The greenhouse learning team has three pieces of information on its desk. Chemistry has classified several fictional solutions with an indicator. Mathematics has converted a model sensor signal into an estimated concentration. English has found an unsupported claim in a draft exhibition report. The final task is not to calculate another number. It is to decide what the school can responsibly publish, what it should revise and what remains unknown. This guide shows how to conduct an evidence audit that connects, but never confuses, the three subjects.
The Singapore–Cambridge Secondary Education Certificate (SEC) begins in 2027. Relevant official G3 subject codes are K300 English Language, K310 Mathematics and K324 Chemistry. They are separate assessed subjects, each with its own syllabus and papers. EMS here means an independent English–Mathematics–Science learning exercise; there is no official combined EMS paper. The school committee, samples, sensor, readings and draft poster are fictional. Nothing in this educational exercise certifies real irrigation, drinking or food safety.
Quick route: State the release decision; classify each piece of evidence; check Chemistry’s qualitative scope; verify Mathematics’ equation and interval; repair the English claim; identify missing information; decide what can be displayed; finish with examination-style transfer questions. Vol 1387 teaches the acid-base test, Vol 1388 teaches the numerical calibration, and Vol 1389 teaches the incident report. This final article integrates their evidence into one limited publication decision.
1. Name the actual decision
The fictional exhibition committee can decide whether to display corrected classroom materials. It cannot authorise real water use, declare food safe, or award regulatory approval. Write the decision question as precisely as possible: “May we publish an educational poster containing the supplied indicator observations and the stipulated model concentration estimate, with clear qualifications?” A broad question such as “Did the greenhouse pass?” has no defined test or authority.
This framing saves work. We do not need to inspect every scientific possibility to approve a truthful explanation of a classroom model. We do need to prevent the poster from implying that its fictional data constitute a professional field test. A good student starts with the decision that the evidence is supposed to inform. In a school examination, the same habit helps identify what a long passage is really asking.
2. Establish an evidence ledger
Create an evidence ledger with columns for source, quantity or observation, method, claim supported and limitation. Chemistry supplies colour observations from teacher-provided classroom samples; Mathematics supplies a reference table and a stipulated linear relation; English supplies the draft and corrected report. Every item should be traceable to its source. A recommendation can then point to the evidence that justifies it.
This simple structure prevents one assertion from becoming three pieces of evidence. If the Mathematics team calculates a concentration, and the English team repeats it in a report, the report is not a second independent measurement. It is a communication of the first calculation. Likewise, an exhibition headline is not evidence about the sample merely because it appears beside a scientific-looking graph.
3. Use a common vocabulary for claims
The committee adopts four labels: observation, model result, inference and decision. “Sample A turned Universal Indicator orange” is an observation in the fictional exercise. “R = 22 corresponds to C = 3.4 mg/L” is a result from the stipulated linear model. “The colour indicates acidic character” is an inference based on the indicator. “Publish the amended display” is a decision.
Each label invites a different check. Observations require trustworthy recording; model results require correct arithmetic and appropriate assumptions; inferences require a relevant scientific principle; decisions require stated goals and limits. A strong written response signals these levels through verbs: recorded, calculated, suggests, recommends. It does not treat an attractive poster as a substitute for an evidential chain.
4. Keep the Chemistry and Mathematics tasks separate
The Chemistry test addresses relative acidity and alkalinity through indicator colour. The Mathematics calibration addresses the signal of a different fictional sensor responding to a model solute. The two are not interchangeable. In particular, a pH reading must not be inserted into R = 5C + 5 as though it were a solute concentration. That equation belongs only to the instrument and reference table stipulated in the Mathematics guide.
Students are often tempted to use a familiar calculation wherever they encounter a number. The committee must instead ask which quantity the number measures. Orange colour, pH category, arbitrary sensor signal and milligrams per litre are distinct. A cross-subject study is successful when the boundaries remain clear, not when all information is forced onto one invented scale.
5. Reconstruct the supplied Chemistry findings
The fictional solution A appears orange with Universal Indicator, B appears green and C appears blue under the same classroom viewing method. The cautious interpretation is that A is acidic, B near neutral and C alkaline according to the supplied indicator reference. These are broad categories. The test does not identify the dissolved substances or prove any sample is suitable for consumption or real irrigation.
The team’s audit should preserve the qualitative wording. “B is safe because it looks green” is not an allowed conclusion. A neutral solution may contain unwanted substances, and the test was not designed to assess them. If a task asks for the role of the indicator, explain that it provides information about acid-base character. If a task asks for safety, explain that the supplied evidence is insufficient.
6. Check the reaction explanation
The Chemistry guide’s central ionic relationship is H⁺ + OH⁻ → H₂O. It explains neutralisation when acid-derived hydrogen ions react with alkali-derived hydroxide ions. The committee can include this teaching equation on an educational display if it is labelled as an example of the principle. But an equation alone does not prove what substances were in a particular sample or what happened in a real greenhouse.
An important transfer skill is distinguishing a general scientific law or model from an observed event. A correct neutralisation equation cannot validate an unknown field procedure that no one performed. Similarly, a fair plan does not become a completed experiment. In the exam, students should use relevant principles to explain provided observations, not use principles to invent observations that the question never supplied.
7. Separate neutralisation from dilution
Suppose the draft text says, “The water became neutral when we added more water.” That sentence confuses dilution with a reaction. Adding solvent can reduce concentration without necessarily achieving neutrality. Neutralisation specifically involves acid-base reaction. The audit should correct both the scientific concept and the English wording, because an attractive sentence can still misteach chemistry.
A useful correction is: “In a model acid-base reaction, added alkali can supply hydroxide ions that react with hydrogen ions. The amount required depends on the reagents and their concentrations.” The committee need not give instructions for treating real water. This is a conceptual teaching statement. The distinction also demonstrates why checking facts before polishing prose is more efficient than editing language alone.
8. Check whether the methods were fair
A credible classroom comparison holds relevant conditions constant. The colour readings should be made using the same indicator procedure and viewing conditions; the sensor calibration should use a consistent reference method. If either method changes between samples, the committee cannot automatically attribute the differences to the sample property of interest. A report should identify which controls were actually specified.
The fictional dossier provides a controlled instructional setup but not a professional validation package. Its strongest claim is a comparison within that setup. If asked how to improve the method, name a specific weakness: standardise lighting for colours, check the instrument’s reference reading, repeat the measurement or extend the calibration range where appropriate. Avoid “be more careful” as a generic substitute for scientific reasoning.
9. Reconstruct the linear model from first principles
The calibration pairs are concentration C in mg/L and signal R in arbitrary units: (0,5), (2,15), (4,25), (6,35). The change in signal from 5 to 35 is 30 units, while concentration changes by 6 mg/L. Gradient = 30 ÷ 6 = 5 signal units per mg/L. At zero reference concentration the signal is 5, so the vertical intercept is 5. The relation is R = 5C + 5.
The committee does not have to trust the formula because it appeared in a previous article. Anyone reviewing the poster should be able to rebuild it from the table. That independent check is the purpose of an audit. The data fit this exact classroom line, but the line is stipulated over the observed interval. It is not a universal instrument law. A perfect algebraic fit is not a substitute for method validation.
10. Verify the corrected number
The unknown model sample has a signal reading R = 22. Substitute into 22 = 5C + 5, giving C = (22 − 5) ÷ 5 = 3.4 mg/L. The concentration lies between the 2 and 4 mg/L reference levels, so it is an interpolation within the tested interval. The original poster had wrongly labelled the raw signal as “22 mg/L”.
The audit should display both quantities with correct labels if they are important to the learning objective: “Signal reading: 22 units. Model-derived concentration estimate: 3.4 mg/L.” A reader can then see that the first number was recorded and the second calculated. This is a small but essential form of transparency. Without labels, apparently precise figures can become meaningless or misleading.
11. Test the arithmetic independently
One reviewer should check the subtraction and division without copying the first student’s work. Another may use a graphical reasoning check: 22 lies between signals 15 and 25, so the predicted concentration must lie between 2 and 4 mg/L. The value 3.4 fits. A mistaken answer of 34 mg/L would fail this interval check even if a calculator displayed it without warning.
Independent checking is not the same as asking two people to agree. If both copy the same equation with an incorrect intercept, the error persists. A better audit uses different routes where possible: derive the line from two points, test a third reference point, solve the equation and compare the answer with nearby observations. Diverse checks reduce the chance that one misunderstanding will pass through the system unnoticed.
12. Check the stated uncertainty separately
Assume the fictional sensor reading is R = 22 ± 1 signal unit. Under the stipulated increasing line, R = 21 yields C = 3.2 mg/L and R = 23 yields C = 3.6 mg/L. Thus the model interval is 3.2–3.6 mg/L under that input bound. Do not convert the interval into an unsupported probability statement.
The committee may choose whether that degree of detail is useful for a public educational poster. For a learner-facing explanation, it could say “about 3.4 mg/L under the model, with a corresponding 3.2–3.6 mg/L range if the signal varies by one unit.” The condition must appear with the range. Leaving it out would suggest that the interval came from a validated field instrument or a statistical confidence analysis, neither of which the task provides.
13. Distinguish measurement spread from accuracy
Repeated sensor readings such as 21, 22 and 23 show some variability. Their mean is 22, and their range is 2 signal units. But close readings do not prove that the sensor is accurate. If the instrument has a systematic offset, all readings may agree and still be wrong relative to a valid reference. The committee should not turn repeatability into a guarantee of true concentration.
This is a transferable Statistics lesson. A mean summarises centre, a spread statistic describes variation and calibration connects the reading to a reference. Each answers a different question. An audit that uses “consistent” as a synonym for “correct” erases those distinctions. State what was tested: the group has repeat readings under an instructional method, not proof of independent chemical validation.
14. Mark the calibration boundary
The calibration reference interval is 0–6 mg/L. A reading of 22 gives an estimate within that range. If another sample produced a signal of 41, the formula would return 7.2 mg/L, but that result would be an extrapolation beyond the observed interval. The committee should not publish it as a validated concentration without further information.
Write the boundary directly in the evidence ledger. The phrase “within the calibrated range” is not bureaucratic caution; it tells the reader which inferences the model can reasonably support. A careful report communicates the difference between mathematically extrapolating a line and demonstrating that a physical instrument continues to behave linearly. Knowing when not to use a formula is a high-level examination skill.
15. Explain the disputed headline
The draft poster contains the statement “22 mg/L confirmed — safe greenhouse rinse water.” It commits more than one error. The first number confuses raw signal units with concentration. The word “confirmed” overstates the status of a model estimate. The safety claim introduces a new assertion that neither indicator colour nor the fictional sensor model tested. Correcting only the arithmetic would leave the most serious communication problem in place.
A proper audit separates these errors so that each can be fixed. Replace the numerical label, qualify the inference and delete the unsupported safety judgement. The result may be less exciting, but it is more educational. The aim of the exhibition is to teach visitors how evidence works, not to advertise a conclusion that was never tested.
16. Review the English report’s chronology
The incident report from Vol 1389 states that the error was discovered during final review before printing. That timing changes the committee’s responsibility. It should amend and verify the poster before publication; it should not imply that a public recall of material already distributed has occurred. A false chronology can create an incorrect narrative even when every scientific number is right.
Check verbs such as “was published”, “will be published” and “has been submitted”. The report must distinguish completed events from planned actions. In a situational-writing examination, this is task fulfilment, not merely grammatical polish. The story’s conditions determine what the recommendation should be. A reader must know whether the next action is prevention, correction or public follow-up.
17. Read for unsupported agency claims
The exhibition committee is a school learning group. It should not use words that suggest official inspection, government certification or a real professional water-quality approval. An educational display can cite official syllabus documents for what students are expected to learn, but that is not the same as an agency approving the group’s samples. Authority and evidence are separate.
If the poster says “certified safe by the Chemistry syllabus”, the sentence is conceptually wrong. A syllabus defines learning content and assessment; it cannot certify a particular sample. The editorial repair is not simply to remove the agency name. The committee should replace the claim with a statement about the activity’s actual scope: demonstration of acid-base classification and linear-model interpretation using supplied fictional figures.
18. Build a claim-by-claim audit table
For each sentence, ask: What is the exact claim? Which source supports it? Does the source directly observe the fact or only model it? What qualification is needed? What wording should replace it? For example, “Sample B appeared green with Universal Indicator” has a recorded classroom observation; “Sample B is safe to drink” has no support from that observation.
Another row might read: “The model estimate is 3.4 mg/L.” That is supported by the calibration table and arithmetic, subject to the model’s stated range and assumptions. The committee should classify the first as qualitative evidence and the second as quantitative inference. A tidy audit table helps expose where a sentence leaps from a narrow result to a much wider claim.
19. Choose a release standard
A useful internal standard for a school display is that every numerical statement must have a named quantity and unit; every inference must name the model or relevant evidence; every important limit must be visible; and every instruction must have an appropriate audience and authority. This standard is not an official government rule. It is a practical editorial checklist for truthful educational communication.
The standard may seem strict, but it prevents avoidable mistakes. It also helps learners revise efficiently. If a number lacks units, fix the label before debating its meaning. If a model is extrapolated, flag the range. If a safety claim has no relevant evidence, remove it. If a recommendation is too broad for the committee’s authority, narrow the action. Each weakness maps to a concrete edit.
20. Decide what can be printed
The committee can print a poster that says: “In a fictional classroom exercise, Universal Indicator gave different colours for supplied model solutions. A separate hypothetical sensor calibration related signal R to model-solute concentration C by R = 5C + 5 for the reference range 0–6 mg/L. A signal of 22 units corresponded to a model estimate of 3.4 mg/L.”
That wording is longer than the misleading headline because it carries necessary distinctions. It can be simplified for younger visitors without sacrificing truth. For example: “Colour tests classify acidity; a separate sensor model estimates another quantity. These classroom examples do not test whether real water is safe.” The two versions serve different readers but express the same evidence boundary.
21. Decide what must be withheld
The committee should withhold claims that actual greenhouse water was tested, that any produce is fit for consumption, that 3.4 mg/L is a safety threshold, or that a universal solution-treatment procedure has been validated. None of those propositions follows from the fictional dataset. It should also avoid implying that one colour or concentration estimate measures every relevant substance.
Withholding a claim is not hiding evidence. The audit can openly state that the available information does not address those questions. A good student learns that uncertainty has different forms: missing measurements, an untested range, method variation, ambiguous wording or lack of decision authority. Naming the particular gap makes the limitation informative rather than defensive.
22. Do not invent a pass/fail threshold
Suppose someone proposes declaring all samples below 4 mg/L “safe”. That threshold did not appear in the supplied material, and the measured model solute is not specified as a hazard with a validated standard. The committee cannot create a scientific safety threshold from convenience or a desire for a green tick on the display. A mathematical comparison is only meaningful when the criterion itself has a legitimate basis.
In real contexts, criteria depend on substance, use, exposure and applicable standards. The fictional school exercise does not provide such a framework. Therefore the report should avoid real-world approval language altogether. In an exam, if a threshold is explicitly supplied as part of a hypothetical problem, students may apply it mathematically while explaining that their conclusion is conditional on the stated rule.
23. Make a fair comparison of alternative statements
Consider three captions. Caption A says “22 mg/L proven safe”. Caption B says “sensor reading 22; estimated 3.4 mg/L under the classroom calibration”. Caption C says “all samples passed”. The only caption supported by the stipulated numerical data is B, though it should still carry the educational context. Captions A and C introduce unsupported identity, certainty or pass/fail conclusions.
The skill is not simply choosing the most cautious sentence. A caption that says “we know nothing” would also be inaccurate, because the exercise supplies meaningful information about a sensor model. Good communication preserves warranted findings while refusing unwarranted extensions. Precision requires both positive evidence and clearly bounded limitations. This balance is useful in English comprehension, scientific conclusions and mathematical modelling.
24. Compare two independent checking methods
One reviewer might rederive R = 5C + 5 from the table and recompute 3.4 mg/L. A second reviewer might read the corrected report without seeing the previous draft and check whether the conclusion matches the supplied stimulus. These are complementary tasks: one tests calculation, the other tests communication. A third Chemistry reviewer can check that the pH discussion has not been merged with the solute-calibration exercise.
Separate checking roles help prevent a common failure: an entire team unconsciously accepts a misleading statement because it sounds familiar. A calculation check cannot detect every unsupported health claim, and a grammar check cannot catch every unit error. Each discipline contributes a distinct form of scrutiny. The value of EMS learning lies in coordinated reasoning, not an invented combined examination grade.
25. Consider a new conflicting reading
Imagine a second model instrument reports a concentration estimate of 4.1 mg/L for the same fictional sample. That differs from the original estimate of 3.4 mg/L. The committee should not automatically average them to obtain 3.75 mg/L and announce an improved truth. It must ask whether the instruments measure the same property, whether their calibrations are valid and whether the readings were made under comparable conditions.
An average is not a conflict-resolution device. If two estimates arise from different assumptions or systematic biases, the midpoint may have no special authority. The right next step is to investigate the discrepancy and report it. In a mathematics problem that explicitly assumes independent comparable measurements, an average may be requested. In an evidence audit, the meaning of the measurements comes first.
26. Consider a damaged reference record
Suppose the reference table has a smudged entry at 4 mg/L. The other three reference pairs might still suggest a consistent line, but the audit should not silently reconstruct the missing number and claim it was observed. It may label a predicted value and explain how the model gives it, then request a check of the original record. The difference between a measured point and an inferred point must remain visible.
This is another useful examination habit. A question may ask students to predict a missing entry from a relation, and the calculation can be entirely appropriate. But the completed table then contains a model prediction, not an independent original observation. When writing a scientific conclusion or a public caption, that distinction matters. The credibility of a report depends on preserving provenance.
27. Consider an outlier without deleting it
Suppose repeated signals are 21, 22 and 40 units for a supposedly unchanged sample. The third value is unusual relative to the others. The committee should record and investigate it rather than quietly remove it to make the graph neat. A possible check is to repeat the measurement under controlled conditions and inspect the instrument record. No single explanation can be asserted without more evidence.
If the high reading recurs, the model or sample conditions may need reconsideration. If a repeat under corrected procedure returns to the earlier range, a procedural issue may be plausible. The appropriate conclusion depends on what the new observations show. A mature answer describes the uncertainty and the next informative test. It does not invent a convenient cause simply to end the story.
28. Consider the limits of a beautiful graph
A neatly drawn straight line can create an impression of certainty. But the line is a representation of data and assumptions; its appearance does not prove the instrument’s reliability outside the supplied conditions. The committee should show axes, units and reference points, then explain the valid interval. A missing or distorted scale could mislead visitors even if the equation printed nearby is algebraically correct.
Students can practise reading graphs defensively. Ask whether the vertical axis begins at zero, whether the intervals are consistent, whether an apparent change is large relative to the measurement scale and whether the plotted line extends beyond observations. This skill connects G3 Mathematics data interpretation with English evaluation of visual texts and Science’s concern for fair measurements.
29. Revisit the 2027 English assessment
The K300 English Language syllabus assesses Writing, Comprehension, Listening and Oral Communication across four papers. The weights are 35%, 35%, 10% and 20%, respectively. The report in Vol 1389 is a Paper 1 Situational Writing practice activity. It is not a new official report paper, and this EMS review is not a fifth English component.
The connection to the wider syllabus is through transferable abilities: writing for audience and purpose, interpreting texts and visuals, tracking evidence, taking accurate notes and discussing a reasoned position. A student who can detect “confirmed” as an overclaim is practising language evaluation. They should still prepare for the formal task types and timings in the official document, rather than using this interdisciplinary activity as their only revision.
30. Revisit the 2027 Mathematics assessment
G3 Mathematics K310 has two papers, each 90 marks and 50%, lasting two hours 15 minutes. The second paper includes an extended real-world application problem. The sensor exercise practises finding a linear equation, interpreting a graph, solving for an unknown and respecting constraints. It should be treated as one context for the subject skills, not a prediction about any actual SEC examination question.
In a formal response, the student needs working and a contextual answer. For R = 22, writing “3.4” alone loses essential meaning. The complete statement is “the model concentration estimate is 3.4 mg/L within the supplied reference interval”. Mathematical accuracy and communication reinforce each other. The result can be wrong even if the algebra is flawless when the variable has been misunderstood.
31. Revisit the 2027 Chemistry assessment
G3 Chemistry K324 consists of Paper 1 multiple choice at 30%, Paper 2 structured and free response at 50%, and Paper 3 practical at 20%. The Chemistry volume practises acid-base definitions, indicator interpretation, balanced equations, planning and evaluation. The committee audit draws on those skills but does not replace laboratory competence or the formal examination syllabus.
A strong Chemistry student explains what the indicator measures, what a reaction means and why a result is limited. They should not confuse a numerical signal from a different model instrument with a measured pH value. The word “chemical” is not a licence to generalise; each measurement has a method and meaning. Good examination thinking starts by linking a question to the right concept.
32. Transfer task: answer a Chemistry question
Question: In the fictional exercise, Sample B turns Universal Indicator green. State one supported inference and one unsupported conclusion. A model answer is that the sample is near neutral according to the supplied indicator chart; it is unsupported to claim that the sample is safe to drink. The indicator result concerns acid-base character, not the full composition or suitability of the sample.
A better answer adds a reason rather than only a disclaimer. Green is an indicator colour interpreted against a chart; suitability for consumption would require separate evidence. Keep the response concise when an examination asks for two points. In an integrated review, the same distinction can be expanded into a policy for what may appear on a school poster.
33. Transfer task: answer a Mathematics question
Question: The calibration pairs are (0,5) and (6,35), and the relationship is linear over the supplied reference interval. Find C for R = 22. A model solution finds the gradient (35 − 5) ÷ 6 = 5, intercept 5, equation R = 5C + 5 and C = 3.4 mg/L. The estimate is an interpolation because C lies between zero and six.
For a higher-demand extension, let R = 22 ± 1 and derive the interval 3.2–3.6 mg/L. Then ask whether a signal of 41 should be reported as validated. The mathematical formula returns 7.2 mg/L, but the tested range does not establish its reliability. This kind of multi-step interpretation trains students to separate what can be calculated from what can responsibly be concluded.
34. Transfer task: answer an English question
Question: Replace the sentence “Our certified water test proved the sample is safe and contains 22 mg/L” with an accurate line for a school display. A suitable revision is: “In our fictional classroom sensor model, the recorded signal was 22 units, corresponding to an estimated 3.4 mg/L within the supplied calibration range. This is not a water-safety test.”
The revision is not timid; it gives the actual finding. It fixes the unit error, removes unsupported authority and limits the claim to the exercise. Students should practise making such corrections without turning a short report into a long disclaimer. The purpose is to let readers understand what happened without misunderstanding the evidence. This is clear language serving a precise public purpose.
35. Transfer task: make the final release decision
The fictional committee chooses to publish only the corrected educational display. It keeps the sample colour observations labelled as qualitative, the signal model labelled as hypothetical and the concentration estimate tied to its calibration range. It removes claims about certification, water treatment or real food safety. It records that the mistake was found before public printing and asks for an independent final check of units and wording.
This decision is stronger than either extreme of “publish everything” or “publish nothing”. The evidence is sufficient to teach some scientific and mathematical principles but not sufficient for the broader claims. A good decision states the allowed action, the supporting evidence, the withheld assertions and the conditions for release. The result is auditable by someone who did not participate in the original drafting.
36. Diagnose failures by discipline
If the poster says that green indicator colour means drinkable water, revisit Chemistry’s scope. If the signal of 22 is treated as 22 mg/L, revisit Mathematics’ calibration and units. If the correct number appears alongside the word “certified”, revisit English’s evidence and audience. If all three are correct separately but the final recommendation still calls for real-world use, revisit the committee’s decision authority.
This diagnostic habit makes revision efficient. Instead of assigning another generic EMS project, identify which reasoning step failed first. A student might know the equations but struggle to write a cautious conclusion, or write fluently while misunderstanding a graph. The right practice should target the missing link. Success in a combined learning task depends on strength in the component disciplines, not vague confidence.
37. Organise the work over four study sessions
Session one revisits Chemistry: students classify indicator results and explain what the test does not identify. Session two revisits Mathematics: derive the calibration, check the unknown estimate and calculate the stated interval. Session three revisits English: rewrite misleading captions and produce an accurate report for a named audience. Session four conducts the release audit and requires a decision with clear reasons.
Each session should produce something inspectable: a corrected inference, worked calculation, revised paragraph or decision ledger. Parents can ask the learner to explain the difference between “recorded”, “estimated” and “approved”. The answer reveals understanding better than asking whether the entire project was easy. If a distinction collapses, return to the relevant earlier volume instead of adding more loosely connected tasks.
38. Use the existing learner pathways
Continue through the Science Hub, Mathematics Hub and Secondary English Learning Hub for separate subject mastery. The PSLE Learning Guide helps primary pupils establish reading accuracy, recorded working and evidence-based answers before the G3 challenge.
Earlier EMS review chapters address other contexts, including Vol 1386 on simulated cold-chain delivery. This volume instead owns the distinction among an acid-base colour test, a calibrated numerical model and an editable public claim. Its purpose is not to recycle the delivery decision; it teaches how to audit evidence across a new measurement problem.
39. A final evidence-release checklist
Before approving the display, confirm that every quantity has a name and unit, every chemical observation is interpreted within the indicator’s scope, every model result states its assumptions, every graph has an honest scale, every incident report preserves chronology and every recommendation falls within the decision-maker’s authority. Check that no image or headline implies a real safety certification. These checks belong to responsible educational communication, not a claim of official approval.
The checklist can be used in other contexts: household-energy experiments, local-history exhibits, student surveys or school recycling projects. The names of the variables change, but the habits remain. State the question, identify the source, verify the method, inspect the inference, revise the claim and decide what the audience can be told. A concise, checked statement is more valuable than a confident statement with no source.
40. Final judgement
The committee approves the corrected fictional display, not a real-world water-use claim. It reports three qualitative indicator observations and a separate modelled concentration estimate of 3.4 mg/L from a signal of 22 under R = 5C + 5. It explains that the model was calibrated in a limited range and that neither exercise establishes real water safety. The English incident report records the error and its correction before publication.
The strongest lesson is that excellent performance does not mean forcing certainty out of incomplete information. Students should know when to calculate, when to explain a scientific mechanism, when to qualify a sentence and when to say a question remains unanswered. Those habits can strengthen performance across G3 English, Mathematics and Chemistry while respecting that each is separately assessed under the SEC.
Official syllabus and subject references
The SEAB 2027 G3 syllabus directory lists the official 2027 subject codes. Consult the original K300 English Language, K310 Mathematics and <a href="https://isomer-user-content.by.gov.sg/334/db643e6e-c6a2-4da8-9a29-f12b8cbadcc9/K324 Chemistry syllabuses for formal requirements. The examples here are invented study materials, not official specimen questions or real safety advice.