Three students are preparing the final recommendation for a fictional school field-edge project. The Biology student draws a persuasive food web and explains how plants could provide resources for insects and birds. The Mathematics student notices that the two pilot samples have the same mean insect-visit count and different levels of variation. The English student writes that the proposed habitat ‘has already proved a remarkable success’. Each has contributed something useful; together, they have also exposed a dangerous jump from possibility to proof. Their challenge is not to make the result sound bigger. It is to reach a decision that the evidence can bear.
This integrated English–Mathematics–Science (EMS) guide completes a deliberate four-volume sequence. Volume 1419 established the Biology of food webs, energy and conservation; Volume 1420 examined distributions, standard deviation and realistic survey constraints; Volume 1421 taught balanced G3 English Paper 1 argument and editing. Here we combine the skills in an original review-board simulation. All case material is invented for learning and does not describe a real school’s project or collected data.
Quick route: define the decision; label each source; separate biological mechanism from observation; recompute the numbers and units; test sampling limitations; identify the point at which an inference becomes too strong; repair public language; choose a reversible action within constraints; and record what evidence would justify a later update.
Three real G3 subjects, not an extra EMS examination
The Singapore–Cambridge Secondary Education Certificate begins in 2027. SEAB lists G3 English Language as K300, Mathematics as K310 and pure Biology as K325, with separate English, Mathematics and Biology syllabuses. ‘EMS’ is a study method for transferring knowledge across subjects; it is not an official SEC examination subject, paper or syllabus code. Students offering Combined Science should consult their actual combination’s syllabus. The exercises below do not predict an integrated SEC paper or replace practice in each distinct assessment format.
K300 English has Writing, Comprehension, Listening and Oral Communication. Its Paper 1 Writing lasts 1 hour 50 minutes; Sections A, B and C assess Editing, Situational Writing and Continuous Writing. K310 Mathematics has two written papers, each 2 hours 15 minutes, 90 marks and 50% weighting; its second paper includes a final real-world problem. K325 pure Biology uses a one-hour forty-mark multiple-choice paper, a one-hour-forty-five-minute eighty-mark structured/free-response paper, and a one-hour-fifty-minute forty-mark practical paper. Their different requirements explain why the integration exercise must return to specific subject skills rather than replace them.
The fictional evidence folder
The school board is considering one small planted strip beside the field instead of a larger immediate redesign. Folder A contains a proposed food web: plant → caterpillar → bird, plus notes about decomposers. Folder B contains a hypothetical five-period observation for current verge A: 2, 2, 3, 3, 4 visits, and test layout B: 1, 2, 3, 4, 4 visits. Folder C lists a $140 budget, a small monitoring plan priced at $90 and a larger one priced at $170. Folder D contains a draft notice: ‘The garden has doubled biodiversity, stored large amounts of carbon and proved that every unused space should be planted.’ No habitat has yet been installed in this fictional planning scenario, so that draft is especially misleading.
The board must choose among three actions. It can approve the full conversion now; approve a small, reversible trial with consistent monitoring; or decline the project entirely. The goal is not to choose the most enthusiastic option. It is to choose a feasible option that acknowledges the difference between a hypothesis and a result. More information may change the decision later. Scientific reasoning does not require absolute certainty, but it does require honest boundaries on what is known.
A disciplined evidence chain
Every recommendation should have four linked parts: source, calculation or mechanism, inference, and action. First describe the available record exactly. Next show the scientific concept or calculation used to interpret it. Then state what conclusion follows and where it is limited. Finally recommend a feasible step or further check. If one link is missing, the claim can sound plausible while remaining unsupported. A correct mean does not make the sampling unbiased. A sound food-chain diagram does not prove that organisms already occupy the proposed planting area. An elegant essay does not make a guessed fact true.
For the sample numbers, both groups have a total of fourteen visits across five periods, so both means are 2.8. Set A has a population standard deviation of approximately 0.748 visits; set B has approximately 1.17. Set B varies more within these invented observations. Neither sample proves that layout B supports more species or unique insects. The measurement is visits per interval, and the two groups may also have design limitations. This is a useful example of a valid mathematical comparison that must not be promoted into an unwarranted ecological conclusion.
Energy flow and conservation claims
Plants act as producers; consumers gain chemical energy by feeding; decomposers break down dead matter. Energy transfer between trophic levels is incomplete, while carbon atoms cycle among organisms and the environment. These concepts justify possible mechanisms to investigate at a planted edge. They do not justify a carbon-storage total or guaranteed increase in biodiversity without measurements. Even a sound calculation of trophic efficiency based on illustrative energy values is a model, not a measurement of this school. The public statement must distinguish mechanism, estimate and observation.
The board should also distinguish environmental aspirations from measured human constraints. A project can be worthwhile as a learning experiment without already being a demonstrated ecological success. The school must preserve safe access, choose appropriate plants and avoid needless disturbance of wildlife. A recommendation that treats accessibility or maintenance as unimportant has not become more scientific; it has merely narrowed the evidence to the part that supports its preferred answer.
Thirty-six integrated audit exercises
The following case files are organised from straightforward checking to increasingly demanding judgement. In each, read the claimed fact, identify the missing link, propose an accurate replacement and decide what to check next. These are original learning exercises, not official examination questions. The detailed answer after each case demonstrates how to carry reasoning across subjects without confusing their formal assessments.
Review-board case 1: An arrow with the wrong meaning
Questionable evidence or claim. A poster shows bird → caterpillar → plant and says the arrows identify what each animal hunts.
Audit finding. For energy transfer, reverse the arrows to plant → caterpillar → bird. The original poster reverses food-to-feeder relationships. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Rebuild the scientific diagram before any public explanation; otherwise a grammatically clear sentence still teaches incorrect Biology. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 2: An unsupported consumer
Questionable evidence or claim. The food web features a bird species that has never been observed at the school.
Audit finding. Mark the bird as a plausible example rather than a recorded inhabitant. Food-web possibility is not site evidence. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Carry a ‘hypothetical’ label through every chart and caption derived from this diagram. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 3: Energy and matter
Questionable evidence or claim. The brochure says that plant energy cycles forever through birds and back to leaves.
Audit finding. Energy is transferred and becomes less available at successive trophic levels; carbon and other matter are cycled. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Use separate diagrams for energy flow and carbon transfer to avoid blending the mechanisms. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 4: Energy efficiency
Questionable evidence or claim. Producers contain 1,500 kJ and primary consumers 120 kJ in an illustrative model.
Audit finding. Efficiency = 120/1500 × 100% = 8%, for that specified transfer. The remaining energy is not destroyed. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. The English caption must say ‘in this model’, not ‘measured at our school’. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 5: Biomass and abundance
Questionable evidence or claim. A tree supports many caterpillars; the count pyramid is narrow at the producer level.
Audit finding. Numbers depend on individuals; biomass measures living mass. One tree can outweigh many smaller consumers. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Name which pyramid is being interpreted before reaching any conclusion. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 6: A carbon promise
Questionable evidence or claim. A leaflet claims twelve proposed seedlings have already trapped a quantified mass of carbon.
Audit finding. No seedlings have been installed or measured. Photosynthesis can lead to carbon storage, but the project-specific amount is unknown. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Replace the quantity with an evidence-based goal, such as monitoring growth and survival. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 7: A false biodiversity doubling
Questionable evidence or claim. Five observation periods produce fourteen insect visits in both two groups.
Audit finding. The means are equal at 2.8 visits per period. They do not even show a doubling of visits, much less species richness. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Remove the doubling claim and specify the actual measured quantity. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 8: The spread difference
Questionable evidence or claim. Two visit datasets have standard deviations about 0.748 and 1.17.
Audit finding. The second sample is more variable on this measure. It does not necessarily have fewer or more unique insects. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Use the mathematical comparison as a description of distribution, not a verdict on habitat quality. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 9: Unspecified units
Questionable evidence or claim. A chart says one site averaged 2.8 but omits the unit.
Audit finding. Write 2.8 recorded visits per observation period, and state how long the period lasted if known. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. A number without a unit invites readers to mistake events for organisms or minutes. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 10: A different observation window
Questionable evidence or claim. One team counted for 30 minutes and another for 60.
Audit finding. Normalise to a common time rate before comparing, and still check observation area and protocol. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Comparable rates on one denominator do not guarantee a fair ecological experiment. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 11: Rain interrupts records
Questionable evidence or claim. Three planned observations were cancelled during rain.
Audit finding. Do not fill cancelled periods with zeros. State valid periods and reasons for missing data. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. The missingness may be related to the outcome, so report it explicitly. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 12: The single best day
Questionable evidence or claim. A school report chooses the most active survey day as typical.
Audit finding. A maximum is not the mean or median; it is only one extreme observed value. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Show the full distribution or a fair summary rather than selecting only a flattering day. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 13: A faulty bar chart
Questionable evidence or claim. A graph starts at 13 and ends at 16, making a small difference look huge.
Audit finding. Explain the truncated axis and consider a chart with a clear scale and labelled counts. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. The visual design must not intensify a conclusion beyond the underlying data. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 14: Bigger sample, different method
Questionable evidence or claim. A larger sample was gathered by more experienced observers using a different identification guide.
Audit finding. Sample size has increased, but measurement method also changed. Differences may reflect identification rather than ecology. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Keep records of method and observer training; compare like with like or state the limitation. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 15: Wrong population
Questionable evidence or claim. A survey records visible insects but a notice refers to all animals at the site.
Audit finding. The observed group is narrower than the alleged population. Unknown organisms were not counted. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Correct the claim’s subject before editing its tone. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 16: A leading question
Questionable evidence or claim. Students ask visitors whether they support the ‘wonderful new wildlife garden’.
Audit finding. The question may influence answers. Use neutral wording such as ‘Do you support the proposed planting trial?’ Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Survey language is part of evidence quality, not merely English style. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 17: A weak school objective
Questionable evidence or claim. The board’s sole aim is to create the most attractive photograph.
Audit finding. A scientifically meaningful goal would define what to learn, what to observe and what constraints to preserve. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. A visually attractive outcome and a defensible environmental outcome are different goals. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 18: Budget arithmetic
Questionable evidence or claim. A small pilot costs $90, a large programme $170 and the budget is $140.
Audit finding. The small pilot fits and leaves $50; the larger plan exceeds the budget by $30. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. State the feasible option before considering ecological effectiveness. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 19: Hidden maintenance
Questionable evidence or claim. The $90 pilot excludes an unpriced maintenance activity.
Audit finding. The known cost is $90, but total future cost is not established. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. A good decision states that remaining funds are provisional until the omitted cost is assessed. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 20: Ignored access
Questionable evidence or claim. The proposed bed crosses the normal route to a gate.
Audit finding. The ecological attraction is not sufficient to establish site feasibility. Routes and access must be checked. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. The recommendation should preserve safety and function even if the Biology model is sound. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 21: Overconfident introduction
Questionable evidence or claim. The announcement begins, ‘Our project has solved biodiversity loss.’
Audit finding. That sentence asserts global and site-specific outcomes not supported by any measurement. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Begin with the actual status: ‘The school is considering a monitored planting trial.’ Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 22: Misleading statistical verb
Questionable evidence or claim. A draft says the numbers prove the garden worked, but the project has not begun.
Audit finding. The numbers are hypothetical pilot examples, not results of an implemented intervention. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Use ‘illustrate’ for training data and ‘show’ only for genuinely observed data. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 23: Missing time boundary
Questionable evidence or claim. A report says insect visits increased permanently.
Audit finding. Without long-term monitoring, persistence cannot be established. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. State the observation dates and avoid ‘permanent’ unless there is appropriate evidence. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 24: One photo, many claims
Questionable evidence or claim. A photograph of a bird beside flowers is used to claim breeding success.
Audit finding. It can at most support the observation of the bird at the photographed location and time. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Breeding, feeding and nesting require separate observations; choose accurate verbs. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 25: Failure to distinguish seed and visit
Questionable evidence or claim. More insect visits are reported, and the caption says seed production increased.
Audit finding. Visits do not directly measure pollination success or seed production. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Keep the mechanism as a hypothesis and identify the outcome that would need recording. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 26: Food web resilience
Questionable evidence or claim. Multiple feeding links appear in a model, so the caption promises a stable ecosystem.
Audit finding. Alternative links can reduce dependence on one prey, but stability depends on population dynamics and other conditions. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Write a conditional mechanism and avoid an absolute forecast. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 27: Ecological harm not addressed
Questionable evidence or claim. A proposal suggests introducing an unfamiliar species to attract birds quickly.
Audit finding. Introducing organisms without appropriate assessment can create ecological risk. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Choose observation and suitable habitat management over reckless release. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 28: Observation effort changes
Questionable evidence or claim. After planting, the school doubles the number of observation hours and reports more sightings.
Audit finding. More effort alone may raise the total count. Compare equal-effort records or use transparent standardisation. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Report exact hours rather than hiding the changed method behind a percentage. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 29: Evidence versus opinion
Questionable evidence or claim. One student says the verge looks better, so the treatment is scientifically superior.
Audit finding. Appearance is a subjective judgement; ecological performance needs appropriate observations and agreed measures. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Acknowledge aesthetic value separately rather than disguising it as a scientific result. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 30: Unfair counterargument
Questionable evidence or claim. The essay calls access concerns ‘anti-environmental’.
Audit finding. Access and safety are legitimate constraints; the writer should answer them respectfully with a feasible design. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. The quality of an argument depends partly on treating the opposing case fairly. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 31: Grammar changes certainty
Questionable evidence or claim. An editor changes ‘may improve’ to ‘will improve’ to make the sentence stronger.
Audit finding. The revision changes the truth conditions, not merely its style. Without evidence, ‘may’ better reflects uncertainty. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Editing must preserve warranted certainty rather than sharpen claims by removing qualifiers. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 32: A hidden causal confounder
Questionable evidence or claim. Two survey groups differ in sunlight as well as planting.
Audit finding. If both conditions vary, the result cannot be uniquely attributed to planting. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Name at least one alternative explanation and propose a stronger comparison. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 33: A selective public chart
Questionable evidence or claim. The announcement displays the mean but omits sample size and spread.
Audit finding. Readers may infer stability that is not supported. Include observation count and an appropriate measure of spread. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Transparent presentation can be concise without withholding context essential to interpretation. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 34: A correction policy
Questionable evidence or claim. A figure in the draft notice is found to have the wrong denominator.
Audit finding. Recalculate from source records, amend the statement clearly and record what changed. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Trust is strengthened by prompt correction, not by quietly preserving a claim after its basis fails. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 35: Evidence threshold for expansion
Questionable evidence or claim. The board wants to expand after a single observation day.
Audit finding. A better threshold uses repeated comparable sessions, plant survival, cost and access review. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Set success criteria before seeing the result to reduce selective interpretation. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
Review-board case 36: Final recommendation
Questionable evidence or claim. The board has a plausible ecological mechanism, illustrative statistics, a $140 budget and unresolved access questions.
Audit finding. Approve only a small conditional trial after access checks, within a verified budget and with consistent monitoring. Do not claim success in advance. Trace the statement back to its actual source. In integrated reasoning, a sound numerical answer and a sound scientific explanation may still be insufficient if they do not address the same measured quantity.
Decision-ready correction. Give the reader an action, a limit and a review condition. This is stronger than a slogan. Before signing off a statement, check whether its subject, unit, time boundary and degree of certainty match the record. State which conclusion is supported now and what additional evidence would justify a broader statement later.
A model board decision with explicit conditions
The board should not approve a claim of demonstrated biodiversity improvement. The food-web diagram identifies possible relationships between plants, insect consumers and birds, but does not demonstrate which organisms are present or how their populations respond to planting. The two miniature datasets have equal means, and the difference in their standard deviations describes variation in visits, not a difference in species diversity. The data are teaching examples, not verified observations of the proposal. Their use in a public statement must remain clearly labelled.
A limited planting trial may nevertheless be a reasonable learning project. Its scope should protect essential access and keep the proposed cost within the available budget, including provision for maintenance. The school should define its observation units in advance and make future comparisons with consistent time, area and method. It should record weather and other factors that might affect detection. Its public statement can say it is trialling habitat provision and learning how to monitor it, without promising results that do not yet exist.
At a scheduled review, the board should ask whether plants survived, whether access remained safe, whether monitoring took place as planned, and what the observations support. The next decision could be to maintain, adjust, expand or stop the trial. A useful report gives the measured result, the method, the limit and the response. A correction should be issued if an earlier headline proves inaccurate. That is an accountable way to use evidence.
Practice plan across separate SEC papers
For Biology, rehearse the food-chain direction, energy-transfer explanations, carbon-cycle statements and interpretation of unfamiliar observations. Use the specific learning outcomes and practical skills in K325, not an invented EMS paper. For Mathematics, calculate a mean and standard deviation with visible working, analyse an appropriate graph and finish with a decision that respects all stated constraints. In K310 Paper 2, train yourself to identify useful mathematics in a long real-world problem. For English, write a 350–500-word continuous response to a new proposition; then revise the position, paragraph logic, counterargument and grammar separately. Each subject needs direct examination practice as well as cross-subject transfer.
For a thirty-minute home routine, spend ten minutes on a Biology explanation from unseen data, ten minutes on a Mathematics comparison from a fresh dataset, and ten minutes rewriting a misleading claim into precise English. For a longer session, add a complete short essay plan and a verbal defence of the recommendation. The aim is not to memorise this invented field-edge story. It is to notice what kind of evidence a new question supplies, choose the right tool, and avoid making the conclusion larger than the available record.
Learning links and sources
Begin with Volume 1419: ecological mechanisms and conservation, continue through Volume 1420: survey statistics and fair comparisons, and use Volume 1421: balanced Paper 1 writing. The previous sports-day EMS audit in Volume 1418 teaches a related but different problem about materials recovery. For subject routes, see Secondary English, Secondary Mathematics, and PSLE-to-Secondary Science. The PSLE-to-Secondary 1 transition guide supports earlier learners.
Official source check: the SEAB 2027 G3 school-candidate directory and the K300 English, K310 Mathematics and K325 pure Biology syllabuses, as available in November 2026. Examination formats and subject combinations should always be checked for the candidate’s actual examination year.