A school display says, “Everyone can read faster when looking farther away.” The presenter points to a small table, a picture of an eye and a recording of a teacher explaining the activity. There seems to be evidence from Science, Mathematics and English. But a claim does not become true just because three subjects appear on the same poster. Each source has a different job, and each has limits. The Biology explains possible mechanisms of focus and light response. The Mathematics describes a set of matched completion times. The English listening report tells us what the speaker actually said, including corrections and exclusions.
This guide teaches you to conduct an evidence hearing before a school exhibition, project report or examination response. You will sort facts from interpretations, verify arithmetic, restore omitted conditions and decide whether a sentence can be published as it stands. The six-person numerical example is invented for learning; it is not a study of students’ eyesight. It cannot diagnose a visual condition, and an ordinary classroom reading activity should never be described as a medical screening procedure.
**Quick answer:** identify the exact public claim; name the evidence offered for it; check whether the Biology mechanism matches the variable; verify the Mathematics on complete matched records; restore any correction or missing-data condition from the spoken report; then approve, qualify or reject the claim with a short explanation. The final sentence should be no stronger than the evidence supporting it.
Contents
- Where this volume belongs in the G3 SEC sequence
- The 2027 examination reference point
- The hearing’s fictional case file
- A claim needs a defined population
- Biology gate: pupil and lens are not the same mechanism
- Mathematics gate: keep each participant paired
- Mathematics gate: centres and distributions tell different stories
- Missing data gate: do not fill the silence with zero
- English listening gate: which number survived correction?
- A proper evidence ledger
- Approve, qualify or reject
- Causation requires more than agreement across subjects
- Display wording is part of scientific accuracy
- The exam answer is not a full research paper
- Comparing two different versions of the report
- Follow the route from PSLE foundations to SEC judgement
- 30 public-claim hearing exercises
- Four-week mastery plan
- Frequently asked questions
Where this volume belongs in the G3 SEC sequence
The learning route has moved from recognising a response to evaluating a whole explanation. Vol 0787 separated accommodation from the pupil reflex. Vol 0788 worked through six matched reading-time records and missing entries. Vol 0789 trained listening for participant labels, revised numbers and exclusions. This fourth volume asks how to use all three without claiming that their combination proves more than any one of them can establish. It is an EMS transfer exercise, not a separate examinable subject called EMS. In the actual SEC, English, Mathematics and individual Science subjects have their own syllabuses, paper formats and marking demands.
The 2027 examination reference point
Singapore’s new Singapore–Cambridge Secondary Education Certificate begins with the 2027 examination cohort. In the 2027 G3 syllabuses, English Language is K300, Mathematics is K310, and pure Biology is K325; Combined Science combinations have their own codes. Do not assume an older O-Level code is the new SEC code. Official English K300 comprises writing, comprehension, listening and oral components. Mathematics K310 has two written papers, each 90 marks and 50% of the subject grade. Pure Biology K325 has multiple-choice, structured/free-response and practical components. This interdisciplinary clinic supports reasoning in those subjects, but it is not a claim that SEAB sets one integrated EMS paper. Students should verify their registered combination and paper requirements against the SEAB 2027 G3 syllabus index.
The hearing’s fictional case file
Six fictional participants, A to F, complete one near-card reading task and one distant-card reading task. The near times in seconds are A 18, B 20, C 22, D 24, E 26, F 30. The corresponding distant times are A 16, B 19, C 19, D 23, E 25, F 28. These values are constructed solely for arithmetic and interpretation practice. They are not measurements from a clinical study. The case also includes a teacher’s spoken note, a draft poster, and a simplified account of the eye. The spoken note later corrects one record and explains a missing result in a revised version of the data. A responsible reviewer must know which version of the record is being evaluated before drawing a conclusion. Mixing the original complete example with an updated incomplete table will produce misleading arithmetic.
A claim needs a defined population
The sentence ‘all six participants in the fictional complete-pair table took less time on the distant card’ can be checked against six pairs. The sentence ‘every student everywhere will always read faster at a distance’ cannot. The difference is not pedantic: the first claim names a specific set and an observed outcome; the second adds an enormous population and an absolute prediction. A school display should describe only the participants and conditions actually represented. Even six consistent observations do not establish a general rule for every learner, and small samples provide a poor basis for far-reaching claims. Look for words such as all, always, guarantees, proves and never. These are warning signals to check whether the evidence has the breadth the sentence implies.
Biology gate: pupil and lens are not the same mechanism
A pupil controls how much light enters the eye through iris muscle action. Near–far accommodation depends on the ciliary muscle, suspensory ligaments and changes in lens shape that help focus an image on the retina. When the object becomes nearer, the ciliary muscles contract, tension in the suspensory ligaments falls and the lens becomes thicker or more convex. When the object becomes farther, the ciliary muscles relax, ligament tension increases and the lens becomes thinner or less convex. A school claim that a smaller pupil directly makes the lens thinner confuses mechanisms. The correct response identifies the changed condition and traces only the relevant chain. A photograph of pupil diameter cannot show lens curvature or alone establish whether accommodation was effective.
Mathematics gate: keep each participant paired
The complete fictional table gives near-minus-distant differences for A through F of 2, 1, 3, 1, 1 and 2 seconds. Their total is 10 seconds; the mean paired difference is 10 divided by 6, about 1.67 seconds. All six signs are positive under this convention, indicating that the distant task was completed in less time for each participant in that constructed example. This is not the same as saying that distance caused the difference. To preserve a matched design, compare A with A and B with B. If someone sorts both columns separately and then subtracts, the resulting differences may no longer represent real pairs, even if the two column means remain arithmetically unchanged.
Mathematics gate: centres and distributions tell different stories
The near times total 140 seconds and have a mean of about 23.3 seconds; the distant times total 130 seconds and have a mean of about 21.7 seconds. The near median is 23 seconds and the distant median is 21 seconds. But the median of the six paired changes is 1.5 seconds, not the 2-second difference between the two marginal medians. Those are different statistics. The interquartile range of each column is 6 seconds under the median-of-halves method used in this learning example. Equal IQRs do not make the data identical. A reviewer must say whether a number is a mean time, a mean change, a median time or a median change. Never label every statistic simply ‘the average improvement’.
Missing data gate: do not fill the silence with zero
Now imagine the teacher announces a revised recording sheet in which E’s distant reading was never recorded. That revised sheet has only five complete pairs: A, B, C, D and F. Their near-minus-distant differences are 2, 1, 3, 1 and 2 seconds, summing to 9 seconds. The complete-pair mean is 9 divided by 5, or 1.8 seconds. The claim ‘all six improved’ is no longer warranted by the revised sheet because E has no second observation. The correct conclusion is about the five available complete pairs. A blank cell does not equal a measured zero; assigning zero would fabricate a result and radically distort the average.
English listening gate: which number survived correction?
A teacher might say, ‘I first reported D as twenty-one seconds, but the corrected distant time is twenty-three.’ A learner must record 23 seconds as the final value for D; 21 was superseded. Another sentence might say, ‘Six students completed the near task, but only five completed the distant task.’ The second count matters even though the first number is more memorable. English Paper 3 assesses whether a listener understands speaker meaning, including corrections and conditions. For 2027 G3 English K300, Section A recordings are heard twice, whereas the informational text in Section B is heard once. During a hearing, do not assume every fact will be repeated; use concise labels and listen for not, only, except, actually and the final confirmed figure.
A proper evidence ledger
A useful ledger has five conceptual columns: proposed claim, source, actual observation, inference and limitation. You do not need to draw a complex table in every examination, but the structure helps you think. For example, the proposed claim ‘pupil size caused the time difference’ is associated with the Biology explanation; yet no direct pupil measurements appear in our reading-time table. The actual observation is completion time, not pupil size. Thus the causal inference is unsupported by the supplied numbers. Another claim, ‘the six complete pairs have a mean difference of about 1.67 seconds’, is directly supported by arithmetic from the first fictional table. Always identify the correct version of the table before doing the calculation.
Approve, qualify or reject
An evidence hearing needs three possible outcomes. Approve when wording directly and accurately describes the supported observations, with correct units, population and conditions. Qualify when a modest correction makes the claim defensible: replace ‘every student’ with ‘the six participants in this illustrative dataset’, or add ‘in this activity’. Reject when a claim invents a result, confuses a mechanism or diagnoses health from classroom completion times. This is not a debate about whether the writer sounds confident. It is a comparison between the wording and the evidence. Explain the reason in one or two sentences and provide an accurate replacement if one is possible.
Causation requires more than agreement across subjects
Imagine that the distant card was also larger, the room was brighter, and students attempted the distant task second. Any of these features could affect task completion. The Mathematics can still calculate the observed differences, and the Biology can still explain accommodation in general, but the combined evidence does not isolate a cause of faster completion. A good improvement plan would hold target design and lighting comparable and vary trial order appropriately while preserving the same participants. The report should state what was controlled and how, not merely assert that the test was fair. A small classroom task cannot establish a medical conclusion, even if its numbers appear orderly.
Display wording is part of scientific accuracy
A poster headline is not a licence to remove conditions for dramatic effect. Compare ‘Our class proved that distant vision is better’ with ‘In six fictional paired observations, the distant-card completion times were lower; this exercise does not identify the cause’. The second is less exciting but much more precise. A reader can see the population, variable, result and limit. Use plain English, not hedging piled on hedging. Phrases such as ‘in this record’, ‘under these conditions’ and ‘the data do not show’ help the reader separate what was established from what remains unknown. Clear public language is a form of evidence discipline.
The exam answer is not a full research paper
An integrated case can be reduced to a sequence that fits a short-answer question: identify the error; quote or calculate the relevant evidence; explain why it matters; state the corrected conclusion. For Biology, use the right structures and mechanism. For Mathematics, show the matched subtraction, denominator and units. For English, attribute the speaker’s correction and preserve the qualifying words. A lengthy account of every feature of the eye is not useful when the question asks about one conclusion. Equally, a correct calculation without context does not answer an evaluation question. Let the command word determine the amount of working and explanation.
Comparing two different versions of the report
The first teaching table contains all six complete pairs and supports a mean paired difference of approximately 1.67 seconds. The revised version declares E’s distant value missing and supports a mean of 1.8 seconds among five complete pairs. Neither answer is an error when its population and data version are named correctly. Trouble begins when a presenter announces ‘1.8 seconds for six students’ or mixes the first table’s count with the revised table’s mean. In a public exhibition, version labels and correction notes are part of the evidence. If the recording changed, update the caption or clearly identify which dataset the caption describes.
Follow the route from PSLE foundations to SEC judgement
Primary learners first practise noticing the stimulus, reading a table and answering exactly what was asked. Upper-secondary SEC work demands a more reliable transfer: choose an appropriate mechanism, keep paired observations intact, interpret a spoken correction, and judge the scope of a public claim. The PSLE stimulus-response guide rebuilds the scientific chain. The PSLE Mathematics word-problem route reinforces what a number refers to. The PSLE English revision route builds careful attention to wording. Each foundation remains useful as the questions grow harder.
Thirty original evidence-hearing cases
For each case decide whether the proposed claim should be approved, qualified or rejected. The short verdicts explain how one detail changes the logic. All numbers and people in this guide are fictional, and the examples are not copied from SEC past papers.
Hearing 1: Headline: ‘Pupils shrink when the student looks farther away.’
Verdict and reason: Reject as a biological explanation of accommodation. The main near–far change is lens curvature produced through ciliary muscles and suspensory ligaments. Pupil diameter responds especially to light and cannot be used as a direct substitute for lens-shape evidence. A replacement should name the viewing-distance condition and relevant mechanism. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 2: Headline: ‘The retina changes shape to focus an image.’
Verdict and reason: Reject. In the school-level accommodation chain, lens shape changes and the image is focused on the retina. The retina receives light-sensitive input; it is not the adjustable lens. A repair would state that the ciliary muscles and suspensory ligaments change the lens curvature according to viewing distance. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 3: Caption: ‘All six complete pairs had shorter distant times.’
Verdict and reason: Approve for the original six-person fictional dataset. Compare each participant’s near and distant values; every distant time is less. Preserve the words ‘all six complete pairs’ and the fact that the example is constructed. Do not extend the conclusion to all students or to eyesight quality. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 4: Caption: ‘The mean paired change was 1.67 seconds.’
Verdict and reason: Approve for the original complete table, if the difference is defined as near minus distant and the value is rounded appropriately. The six differences sum to 10 seconds. Specify that this is a mean change in task completion time, not a change in pupil diameter or accommodation strength. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 5: Caption: ‘The mean change for the five complete revised pairs was 1.80 seconds.’
Verdict and reason: Approve for the revised sheet where E’s distant record is missing. The five valid differences sum to 9 seconds, giving 1.8 seconds. The denominator and dataset version are essential. A stronger statement about all six participants would be unsupported. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 6: Caption: ‘E recorded a distant time of zero.’
Verdict and reason: Reject for the revised sheet. The note says E’s distant time was not recorded. Zero would be a numerical observation that was never supplied, and could imply instantaneous task completion. Replace with ‘E had no recorded distant-task time, so the paired calculation used five complete records’. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 7: Audio: ‘D took twenty-one seconds—sorry, twenty-three.’
Verdict and reason: Use 23 seconds as D’s final corrected distant time if that is the source recording under review. The number 21 is an earlier mistaken report. The listening skill is to recognise the repair and attach the final number to D, not to copy two values as separate trials. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 8: Audio: ‘Only five students finished both tasks.’
Verdict and reason: Record five complete pairs for that version. The word ‘only’ restricts the count. It does not mean five students attended the entire activity or that the absent sixth result was zero. If the display retains a six-person mean, the presenter must reconcile which version was used. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 9: Poster: ‘Equal interquartile ranges mean equal performance.’
Verdict and reason: Reject. In the complete fictional set, both columns have an IQR of six seconds using the stated method, but their centres differ and individual changes remain different. Equal IQRs describe comparable middle-half spread, not identical distributions or the same performance. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 10: Poster: ‘The median difference is two seconds.’
Verdict and reason: Qualify or reject according to what ‘difference’ means. The near median is 23 seconds and the distant median is 21, a difference of two seconds. However, the median of the six paired differences is 1.5 seconds. A responsible caption names which calculation it reports. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 11: Question: ‘Do the numbers prove that distance causes faster reading?’
Verdict and reason: Reject. Completion times differ under the fictional conditions, but possible confounding factors include target size, illumination, prior familiarity, attention and trial order. State that the data show an association within the constructed task, not a unique explanation or universal biological effect. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 12: Question: ‘Can a pupil photograph establish lens accommodation?’
Verdict and reason: No. A pupil photograph may display pupil diameter under those conditions but does not directly show lens curvature. Explain the different functions of iris and lens and identify what direct evidence is missing. Do not treat appearance alone as proof of the accommodation process. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 13: Claim: ‘A decreased by eleven per cent.’
Verdict and reason: Qualify the comparison. Participant A decreased from 18 seconds near to 16 seconds distant, a decrease of 2/18 times 100, or approximately 11.1% relative to the original near time. Without a named reference and unit the percentage is incomplete. It reports task time, not biological capacity. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 14: Claim: ‘A’s near time exceeded the distant time by 11.1%.’
Verdict and reason: Reject that particular base. A’s near time is 18 seconds and distant time 16 seconds; the amount by which near exceeds distant is 2/16 times 100, or 12.5%. Eleven point one per cent uses 18 as the base for a decrease from near to distant. The wording determines the denominator. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 15: Speaker: ‘The demonstration guarantees clearer sight.’
Verdict and reason: Reject. A classroom reading-time task cannot guarantee a visual-health outcome. A safe replacement says the demonstration practises interpreting accommodation, pupil responses and reading-time data. A real concern about a person’s vision belongs with a qualified eye-care professional, not a school poster. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 16: Speaker: ‘The light and card distance were changed together.’
Verdict and reason: Flag a fair-test limitation. Two explanatory conditions changed at once. You cannot attribute any observed time difference solely to lighting or solely to viewing distance. Suggest a design that controls one while deliberately varying the other, with safe ordinary illumination. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 17: Speaker: ‘We gave the distant task second to every participant.’
Verdict and reason: Identify order as a possible confounder. Students might practise, become familiar with the text or tire between tasks. A design might vary or counterbalance order with comparable tasks. The mathematical differences remain calculable, but the causal claim requires more than correct subtraction. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 18: Display shows two bar-chart means with a truncated axis.
Verdict and reason: Qualify the visual interpretation. A truncated scale can magnify a modest numerical difference. State the actual mean values, about 23.3 seconds near and 21.7 seconds distant in the complete example, and avoid saying the graph alone proves a dramatic improvement. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 19: Display says ‘Distant mean = 21.7%’.
Verdict and reason: Reject the unit. The mean distant completion time in the complete table is approximately 21.7 seconds, not a percentage. A percentage requires a reference value and comparison. Fix the unit before interpreting what a reader might think the number means. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 20: Committee note: ‘We used all six near results and five distant results for a matched average.’
Verdict and reason: Reject the calculation method as described. A matched-pair mean requires both measurements for each included participant. Use five complete pairs or explicitly analyse the two marginal sets separately, naming their different membership. Do not call a mixed-membership difference of means a matched mean. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 21: A presenter quotes Biology to claim ‘near means bright’.
Verdict and reason: Reject the implied equivalence. Viewing distance and illumination are distinct conditions. A near object can be dim and a distant object can be bright. Accommodation chiefly addresses focusing for distance while the iris controls light entry. The poster must not treat the two descriptors as interchangeable. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 22: The final corrected report is available, but the draft uses the old recording.
Verdict and reason: Request correction before publication. Determine which observation the revised record superseded, update the affected table entry and recompute any dependent statistic. A neat caption is not adequate if it uses superseded numbers. The final source record takes priority over an earlier draft. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 23: A summary omits the word ‘only’ before ‘five completed pairs’.
Verdict and reason: Require the qualifier to be restored. ‘Five completed pairs’ is a count; ‘only five of six participants completed both’ describes the exception and helps explain the missing case. In English, a small omitted word can change the scope of an otherwise accurate statement. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 24: A description says ‘The pupil is the light-sensitive layer at the back of the eye.’
Verdict and reason: Reject. The pupil is the opening controlled by the iris; the retina is the light-sensitive layer lining the inside back of the eye. Correct anatomical identity before evaluating any causal claim, because a wrong structure can make the rest of an explanation incoherent. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 25: A table labels the change ‘distant minus near’ but prints positive 2 for A.
Verdict and reason: Flag a sign error. For A, distant minus near is 16 − 18 = −2 seconds. Positive 2 belongs to near minus distant. Either convention is acceptable when stated consistently; the sign tells readers the chosen direction and must match the arithmetic. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 26: A caption uses the median of paired changes but calls it a mean.
Verdict and reason: Correct the statistic name. The paired changes sorted are 1, 1, 1, 2, 2, 3 seconds, giving median 1.5 seconds, while their arithmetic mean is about 1.67 seconds. The distinction matters because the two measures answer different questions. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 27: A school flyer says ‘Test your eyesight for free using our card game’.
Verdict and reason: Reject the clinical implication. A school demonstration with reading cards is not a medical vision test. Recast it as a learning exercise about biological mechanisms and evidence interpretation, not an assessment or guarantee of a participant’s vision. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 28: The committee is shown only a photocopy of a corrected number.
Verdict and reason: Ask for the context and final source identification. A single number cannot tell whether it belongs to A, D or E, or whether it is near or distant, seconds or minutes, preliminary or final. A reliable ledger attaches the value to a label, condition and version. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 29: An examiner asks: ‘What is one evidence limitation?’
Verdict and reason: Choose a specific limitation and explain its effect. For example, if font size and viewing distance changed together, the trial cannot isolate distance as the cause. A generic statement ‘the results may be unreliable’ is weaker because it does not identify what threatens the inference. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Hearing 30: A final presentation sentence says ‘The fictional five-pair revised record has a mean near-minus-distant difference of 1.8 s; this cannot diagnose vision.’
Verdict and reason: Approve. The statement uses the right revised denominator, statistic, direction, unit and limitation. It avoids extending the finding to people outside the constructed record. A short correct sentence can carry a complete judgement without a dramatic headline. Independent check: Identify the evidence source, the measured quantity and any missing condition. Decide whether the language is narrower than, equal to or broader than what the record permits. A correct correction names the exact part of the claim that changed.
Four weeks from foundations to integrated examination judgement
Week 1 — Build the evidence gates
Review the eye structures and the two different mechanisms. Rewrite near-focus and far-focus chains accurately, then explain what an iris-controlled pupil change does. Practise identifying whether a short scenario changes viewing distance, illumination or both. Use ordinary written examples rather than unsafe bright-light experiments. End the week by correcting five misleading captions and stating exactly what each replacement can claim.
Week 2 — Reconcile the two numerical versions
Recreate the fictional six-pair data set and calculate the mean and median of the paired differences. Then use the revised five-complete-pair version, identify which participant is excluded, and recompute the appropriate mean. Show units, denominator and direction. Add one deliberately misleading graph and explain how its axis affects interpretation. Repair arithmetic mistakes before writing a public sentence.
Week 3 — Listen and restore the conditions
Use short original recorded announcements or have a tutor read a changing report aloud. Capture participant IDs, revised figures, exceptions and final decisions. Some practice should be one-hearing to build the official English Section B skill; some can use two hearings for Section A-style verification. Compare your notes with the transcript after the attempt. Record whether the first missed link was a number, a negation, a speaker or a correction.
Week 4 — Run the hearing independently
Attempt the thirty cases here without reading the verdicts first. For each, write ‘approve’, ‘qualify’ or ‘reject’ followed by a reason. Then conduct a fresh invented case with different quantities, a corrected announcement and an overconfident headline. Keep the final report concise and reader-safe. Revisit the PSLE-to-secondary study plan if the task reveals missing foundations. Your goal is independent evidence judgement, not memorisation of this particular demonstration.
Frequently asked questions
Is EMS an official 2027 SEC paper?
No. EMS here means an integrated learning exercise that transfers English, Mathematics and Science reasoning. Candidates sit the separate syllabus papers for their registered subjects, not a new unified EMS paper.
Can I use the original mean of 1.67 seconds after E’s record is missing?
Only if you explicitly refer to the original six-complete-pair fictional dataset. In the revised five-complete-pair version, the mean is 1.8 seconds. Do not mix the counts and values.
Why is a claim about eyesight unacceptable?
The fictional case measures task completion times, not a clinical eye condition. A classroom exercise does not establish a diagnosis or guaranteed health outcome.
What does ‘approve, qualify or reject’ mean?
Approve directly supported wording; qualify a claim whose scope, population or method needs correction; reject an unsupported, false or misleading statement. Always give a reason tied to the supplied evidence.
How should I use the three subjects together?
Use Biology for mechanisms, Mathematics for records and calculations, and English to interpret the wording and spoken corrections. Keep the role of each source clear before combining conclusions.
What if the speaker corrects a participant’s number?
Replace the superseded value in the appropriate record and recalculate dependent results. Keep the original only if the question asks what was changed.
Is a smaller pupil proof that the lens focused near objects?
No. Pupil diameter and lens accommodation are different mechanisms. The observation alone does not establish lens shape or its focusing success.
How do I revise without memorising this particular case?
Practise the same reasoning on new small tables, diagrams, announcements and public claims. The case facts should change; the habits of checking source, units, population, mechanism and evidence limits should remain.
Your next reading route and official references
Return to the three owners in this batch: Biology Vol 0787, Mathematics Vol 0788, and English Vol 0789. Use the Secondary 3 Science owner, Mathematics statistics owner and English listening and discussion owner for wider preparation. Official documents: 2027 G3 Biology K325, 2027 G3 Mathematics K310, 2027 G3 English K300 and the SEC overview. This site’s learning exercises are independent and not endorsed by MOE or SEAB.
