The question is whether the evidence supports the announcement
The habitat inventory project has completed a drawing of an optical path, a calculation of possible scan outcomes and a reading of two notices. It is tempting to assemble the three pieces into one triumphant announcement: every label has been checked and every carton is ready. That would be an unjustified leap. Each record answers a smaller question. The real examination skill is deciding where a sound result ends and an unsupported conclusion begins.
This final chapter in the inventory sequence is a guided evidence hearing, built for the learner rather than for a fictional inspector. It shows how Physics, Mathematics and English can be practised together without pretending that the SEC has a combined EMS paper. You will check a ray calculation, compare expected and observed counts, read a notice closely, repair flawed statements and decide what a cautious team can say at the end of the exercise.
The first task is to recover the correct foundational work; the last is to draft and defend a short evidence-based conclusion under time pressure. Every number and operational rule in the habitat case is invented for classroom practice. None is a real equipment specification, school assessment prediction or safety authorisation.
What the official 2027 G3 SEC syllabuses actually assess
Singapore–Cambridge Secondary Education Certificate examinations begin in 2027. SEAB’s 2027 G3 syllabus index lists the separate school-candidate subjects: English Language K300, Mathematics K310 and Physics K323. For English K300, Writing, Comprehension, Listening and Oral Communication carry 35%, 35%, 10% and 20% respectively; Comprehension Paper 2 lasts 1 hour 50 minutes with 50 marks. Mathematics K310 has two papers of 2 hours 15 minutes and 90 marks each, with equal 50% weights. Physics K323 has multiple-choice Paper 1 (40 marks, 30%), structured/free-response Paper 2 (80 marks, 50%) and practical Paper 3 (40 marks, 20%).
This is a cross-subject study workshop, not an official G3 SEC integrated examination, not a guaranteed question theme and not an additional assessment component. The Science examples use the pure Physics K323 syllabus; Combined Science candidates must check their actual registered subject, such as K326 Physics/Chemistry or K327 Physics/Biology, rather than assume its papers are identical to pure Physics. No special examination time or mark weighting applies to this original EMS hearing.
The owner articles remain separate. Vol 1285 teaches optical ray-path evidence, Vol 1286 teaches inventory probabilities and review capacity, and Vol 1287 teaches English Paper 2 source comparison and summary. Vol 1284 supplies the previous receiving-bay evidence-boundary habit. This volume deliberately does not replace those subject lessons. It tests whether a learner can carry their distinctions into one decision.
The complete case file
File A is a simplified transparent cover. An ideal ray enters it from air at 30 degrees to the normal. The model cover has a refractive index of 1.50, and the ray meets a flat boundary. The air-to-cover drawing is one selected optical path; no detector geometry or real scanner readings are supplied. The physics question is whether the ray directions and relevant angles follow the stated laws.
File B is a fictional inventory record containing 120 cartons. A preliminary probability model assigns 0.90 to a single check being readable and 0.10 to its being flagged. The model assumes independent repetitions for a particular two-check calculation, an assumption that would need investigation in a real repeated-reading process. It predicts an average workload of twelve flagged cases across 120 checks. The later register, however, reports fourteen flagged cases.
File C is a staffing model. Two manual reviewers each have thirty minutes, and every review takes five minutes. Neither reviewer is assumed to process more than one case at the same time. Therefore each can complete six reviews in the allotted period, and together they can complete twelve. This does not say a real operation has exactly the same review time or that all completed reviews end in a correct match.
File D contains two communications. The early briefing says that about twelve cases may need manual review under the assumed ten-per-cent model. The later status notice lists fourteen actual flags, twelve available review places and a warning that a readable label is not necessarily a verified match. A draft public statement ignores that distinction and reads, ‘All 120 cartons have been confirmed by the optical system.’ Your job is to evaluate the draft, not to improve its confidence.
The judgement rules of this exercise
Before considering the draft, give every statement one of four statuses. ‘Supported within the model’ means the given mathematical or physical result follows from the listed assumptions. ‘Recorded’ means a particular document reports a number, without automatically proving the event was independently audited. ‘Conditional’ means a claim is valid only if an assumption is maintained. ‘Not established’ means the file lacks evidence for the conclusion. A contradiction can be stated separately when a claim directly clashes with a given record.
In this way the hearing examines how an apparently successful school answer can become a misleading headline. A correct optical angle, a correct probability and a correct English sentence are not enough if the sentence claims something that neither numerical result proves. The learner should ask what was actually measured, which quantities were calculated, how the text describes the result and what remains unresolved.
Start with the exact question, not the most impressive fact
The hearing asks which statements are warranted. It does not ask students to design real inventory machinery or settle a commercial dispute. A long explanation of refraction will not rescue a conclusion that forgets the remaining two reviews. Read the final instruction first, then return to the files to identify the evidence needed. This reading habit transfers directly to examination tasks where background information is much longer than the answer the candidate must give.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
Make four columns before reasoning
An efficient evidence ledger has the proposed claim, its source, the relevant calculation or quotation and the status of the claim. A fifth column can record a corrected version. For example, ‘The ray enters a higher-index cover and bends towards the normal’ belongs to the optical drawing. ‘Fourteen cases were flagged’ belongs to the later register. Do not fill the source column with ‘the case’ or ‘the teacher said’; those descriptions are too broad to identify the evidence that supports a statement.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
A physically sound calculation still has a physical boundary
At a 30-degree air-to-cover incidence with refractive index 1.50, sin r equals sin 30 degrees divided by 1.50, giving one third. The refracted angle is approximately 19.5 degrees. The reflected angle for the ideal flat surface is 30 degrees from the normal. These are ray results. They do not report whether a detector receives enough light, whether printed marks are readable, or whether detected identifiers match their physical cartons.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
Do not use a critical-angle result in the wrong direction
For light travelling from the higher-index cover into air, a simplified critical angle is arcsin(1 divided by 1.50), about 41.8 degrees. Total internal reflection requires travel from higher to lower refractive index at an incidence above the critical angle. A ray travelling from air into the higher-index cover cannot be used to claim total internal reflection at this entry boundary. A learner who states this boundary has kept the optical reasoning under control.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
A readable result is not a verified match
The fictional scanner outcome ‘readable’ only concerns detection of a label. Verification compares a detected identifier with an agreed reference and the actual carton. A readable label can be attached to the wrong carton, and a flagged reading does not prove a mismatch. These possibilities are why the public statement ‘all cartons confirmed’ cannot be justified by a high probability of readability alone. The word confirmed needs a definition and supporting observations.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
Probability is a model of uncertainty, not a guarantee
Under a stated independent two-check model, probability of both results readable is 0.90 times 0.90, or 0.81. The probability of exactly one flagged result is 0.90 times 0.10 plus 0.10 times 0.90, or 0.18. The probability of at least one flag is one minus 0.81, or 0.19. These are model-based probabilities, not a report of the 120 actual carton records. They assume that the repeated trials genuinely satisfy independence.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
Expected count and observed count have different jobs
Ten per cent of 120 is twelve, the expected number of flagged cases under a particular probability model. The later record reports fourteen. Both statements can be true. To describe a realised inventory, use the actual register count and date rather than silently replacing it with the model expectation. The fact that the observed total differs from the expected total does not by itself establish that the probability model is wrong; a random process permits variations.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
Capacity is a comparison with whole, available resources
Thirty minutes divided by five minutes per case yields six complete reviews per person. Two people can complete twelve under the simplified parallel-working assumption. Fourteen observed flags exceed twelve available review places, so two flagged cases remain outside the scheduled capacity. A person who states that the system can complete all the flagged reviews because the model expected twelve has answered the wrong problem. The observed queue is fourteen.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
The apparent agreement of sources may be an echo
If a public notice repeats a number copied from the register, the notice is not independent verification of that number. An earlier forecast and a later count are different sorts of information, even if their figures happen to agree. A learner should ask whether two documents used separate observations or merely repeated the same source. This question protects against a common evidence error: treating repeated wording as corroboration.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
Read ‘current’ and ‘about’ as meaningful words
‘About twelve’ marks a prediction with approximate language; ‘currently fourteen flagged’ points to a dated record. Neither expression is decorative. Removing ‘about’ from an expectation or removing ‘current’ from a register statement changes how much certainty a reader may take from the sentence. In English Paper 2, these small qualifiers can be central to literal understanding, inference and responsible summary.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
Separate a correction from a confession of incompetence
A revised notice can say that the earlier estimate was twelve and the later observed count is fourteen. That explains a change in available information without claiming that either document was deliberately deceptive. An effective correction identifies what changed, its source and the practical consequence. A careful learner avoids assigning a motive to an author when the text supplies only a numerical difference.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
Do not confuse three types of limitation
A model limitation arises when ideal rays, independent trials or fixed review times stand in for more complex reality. A data limitation occurs when observations are missing or unverified. A language limitation occurs when a sentence uses a status word more broadly than the source permits. The same final claim can fail for all three reasons, but identifying each separately makes the proposed remedy precise. Extra optics calculations will not repair an absent verification record.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
Use the command word to select the action
‘Calculate’ calls for an equation and an answer. ‘Explain’ requires a scientific relation or a cause linked to evidence. ‘Compare’ calls for similarities and differences with both sources named. ‘Evaluate’ asks whether a conclusion is supported, including assumptions and limits. ‘Suggest’ requires a proportionate next step. A combined hearing rewards flexibility, but a candidate must answer the individual subject question rather than produce one generic paragraph for every command word.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
A good handover message can be short
Students sometimes believe a longer message is more convincing. A concise, accurate update may be stronger: ‘The current register contains fourteen flagged cases. Under the scheduled time and assumed five-minute review rate, twelve can be reviewed, leaving two cases to be addressed in a revised plan. A readable label does not by itself confirm a correct carton match.’ This is calm, useful and within the evidence supplied. It makes no invented operational guarantee.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
Mark the first unsupported step
When a model answer is wrong, identify where the reasoning first loses contact with the record. An incorrect refraction angle needs a ray and normal check. A wrong workload decision needs a denominator or capacity check. A misleading public announcement needs a source and status check. In feedback, a precise diagnosis is more valuable than ‘be more careful’, because it tells the student what to practise next.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
Integrated practice should strengthen, not blur, the subjects
Each subject brings a different standard. The Physics calculation needs the correct ray law and physical assumptions. The Mathematics result needs a defined event, coherent operations and units. The English response needs close attention to wording, context and inference. They can sit in one case file without becoming interchangeable. Their joint usefulness comes from preserving, not erasing, the distinct ways they establish a claim.
Practice move: write a two-line answer to this section’s question. The first line should contain the calculation, quoted phrase or observation that does the evidential work. The second should state the conclusion and its boundary. This habit slows the first attempt by a few seconds but usually makes the final answer shorter, more exact and easier to check. Before moving on, ask which new fact would change the conclusion and which facts are merely interesting context.
Forty worked evidence claims, ordered from retrieval to advanced judgement
1. Identify the entry media
Proposed statement. Claim: ‘The ray starts in air and enters the model cover.’
Reasoned decision. Supported by File A. That is the named direction of travel. It must be preserved before a student decides which way the ray bends. Exam check. The boundary direction determines the correct refraction statement; do not silently reverse it.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
2. Check the reflection angle
Proposed statement. Claim: ‘The ideal reflected ray makes 30 degrees to the normal.’
Reasoned decision. Supported within the ideal flat-boundary model. The law of reflection equates the two normal-based angles. Exam check. The calculation makes no statement about intensity or the position of a detector.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
3. Calculate the refracted angle
Proposed statement. Claim: ‘The angle in the cover is about 19.5 degrees.’
Reasoned decision. Supported: sin r equals sin 30 degrees divided by 1.50, so r is approximately 19.5 degrees. Exam check. The smaller angle agrees with bending towards the normal in the higher-index material.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
4. Reject a false angle
Proposed statement. Claim: ‘The refracted ray must be at 45 degrees because the cover is clear.’
Reasoned decision. Contradicted by the given indices and incidence. Transparency alone is not an angle equation. Exam check. Use the supplied medium ratio rather than appearance of a material.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
5. Clarify the unit
Proposed statement. Claim: ‘Refractive index is 1.50 degrees.’
Reasoned decision. Incorrect units. Refractive index is a ratio and is dimensionless; the ray angles are measured in degrees. Exam check. A unit check is often the quickest way to find a mixed-quantity answer.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
6. Calculate the model speed
Proposed statement. Claim: ‘Light travels through this material at about 2.00 × 10^8 m/s.’
Reasoned decision. Supported under the simplified relation v equals 3.00 × 10^8 divided by 1.50. Exam check. That speed does not directly describe how fast an inventory record is processed.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
7. Locate the critical angle
Proposed statement. Claim: ‘The critical angle to air is about 41.8 degrees for an ideal cover with index 1.50.’
Reasoned decision. Supported for a ray travelling from the cover towards air, not for a ray entering the cover from air. Exam check. Always write the medium direction next to a critical-angle calculation.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
8. Refuse the wrong TIR claim
Proposed statement. Claim: ‘Total internal reflection occurs when the air-to-cover incidence is 30 degrees.’
Reasoned decision. Not supported. Total internal reflection requires the higher-index-to-lower-index direction and incidence above the critical angle. Exam check. The direction gate must be checked before numerical comparison.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
9. Identify what a ray diagram measures
Proposed statement. Claim: ‘The drawing predicts a model path through a flat optical interface.’
Reasoned decision. Supported within File A. It models ray direction using specified media and geometry. Exam check. It does not count cartons or validate the register.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
10. Identify what a diagram cannot measure
Proposed statement. Claim: ‘The drawing proves the scanner identifies every label.’
Reasoned decision. Not established. A light-ray construction lacks detector performance and reference-match evidence. Exam check. Avoid a product-accuracy claim based on optical geometry alone.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
11. Read the single-check assumption
Proposed statement. Claim: ‘The model assigns a ten-per-cent probability of a flag.’
Reasoned decision. Supported as a stated modelling assumption, not as an independently measured device performance rate. Exam check. Say ‘the model assumes’ instead of ‘the scanner is always’.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
12. Calculate both readable
Proposed statement. Claim: ‘Two independent checks are both readable with probability 0.81.’
Reasoned decision. Supported by 0.90 multiplied by 0.90 under the specified independence assumption. Exam check. A repeated test on a persistent damaged label may be dependent in reality.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
13. Calculate exactly one flag
Proposed statement. Claim: ‘Exactly one flag across two independent checks has probability 0.18.’
Reasoned decision. Supported by adding RF and FR, each 0.09 in the given two-outcome model. Exam check. Name the two exclusive routes before adding.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
14. Calculate at least one flag
Proposed statement. Claim: ‘At least one flag has probability 0.19.’
Reasoned decision. Supported by the complement of both readable, 1 minus 0.81. Exam check. This includes the two one-flag routes and the both-flagged route.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
15. Distinguish independence from sequence
Proposed statement. Claim: ‘Two checks are independent simply because one comes after the other.’
Reasoned decision. Unsupported. Timing alone does not establish independence; the cause of a failed reading may persist. Exam check. Use independence only as an explicit modelling premise or justified property.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
16. Expected flags
Proposed statement. Claim: ‘Twelve flags are expected from 120 checks at probability 0.10.’
Reasoned decision. Supported as expectation, 120 times 0.10. Exam check. The expected value is not a guaranteed realised tally.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
17. Observed flags
Proposed statement. Claim: ‘The later register lists fourteen cases flagged.’
Reasoned decision. Supported as a report of that register, File B, and should be attributed as such. Exam check. A document’s report is not necessarily an independent audit.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
18. Avoid replacing the register
Proposed statement. Claim: ‘Only twelve were flagged because the probability model says so.’
Reasoned decision. Contradicted by the later register’s fourteen; twelve describes the expectation. Exam check. Use observed figures for an observed workload question.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
19. Compare model with observation
Proposed statement. Claim: ‘The observed flag count exceeds the expectation by two.’
Reasoned decision. Supported by fourteen minus twelve equals two. Exam check. The discrepancy alone is not evidence of cheating, error or a defective optical cover.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
20. Convert observed count to percentage
Proposed statement. Claim: ‘The current register has about 11.7% flagged.’
Reasoned decision. Supported by 14 divided by 120 times 100, rounded to three significant figures. Exam check. Name the denominator and the fact that this is observed, not modelled.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
21. Calculate one staff member
Proposed statement. Claim: ‘One reviewer can complete six five-minute checks in thirty minutes.’
Reasoned decision. Supported under the uninterrupted working model: thirty divided by five equals six. Exam check. Whole reviews, not fractions, count towards cleared cases.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
22. Calculate combined capacity
Proposed statement. Claim: ‘Two reviewers can complete twelve checks in the available period.’
Reasoned decision. Supported under the model of independent parallel work. Exam check. This is capacity, not an observation of how many have already been completed.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
23. Apply capacity to observed need
Proposed statement. Claim: ‘Fourteen flagged cases can all be cleared in the thirty-minute slot.’
Reasoned decision. Contradicted by the twelve-case capacity under the stated model. Exam check. Do not treat the expected twelve as the actual queue.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
24. Find the unresolved queue
Proposed statement. Claim: ‘Two flagged cases remain outside this shift’s ideal review capacity.’
Reasoned decision. Supported conditionally: fourteen flagged less twelve review places leaves two. Exam check. Do not call the two cases defective; they are unresolved in this model.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
25. Ask for extra time
Proposed statement. Claim: ‘Two remaining reviews need ten person-minutes.’
Reasoned decision. Supported if each still needs five minutes and no interruptions occur. Exam check. If two people each take one case concurrently, the elapsed time could be five minutes.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
26. Name the real-world limit
Proposed statement. Claim: ‘The staffing calculation certifies the receiving bay as safe.’
Reasoned decision. Not established. It covers review counts under fictional timing assumptions. Exam check. Avoid operational certification or an invented regulation.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
27. Read the earlier notice
Proposed statement. Claim: ‘The early briefing projected about twelve flags.’
Reasoned decision. Supported as the earlier source’s statement. Exam check. The word ‘projected’ is essential to the meaning.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
28. Read the later notice
Proposed statement. Claim: ‘The later notice reports fourteen flags.’
Reasoned decision. Supported as the later source’s statement. Exam check. The word ‘later’ indicates chronology, not necessarily superior validation of every detail.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
29. Decide whether the notices contradict
Proposed statement. Claim: ‘One notice must be dishonest because twelve differs from fourteen.’
Reasoned decision. Not established. An expected count and an observed count may differ without any logical contradiction. Exam check. Identify the distinct information types before making a judgement.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
30. Explain ‘readable’
Proposed statement. Claim: ‘A readable scan proves the carton is correctly matched.’
Reasoned decision. Not established. A detected label and a verified reference match are different statuses. Exam check. Precise definitions prevent a broad claim from gaining undeserved credibility.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
31. Explain ‘flagged’
Proposed statement. Claim: ‘A flagged carton is proven to contain wrong goods.’
Reasoned decision. Not established. A flag indicates a reason for further checking. Exam check. Do not turn an unresolved record into a confirmed defect.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
32. Audit the all-clear statement
Proposed statement. Claim: ‘All 120 cartons have been confirmed by the optical system.’
Reasoned decision. Unsupported by Files A–D and inconsistent with the unresolved manual-review records. Exam check. The safe repair reports counts, sources and next checks without asserting confirmation.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
33. Audit the phrase ‘our machine is perfect’
Proposed statement. Claim: ‘No scanner error is possible.’
Reasoned decision. Not established by a ray diagram or a probability assumption. Exam check. A prediction for an ideal ray is not a performance warranty.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
34. Read the source of a claim
Proposed statement. Claim: ‘A notice repeating the register is independent verification.’
Reasoned decision. Not necessarily. A copied register value and its printed notice share the same information source. Exam check. Seek independently collected evidence when an independent audit is claimed.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
35. Distinguish a conclusion from a recommendation
Proposed statement. Claim: ‘Because two cases remain, the team should schedule further checks.’
Reasoned decision. A proportionate recommendation within the fictional planning exercise, not a factual claim that checks have already occurred. Exam check. Use ‘should’ for a suggested next step, not ‘have’ for a completed action.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
36. Draft a precise correction
Proposed statement. Claim: ‘Our earlier projection was twelve; the current register reports fourteen.’
Reasoned decision. Supported, with clear time and evidence distinctions. Exam check. Do not conceal the earlier model or present the updated count as a timeless fact.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
37. Explain a limitation orally
Proposed statement. Claim: ‘The optical model predicts a path, not a guaranteed successful identification.’
Reasoned decision. Supported as an accurate boundary statement and useful for spoken explanation. Exam check. The difference matters more than an impressive formula recital.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
38. Construct a calibrated public update
Proposed statement. Claim: ‘Fourteen are flagged; under the current staffing plan twelve can be reviewed, leaving two in the queue.’
Reasoned decision. Supported within the stated files and conditions. Exam check. Mention that readable status alone is not independent carton verification.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
39. Respond to an examiner’s challenge
Proposed statement. Claim: ‘Why not say all is complete if most checks are readable?’
Reasoned decision. Because the question is verification, not mere readability. There are unresolved flagged cases, and no full verified-match record is supplied. Exam check. State the distinction respectfully without inventing equipment faults.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
40. Deliver the final ruling
Proposed statement. Claim: ‘The combined evidence supports a limited update, not a full release certificate.’
Reasoned decision. Supported. The ray path and arithmetic follow their assumptions, the notices can be reconciled, but the file lacks proof that all labels match. Exam check. A responsible conclusion names what is supported and what remains unverified.
For revision, write a one-sentence counterfactual: what additional observation, assumption or source would be required to turn the disputed assertion into a properly supported conclusion? Where the statement already follows from the model, state the model’s limits. This is how an apparently familiar calculation becomes a flexible examination skill rather than a memorised number.
One original timed mini-mock across English, Mathematics and Physics
Set aside a fixed practice period that suits the learner; this is not an official exam time allowance. Part A supplies the ideal ray entering a cover of index 1.50 from air at thirty degrees. Calculate the reflected and refracted angles and explain one limitation of the result. Part B supplies a 0.90 single-check readability probability and an independence assumption for two checks. Calculate the probability of exactly one flag and explain why this is not a verified-match probability. Part C supplies 120 cartons, fourteen observed flags and two five-minute reviewers with thirty minutes each. Determine the maximum completed reviews and queue remainder.
Part D reproduces an early notice predicting twelve flags and a later notice reporting fourteen. Explain whether the two statements must contradict each other. Then interpret the meaning of ‘readable’ and give one inference that the notices do not justify. Part E asks for a three-sentence public update that avoids a false all-clear claim while accurately reporting the record and the simplified capacity check. Finally, state which official SEC subject would assess each separate skill and why no combined EMS paper exists.
Worked answers and marking rationale
Part A: reflection equals 30 degrees measured from the normal. For refraction, sin r = sin 30 degrees / 1.50 = 1/3, hence r ≈ 19.5 degrees. The result describes an ideal ray through specified media; no device accuracy or verified-match rate follows from it. A wrong answer that says 70.5 degrees may have measured an angle from the surface. A correct diagram with a normal at the incidence point makes this easy to diagnose.
Part B: exactly one flag means readable-then-flagged or flagged-then-readable. With independent repetitions, 0.90 × 0.10 + 0.10 × 0.90 = 0.18. This is a probability for outcomes under the stated model. It neither records how many cases are flagged in the existing inventory nor says whether detected labels match the cartons. A strong answer states the independent-trial condition; without that condition, the multiplication is not justified.
Part C: each reviewer completes 30 / 5 = 6 complete reviews, so both together can complete 12 in the model. The current queue contains 14 flags; therefore two exceed the scheduled review capacity. The expected twelve from the probability model is irrelevant to the subtraction once the actual register count is supplied. The conclusion is conditional on uninterrupted parallel work and the five-minute rate. Neither a review’s outcome nor a real-world certification is specified.
Part D: the two notices need not contradict each other, since the first projected workload and the second reported an observed count. ‘Readable’ indicates that a label was detected; it does not prove a correct physical-item match. An inference not justified is that the optical device itself is defective. The documents report statuses but do not establish the device’s causal failure or its design. A good comprehension response stays close to the documents’ time markers and definitions.
Part E, sample answer: ‘The current register lists fourteen flagged cartons, and a readable scan does not by itself confirm a correct match. Under the stated five-minute review estimate and the two reviewers’ scheduled thirty minutes, twelve flagged cases can be reviewed in this period, leaving two for an updated plan. The team will report the results of verification separately once those records are available.’ This is a classroom update, not a real safety notice. It avoids claiming that outstanding checks have already been completed.
For the subject-mapping part, the ray laws belong to G3 Physics K323, probability and capacity reasoning to G3 Mathematics K310, and close reading and concise evidence-based wording to G3 English Language K300. English writing practice can also use such reports, although this volume has deliberately used the Paper 2 comprehension mindset. A student cannot transfer the Physics paper weight to an integrated EMS task; the schools assess subjects under their separate syllabuses.
How to mark an answer without hiding the true weakness
Do not mark only the final number. For Physics, mark the normal, named media, appropriate law, substitution, units and direction check. For Mathematics, mark defined events, independence, branch operations, whole-number resources and interpretation. For English, mark evidence selection, difference between facts and inferences, the effects of qualifiers and accurate paraphrase. Then circle the first step that did not follow from the source. The student can repair one causal weakness instead of rewriting an entire essay that mostly worked.
Make a compact correction record for each weak attempt: the original claim, the missing premise, the corrected sentence and a new transfer question. If a learner wrote ‘twelve flagged’ because the expectation was twelve, the next question should use a different expected and observed count. If the learner equated readability with verification, the transfer question should replace cartons with library books or laboratory samples. A correction that transfers to an unfamiliar case is more valuable than memorising the teacher’s preferred wording.
Three passes from foundation to advanced examination performance
Pass one is foundation. Draw and label the ray; define readable and flagged; calculate a simple percentage and whole-number rate; underline qualifiers in each notice. Give enough time for correct interpretation. Pass two is examination control. Solve a short Physics calculation, a two-branch probability, a contextual workload comparison and a source-inference question without consulting the worked models. Pass three is advanced judgement. Audit a new public claim with evidence from all three subjects while limiting every conclusion to what the case actually establishes.
For a fresh transfer problem, imagine that the cover’s refractive index changes to 1.40, the flag model changes to 0.08 and the later recorded queue becomes eleven. The candidate should recompute the refracted angle, model expectation and available review places, rather than copying old answers. More importantly, the candidate should still know that readability is not verification and that a register statement needs attribution. These conceptual boundaries survive changes in the numbers. That is the most durable part of the exercise.
A learner who finds the integrated task overwhelming can return to a single owner. Rebuild ray diagrams with Vol 1285, event trees and denominator language with Vol 1286, and source comparisons with Vol 1287. Once the separate skills are reliable, practise the hearing again. Depth does not mean forcing all skills into one attempt before the basic foundations are ready. It means choosing the smallest repair that makes the next performance more independent.
Where to go next
Keep the specialist ownership clear: Physics Vol 1285 for reflection and refraction; Mathematics Vol 1286 for probability and whole-number planning; English Vol 1287 for comprehending notices, inference and summary; EMS Vol 1284 for an earlier evidence handover. For larger learning paths use the Sengkang G3 Science guide, G3 Mathematics application owner, Secondary English Sengkang owner and Sengkang PSLE preparation route. The verified syllabus sources are K300 English, K310 Mathematics and K323 Physics under the 2027 G3 SEC syllabus index.
The conclusion is deliberately modest. The ray law explains a model path. The probability model describes possible outcomes under assumptions. The later register records fourteen flags, while the staffing model provides twelve places for reviews. A careful notice can state those points and identify two outstanding cases. It cannot truthfully turn them into proof that every carton is verified. The central examination lesson is to preserve the distance between what a method shows and what a claim says.
