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The student notice says the school has approved the session and all safeguards. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Use conditional wording and submit the proposed protocol for review rather than claiming approval. Try this check: Rewrite the notice as an approval request and identify what details the teacher should check. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 13: The claim confuses ventilation and pulse
The situation: The graph records breaths per minute. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The writer calls the series a measure of pulse rate or blood oxygen saturation. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Retain the measured variable’s correct name and do not substitute another physiological quantity. Try this check: List the separate devices or observations that would actually be needed for the other quantities. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 14: One volunteer dominates a story
The situation: A single pupil reports a strikingly large change. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The display replaces the full set with that student’s anecdote. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Report the observation as one case and rely on the matched set for a statement about the group. Try this check: Discuss why a carefully chosen anecdote can be persuasive yet statistically unrepresentative. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 15: A class becomes a population
The situation: Eight self-selected students supply completed records. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The report declares the same result true for every secondary pupil in Singapore. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Keep the inference within the observed classroom set and explain the selection limitation. Try this check: Rewrite ‘all teenagers’ into a source-bounded phrase that survives checking. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 16: A conclusion outruns its time horizon
The situation: The activity produces observations immediately after walking. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The committee promises long-term improvement in lung capacity. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Report the measured short-term rate difference and avoid unobserved long-term outcomes. Try this check: Name the gap between the observed time window and the proposed duration of the benefit. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 17: A participation figure is misplaced
The situation: Twenty were invited, fifteen volunteered, twelve attended and eight completed both counts. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The report describes twenty complete paired measurements. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Reserve eight for analysis and use the larger counts only when describing recruitment and participation. Try this check: Draw a simple flow from invitation to complete record without implying missing readings are failures. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 18: A positive result is defined after the fact
The situation: The committee wants the poster to say the pilot was a success. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: It changes the meaning of ‘success’ to match whichever number looks attractive. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Define an educational outcome such as correct unit conversion or accurate identification of limitations before collecting records. Try this check: Distinguish success at learning from proof that the activity improves health. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 19: An average is too precise for the purpose
The situation: The computed mean after activity is 21.125 breaths per minute. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: A public headline uses many decimal places to suggest medical certainty. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Keep the exact working in Mathematics, then choose sensible precision for a clearly identified illustrative report. Try this check: Explain why arithmetic precision and strength of evidence are not the same thing. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 20: A source term changes meaning
The situation: The original data sheet says ‘completed measurements’. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The poster changes it to ‘successful participants’ without defining success. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Use the source’s neutral category rather than introducing an evaluative label. Try this check: Make a glossary for invited, volunteered, attended, paired, observed and inferred. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 21: A chart begins at fifteen
The situation: A bar chart uses a narrow vertical range to magnify the change. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The audience reads the difference as more dramatic than the data warrant. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Explain the axis scale and select a representation that does not hide crucial context. Try this check: Redesign the same result with a clear axis title and visible starting value. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 22: A researcher confuses correlation and mechanism
The situation: The students who volunteered happened also to like sport. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The writer asserts that their higher rates were caused by being sportier. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Treat self-selection and physiological interpretations cautiously; do not assign a cause not tested by the exercise. Try this check: Ask what a fair comparison would need before making a new claim about sport participation. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 23: A line graph invents readings
The situation: Only rest and immediate-after values were collected. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The display draws a smooth curve showing what happened every minute between them. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Use two observed points without inventing intermediate measurements; a connecting line is not evidence of each missing value. Try this check: Describe what additional time-stamped observations would support a genuine recovery curve. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 24: The email and poster disagree
The situation: The teacher email states eight complete pairs; the draft poster claims twelve. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The committee publishes both versions without reconciliation. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Resolve the dataset and wording before release; otherwise the audience receives inconsistent evidence. Try this check: Compare every published figure against the same approved table and retain its denominator. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
Protocol clinic 25: A limitation is treated as defeat
The situation: A student finds the sample too small for broad health claims. Before writing a conclusion, identify which record, condition or comparison is at risk. A good protocol exists to prevent this weakness from being hidden inside a later graph, paragraph or publicity line. This is the same habit that supports careful work in a G3 examination question.
The tempting mistake: The committee abandons any educational purpose for the activity. The error matters because the question being answered is no longer identical to the question the evidence can answer. Better-looking calculations or polished wording will not restore the missing connection between method and conclusion.
The repair: Narrow the goal: collecting comparable observations and learning to recognise limitations can itself justify the classroom study. Try this check: Propose an honest title for the display that would still interest learners without promising a health outcome. Then ask whether a reader who was not present could understand the measured variable, the sample, the timing and the limit of the finding from your final explanation. If not, revise the method or the claim.
A final source-to-decision rehearsal
Give yourself an unseen school scenario. Write one paragraph identifying the actual scientific variable, one displaying a small correct calculation and one explaining what a public statement is allowed to conclude. Then write an approval decision: approve, revise or reject, with reasons. Do not guess what the marker wants to hear; show how each conclusion follows from specified information.
The longer route continues from PSLE respiratory and circulatory foundations to the Secondary Science owner, Secondary Mathematics owner and Secondary English owner. This guide’s distinctive job is to integrate them in a method-before-claim decision. All observations, pupils and committee decisions here are hypothetical.
