G3 SEC EMS Cold-Pack Evidence Audit

The notice is ready to print. The evidence is not.

A fictional school science club plans a short learning booth about the energy involved when ice melts. Its draft notice announces: ‘Sample B is twice as efficient as A. Four packs can keep medicines cold all afternoon, as proved by our experiment.’ The sentence sounds decisive. It is also wrong in several distinct ways. The only supplied observation has a twelve-minute near-melting interval for B and eight minutes for A, not a two-to-one ratio. The energy calculation supports an ideal whole-pack count under stated assumptions, not a measured cooling duration. No medical-storage safety evidence exists.

To repair this notice, you need three kinds of careful thinking. Physics explains why energy can enter ice without a rise in temperature and what a latent-heat calculation actually means. Mathematics turns a continuous energy requirement into a feasible whole-number count without rounding away a boundary. English separates observation, calculation, interpretation and unsupported promise, then writes a clear version for the reader. Each part protects the other two from overstatement.

This is an original, text-only English–Mathematics–Science learning case. EMS is our integrated study approach, not an official SEC subject, paper or examination code. The data are fictional and designed to teach reasoning; they are not a record of an actual experiment, an equipment test or advice about handling, storing or transporting medicine or food. The exercise is about what claims the supplied evidence can support.

The official examinations behind the three skills

The Singapore–Cambridge Secondary Education Certificate begins in 2027. For G3, the published reference codes include English Language K300, Mathematics K310 and Physics K323. English has four assessed components covering Writing, Comprehension, Listening and Oral Communication. Mathematics has two written papers, each two hours and fifteen minutes. Pure Physics has three papers, including practical assessment. Their assessments remain subject-specific. This combined classroom case does not merge them into one examination.

Use the official SEAB 2027 G3 syllabus list and the separate K300 English, K310 Mathematics and K323 Physics documents for details. Specific latent heat is a pure Physics K323 topic; do not assume identical requirements in every G3 Combined Science syllabus. The transferable reasoning used here is useful beyond this one subject combination.

Separate the two evidence cards

The booth’s first card is a modelled energy budget. It asks how much useful energy would be required to take 0.100 kg of ice from −5°C to liquid water at 10°C under idealised conditions, then asks how many complete equal packs could meet that journey with 170,000 J available. The card supplies c_ice = 2100 J/(kg°C), latent heat of fusion ℓ = 334,000 J/kg and c_water = 4200 J/(kg°C). It specifies useful energy transferred into the material, neglecting losses.

The second card is a separate fictional observation. Sample A was near the melting temperature for approximately eight minutes; sample B for approximately twelve minutes. At minute twenty, A had a recorded temperature of 6°C and B of 4°C. The observation card does not identify either sample as one of the standard 0.100 kg packs from the model card. Nor does it provide enough information to isolate a physical cause for the different readings. The audit must keep the cards separate unless the source explicitly connects them.

Why the evidence cards must not be merged casually

A common communication error occurs when one source contains a calculation and another contains measurements. The writer combines both into one confident story without checking whether they describe the same objects or conditions. Here, the first card concerns ideal energy required per specified mass and a total useful budget. The second compares observed time and temperature values for two unspecified samples. Neither card establishes an efficiency ranking, a certified duration or a safety outcome.

The separation is not pedantic. It determines whether a numerical result is meaningful. You may report four packs as the maximum theoretical complete count under the first card. You may report the observed four-minute difference in near-melting intervals under the second card. But saying that the four model packs were observed to outperform A for twelve minutes would invent a connection. Evidence literacy begins by naming what each dataset owns.

The Physics account: warmth without a rising thermometer

During the melting of pure ice at the stated transition temperature, energy can be transferred into the sample while the measured temperature remains approximately constant. In the particle model, transferred energy changes the particles’ arrangement and potential-energy contribution rather than immediately increasing the average kinetic-energy contribution associated with temperature. That is why a plateau in an idealised heating curve can represent a real physical process rather than a pause in heating.

An observation of a steady temperature alone does not prove the whole causal story. It becomes evidence for melting when paired with information about the material, its state, heating and transition. If the sample is an unknown mixture or its temperature probe is unreliable, interpretation needs care. The Physics guide Vol 0422 develops the particle account, latent energy and incomplete-melting branches. This audit uses those skills without repeating every example.

Calculate the first energy stage

The standard pack contains 0.100 kg of ice initially at −5°C. To reach its melting temperature of 0°C, its temperature must rise by 5°C. With c_ice = 2100 J/(kg°C), the ideal energy needed is E_1 = 0.100 × 2100 × 5 = 1050 J. This stage is warming a solid. The right relation is E = mcΔT because temperature changes and no state change is included within this stage.

A useful check is the unit cancellation: kilograms multiplied by joules per kilogram per degree and then by degrees yield joules. If the mass had been left as 100 g in a formula expecting kilograms, the numerical output would be a thousand times too large. The amount also makes physical sense: warming the frozen sample by five degrees needs less energy than melting it fully under the supplied constants. We should not, however, assume the same balance for every possible material or temperature range.

Calculate the melting energy

At 0°C, the 0.100 kg sample changes state from ice to liquid water in the ideal model. The specific latent heat of fusion supplied by the question is 334,000 J/kg. The energy for complete melting is E_2 = mℓ = 0.100 × 334,000 = 33,400 J. The temperature change for this ideal transition is zero, but the energy transferred is not. That is the important conceptual distinction illustrated by the booth.

If a learner mistakenly used E = mcΔT for melting, the result would be zero because ΔT is zero. That is a formula-selection error, not a calculator error. The relevant question is what the material does. During a phase change, latent energy must be accounted for. The direction of energy transfer also matters: melting requires energy entering the sample, while freezing releases energy from it under otherwise comparable conditions.

Calculate the final warming stage

After complete melting, the 0.100 kg of liquid water warms from 0°C to 10°C. The ideal energy needed is E_3 = 0.100 × 4200 × 10 = 4200 J. This is a liquid-warming step, so E = mcΔT applies again, with the appropriate specific heat capacity for liquid water. The mass is still 0.100 kg in the simple model; melting has changed the state, not created extra matter.

Now add the three energy contributions: 1050 + 33,400 + 4200 = 38,650 J. The coefficient for one kilogram under this exact journey is 386,500 J/kg, a convenient shorthand for later algebra. It is not a universal property of all cold packs. If the initial temperature, final temperature, substance or heat-transfer assumptions change, the coefficient must be reconstructed.

The Mathematics account: one continuous bound

The budget is 170,000 J of useful energy. Dividing by 38,650 J per standard pack gives approximately 4.398 packs. In a mathematical model that allowed fractions of a pack, this quotient tells us the continuous upper bound. But the source requires complete, sealed 0.100 kg units. The number of packs must therefore be an integer, not a decimal. We need the greatest whole number n satisfying 38,650n ≤ 170,000.

Four packs require 154,600 J. Five require 193,250 J. The first amount fits within the budget; the second exceeds it. Therefore four is the maximum complete count under this model. This is stronger than simply rounding 4.398 down. We have proved that four work and that the next integer fails. This kind of two-sided boundary check is valuable for many resource-allocation problems.

The spare budget is not a safety margin

Four packs leave 170,000 − 154,600 = 15,400 J unused in the ideal budget. This is a numerical surplus in a model. It is not a guarantee that four real packs will perform as planned if containers absorb energy, heat transfer varies or other sources of loss are present. The phrase ‘safety margin’ can be misleading because readers may interpret it as validated protection rather than unused model capacity.

The unused fraction is 15,400/170,000 ≈ 9.06 percent of the available budget. If a draft instead divides by the four-pack requirement, it obtains about 9.96 percent and must label that different base. The Mathematics guide Vol 0423 treats reverse mass and inequality problems in detail. The audit’s job is to make sure these numbers are not given a meaning they do not carry.

What the observation card actually supports

The near-transition interval for sample A is approximately eight minutes; for B it is approximately twelve. The difference is about four minutes, and the B interval is about fifty percent longer relative to A’s eight-minute interval. At minute twenty, the recorded temperature of B is 4°C, which is 2°C lower than A’s 6°C. These are descriptive results within the fictional observation card.

None of them proves that B is ‘twice as efficient’. A twelve-to-eight duration ratio is 1.5, not 2, and the term ‘efficient’ has not been defined in the source. Nor can the two-degree difference at one time be treated as a universal temperature advantage. A precise statement names the quantity, the comparison basis and the observation window. Accuracy here comes from ordinary arithmetic and careful words.

A longer observed plateau does not identify the cause

A possible cause of a longer near-transition interval is a difference in sample mass, useful energy-transfer rate, surrounding conditions, composition or thermal contact. The supplied observation does not identify the relevant values or controls. It is therefore unsound to say that B’s insulation must be superior or that its latent heat must be larger. The observation can support a comparison without supporting a particular causal explanation.

To establish a cause, an investigation would need an appropriately controlled method that isolates the variable of interest. A student can state this without designing a risky procedure. For example, the audit might recommend comparing equal masses under matched conditions and documenting the measurements, subject to appropriate supervised school procedures. The important language is ‘would need further evidence’, not ‘has already proved’.

The first claim fails: ‘twice as efficient’

The claim has two defects. Numerically, twelve minutes compared with eight is a ratio of 1.5, so ‘twice’ is false even if one equates duration with performance. Conceptually, efficiency usually compares useful output with some relevant input; the observation card supplies no definition or measurement of efficiency. A duration comparison alone cannot establish an efficiency ranking.

Repair the sentence without trying to save the appealing adjective: ‘In this illustrative observation, sample B remained near the melting temperature for about twelve minutes, compared with eight minutes for sample A.’ If the purpose is to describe the later readings, add: ‘At minute twenty, B was recorded at 4°C and A at 6°C.’ The repaired claims are less dramatic but more informative because they show the reader exactly what happened in the source.

The second claim fails: ‘four packs were proved by experiment’

The four-pack result comes from a numerical model. It assumes a specified pack mass, temperatures, constants and useful-energy budget. It is not an observed count from the separate A/B temperature data. A writer who calls it an experimental proof mislabels its evidence source. The fact that the calculation is correct does not transform a calculated feasibility boundary into a measured result.

A faithful replacement is: ‘Under the ideal energy model supplied for this exercise, the budget is sufficient for four complete standard packs, but not five.’ The conditions now travel with the answer. The sentence neither hides the result nor pretends that a physical trial has tested it. This distinction between calculated, predicted and observed is important in English reading and writing as well as in scientific reasoning.

The third claim fails: ‘all afternoon’ and ‘medicines’

The original data do not establish a real-world storage duration, environmental conditions, required temperature limits, regulatory standards or medical suitability. The expression ‘all afternoon’ extends far beyond the short fictitious observation, while ‘medicines’ introduces a safety-critical application that was never tested. No amount of rephrasing can make those additional claims emerge from the supplied numbers.

Remove them. The booth can safely teach why a temperature plateau may occur during melting and how a modelled energy budget is checked. It should not give practical guidance for storing medication. A reader-safe message identifies the activity as educational and fictional. An honest educational explanation often gains clarity when an unrelated high-stakes claim is simply left out.

The corrected visitor caption

A concise visitor caption might read: ‘Why can ice melt without getting warmer straight away? During melting, energy helps the solid change into liquid water, so its temperature may stay near the melting point. In our invented classroom comparison, one sample stayed near that temperature for about eight minutes and another for about twelve. That describes this example; it does not tell us how every sample or container would behave.’

Every sentence has a job. The question gives the visitor a reason to read. The next sentence explains the scientific mechanism. The comparison offers concrete evidence that readers can understand. The final sentence protects the scope of the result. No figures are fabricated, and the caption does not need to mention the whole-pack energy budget because that is a separate piece of the exhibit.

The corrected technical note

The committee-facing note could say: ‘For a model pack of 0.100 kg ice moving from −5°C to liquid water at 10°C, the supplied stage-wise constants give a useful-energy requirement of 38,650 J. A budget of 170,000 J permits at most four whole packs: four require 154,600 J and five require 193,250 J. This is a modelled feasibility statement. It does not establish measured cooling performance.’

The note could continue: ‘A separate fictional observation gives near-melting intervals of approximately eight and twelve minutes for samples A and B, with minute-twenty temperatures of 6°C and 4°C. Because the sample specifications and full experimental controls are unspecified, the observed difference cannot be attributed to one physical factor. Neither dataset supports medical-storage or universal product claims.’ This version is longer than the visitor caption because the committee needs the audit trail.

Why the two texts are compatible

One piece explains the physical phenomenon and a limited observation; the other keeps the calculations and distinctions needed for review. They are not competing versions of the truth. They serve different readers while preserving the same boundary between evidence and inference. The English guide Vol 0424 develops that audience adaptation. Here we test it alongside the actual Mathematics and Physics.

A useful cross-check is to ask whether a visitor could read the technical note and recognise the same case. If the visitor text says the experiment proved a product was safe but the technical note says nothing of the kind, the adaptation has failed. Differences in vocabulary and level of detail are acceptable; differences in factual commitments are not.

Audit every number back to its source

Write a short ledger: 0.100 kg, −5°C, 10°C and the constants belong to the model card. So do 38,650 J per pack, the 170,000 J budget and the result of four complete packs. The eight- and twelve-minute intervals, and minute-twenty temperatures 6°C and 4°C, belong to the observation card. The caption’s phase-change explanation comes from the Physics principle. Statements about safety, product superiority or guaranteed duration belong to neither card and must be removed.

This ledger is a defence against accidental synthesis. When three subjects meet, it becomes easy for a fluent writer to borrow a number from one step and give it the authority of another. A good audit makes the origin of every key quantity visible, even if the final visitor caption remains short.

A table written in words

You do not need an image to organise this case. Imagine three columns titled ‘Statement’, ‘Evidence source’ and ‘Verdict’. ‘Four whole packs’ maps to the ideal energy model and is supported only with its assumptions. ‘B stayed near melting for twelve minutes’ maps to the fictional observation and is supported as an approximate result. ‘B is twice as efficient’ maps to no defined efficiency measure and fails. ‘The display explains latent heat’ maps to the science interpretation and is a reasonable educational purpose.

This text-only structure is useful in written examinations too. A small margin note can separate facts, modelled outcomes and unsupported conclusions before you write the final paragraph. You do not need to reproduce a full table for every answer, but you should be able to perform its logic in your head.

The danger of a correct number in the wrong sentence

The number four is correct for the maximum complete packs in the ideal budget. The number twelve is correct for B’s approximate near-transition minutes in the fictional observation. Yet ‘Four B packs last twelve minutes’ is not established because the cases have not been linked. This is a subtle error: neither number is wrong on its own, but their new relationship is invented.

Readers often notice obvious arithmetic mistakes more readily than evidence-relationship mistakes. In an advanced response, inspect the verbs and prepositions connecting your figures. ‘Requires’, ‘remained’, ‘proved’, ’caused’, ‘suitable for’ and ‘lasts’ each make a different claim. Match the verb to the evidence, not to the tone you want.

What if the energy is insufficient for complete melting?

Consider a revised model card containing 0.100 kg of ice already at 0°C, with only 25,000 J delivered usefully. Complete melting would require 33,400 J. The available energy can melt only 25,000/334,000 ≈ 0.0749 kg, leaving about 0.0251 kg of ice in the ideal model. A solution that warms all the resulting liquid after melting is impossible because the melting stage has not finished.

The audit should mark a changed branch clearly. An English notice cannot describe the whole sample as ‘warm water’ when the calculation predicts a mixture of solid and liquid. The mathematical feasibility decision and the physical state must agree. A calculation that yields a precise mass but is then described with the wrong state is an incomplete answer.

What if the pack size changes?

Suppose the model uses 0.080 kg complete packs instead of 0.100 kg while keeping the same temperature journey. Each smaller pack requires 0.080 × 386,500 = 30,920 J. With 170,000 J, five full smaller packs are feasible, consuming 154,600 J; six would require 185,520 J. The count changes from four to five, but the total ice mass of the feasible sets remains 0.400 kg in either case.

An unqualified notice saying ‘the smaller design is better because it supports five packs instead of four’ hides that difference in unit size. If the decision criterion is total mass handled under the model, the two options are equal. This is an opportunity to practise fair comparisons. A count is not meaningful without defining what one unit represents.

What if a temperature reading is missing?

Imagine that the minute-twenty temperature for sample B was not recorded. We could still compare the supplied near-melting intervals, but we could not state that B was two degrees cooler than A at that time. Filling in the missing value from a trend or intuition would be fabrication. The correct response is to name the missing observation and either omit the particular comparison or request further evidence.

A Mathematics paper may sometimes ask for a modelled estimate from given points, but that estimate must be labelled as an estimate, not a measurement. An English summary should preserve the same distinction. A scientifically plausible guess does not become recorded evidence simply because it helps a neat story.

What if two datasets conflict?

Suppose a committee note claims that B stayed near the transition for twelve minutes, while a later checked record shows ten minutes. The first step is to identify which source has authority for the final statement and whether a correction has actually been verified. Do not average the two claims as though they were independent temperature observations. A discrepancy between records is a source-verification problem, not automatically a statistical problem.

If the checked record replaces the earlier value, update the report before publication and make the correction clear. The relative comparison with A’s eight minutes would change from four minutes longer to two minutes longer, and the percentage increase from about fifty percent to about twenty-five percent. One corrected observation can propagate through both calculations and prose. This makes careful source tracking essential.

Reading a question’s command words

An integrated practice prompt may ask you to calculate, explain, evaluate and draft. Answer each command explicitly. For calculate, show the staged arithmetic and appropriate units. For explain, relate energy transfer to the particle model. For evaluate, identify what the evidence supports and what remains uncontrolled. For draft, produce audience-appropriate language that preserves the truth. A single long paragraph should not be used to disguise that one of these jobs has been skipped.

Under time pressure, allocate a line or short heading to each demand before writing. This makes it easier to identify whether the final communication contains the correct numbers. It also prevents you from responding to a request for a numerical bound with only a general discussion of energy conservation.

Three levels of checking

At the first level, check arithmetic: multiplication, conversion from grams to kilograms, and the difference between twelve and eight. At the second level, check the model: which stages are included, whether useful energy is assumed, whether packs must be whole and whether samples from different cards have been mixed up. At the third level, check the communication: whether ‘proved’, ‘safe’ or ‘best’ has been introduced without evidence.

A learner who checks only arithmetic may miss the most serious mistakes in the original notice. A learner who checks only grammar may polish a false claim. These three levels create a repeatable process: calculate correctly, interpret within the model, and communicate faithfully. They can be applied to many unfamiliar contexts beyond the example used here.

Surgical links for the learner route

The immediate foundation is Vol 0422 on specific latent heat. Next is Vol 0423 on reverse budgets and integer feasibility, then Vol 0424 on explaining one finding for two audiences. The preceding Vol 0421 insulated-cup audit provides a related earlier transfer task, but it concerns cooling comparisons rather than melting energy.

For a primary-school foundation, the PSLE guide on comparing the same things and the PSLE guide on temperature versus energy supply predecessor habits. The English reports and proposals owner gives a wider writing route. Read these as complementary owners, not repeated versions of the same question.

The final corrected evidence desk

The repaired answer has four parts. Physics: during melting, energy can change a material’s state without immediately raising the temperature. Mathematics: under the stated three-stage useful-energy model, one standard pack requires 38,650 J and the 170,000 J budget permits at most four whole packs. Observation: in the separate illustrative case, B stayed near the melting temperature for about twelve minutes and A for about eight, with recorded minute-twenty readings of 4°C and 6°C respectively. Communication: neither source establishes a universal efficiency ranking, a guaranteed duration or medical-storage safety.

Those four parts are precise enough to form a technical note and simple enough to guide a shorter public caption. If a student can carry them across different forms without adding a false connection, they have learnt something larger than the particular energy calculation. They have learnt how to keep evidence intact when moving between subjects.

Evidence-check 01: Claim: ‘twice as long’

Given near-transition intervals of eight and twelve minutes, the duration ratio is 12/8 = 1.5. B’s observed interval is 1.5 times A’s, or fifty percent longer relative to A. It is not twice as long. Draft a replacement sentence that identifies the two samples, the approximate duration and the single illustrative context. Then explain why neither the ratio nor the percentage measures efficiency unless useful output and relevant input have been defined.

For each correction, identify the source of the original claim: model, observation, physical explanation or unsupported assertion. Write one revised sentence that says no more than the evidence allows. If arithmetic is involved, show the relevant unit and comparison basis. If a cause is asserted, ask what controls would be needed to justify it. A strong answer is not the longest one; it is the one whose quantities and qualifications survive a careful second reading.

Evidence-check 02: Claim: ‘four packs were observed to work’

The calculation card has 170,000 J and a model requirement of 38,650 J per complete standard pack. It yields four feasible units, with five too costly. No observation of those model packs is supplied. Correct the wording to ‘four packs fit the stated ideal useful-energy budget.’ Identify the additional evidence that would be needed before claiming a real trial tested that count. The repair is about the status of evidence, not another multiplication.

Evidence-check 03: Claim: ‘B stayed two degrees cooler throughout’

The source gives a difference of two degrees only at minute twenty: B 4°C, A 6°C. A writer cannot extend that reading to the entire interval without further data. Replace ‘throughout’ with ‘at minute twenty of the illustrative observation’. This small edit prevents a single-time comparison from becoming a continuous claim. Decide whether a visitor caption should include the exact values or a shorter qualified statement, depending on its purpose.

Evidence-check 04: Claim: ‘the melting energy is zero’

The standard 0.100 kg sample melts at its transition temperature without a temperature rise in the ideal model. Its latent-energy requirement is 0.100×334,000 = 33,400 J. The zero temperature change does not mean zero energy transfer. Repair an explanation that uses mcΔT during the melting interval. Name the state change before choosing the equation and explain the role of the particles’ arrangement.

Evidence-check 05: Claim: ‘five standard packs fit’

The continuous quotient 170,000/38,650 is about 4.398. Because complete packs are required, four fit and five do not. Five would require 193,250 J, exceeding the budget. Test the upper neighbouring integer every time you make a maximum count claim. A decimal quotient is the starting point of the decision, not the final discrete outcome.

Evidence-check 06: Claim: ‘the budget is ten percent spare’

Four standard packs use 154,600 J, leaving 15,400 J of a 170,000 J total. The unused percentage of the budget is approximately 9.06 percent. Relative to the four-pack requirement, the surplus is approximately 9.96 percent. Neither can be called simply ‘ten percent’ without naming a base and indicating appropriate rounding. Rewrite the claim with the denominator made explicit.

Evidence-check 07: Claim: ‘the two cards test the same packs’

The energy model describes a specified 0.100 kg ice pack. The observation card gives data for A and B without establishing their masses or identifying them as those model packs. Do not merge them into one experiment. Draw a simple two-source ledger in words and assign each number to its owner. State what additional matching information would be needed before a combined interpretation became legitimate.

Evidence-check 08: Claim: ‘longer plateau proves superior insulation’

The fictional observation contains near-transition durations but not the relevant controlled variables. Different masses, heating conditions or materials could affect the durations. The safe statement is descriptive: B’s supplied near-transition interval was longer in this example. A proposed controlled comparison might help investigate the cause, but the current evidence does not isolate one. Explain the difference between a possible mechanism and a demonstrated cause.

Evidence-check 09: Change the starting ice temperature

Recalculate the standard pack if it starts at −2°C rather than −5°C. The solid warming term becomes 0.100×2100×2 = 420 J. Melting remains 33,400 J and liquid warming to 10°C remains 4200 J, giving 38,020 J per pack. At 170,000 J, four complete packs still fit. Report the changed per-pack requirement without implying that the observation card’s A/B plateau durations must also change.

Evidence-check 10: Change the final target temperature

Keep the standard 0.100 kg sample at −5°C initially, but end at liquid water 6°C. Solid warming uses 1050 J, melting 33,400 J and liquid warming 0.100×4200×6 = 2520 J. Total useful energy becomes 36,970 J per pack. With 170,000 J available, four standard packs fit and five require 184,850 J. Identify the altered stage and the unchanged ones. This is a model calculation, not a real display reading.

Evidence-check 11: A branch with ice remaining

A separate model supplies 25,000 J to 0.100 kg ice initially at 0°C. The full melting requirement is 33,400 J, so only about 0.0749 kg melts and about 0.0251 kg remains ice. A notice that describes the final sample as entirely liquid is false. Correct both the physical state and the number. If the problem later supplies additional energy, identify first whether it completes the remaining melting before raising liquid temperature.

Evidence-check 12: An anonymous quotation

A source quotes a visitor as saying ‘That pack will keep anything safe.’ The visitor’s opinion is not evidence that the statement is true. If the task asks you to report it, attribute it as a remark. If asked to evaluate it, explain that the provided data do not test safe storage and that no such guarantee follows. The difference between reporting and endorsing a claim is crucial when combining source material in an English response.

Evidence-check 13: Create a four-sentence visitor note

Write four sentences, each with a distinct purpose: a question about melting; a simple explanation of energy and state; the approximate eight-versus-twelve-minute observation; and a limit that the example is not a universal performance test. Check that the numerical facts come only from the observation card. Do not insert the four-pack result unless the note explicitly introduces a separate model and explains the change of source.

Evidence-check 14: Create a technical correction

Draft a short committee correction that states 38,650 J per model pack, four whole packs under the budget and the separate fictional A/B observation. Add that the samples’ specifications are not provided and that medical-storage conclusions are unsupported. A technical correction should show the evidence basis without apologetic padding. It should be possible for someone reading it to tell exactly which original claim was removed and why.

Evidence-check 15: Detect a changed denominator

A learner says B had ‘four minutes more’ and ‘fifty percent more’. Both statements can describe the same duration difference when A’s eight minutes is the baseline. Now use B’s twelve minutes as the comparison base: four minutes is one-third of B’s duration, approximately 33.3 percent. The different percentage does not mean the data changed. It means the denominator changed. Never publish a percentage without naming the reference.

Evidence-check 16: Check if a lower temperature is ‘better’

At minute twenty, B is 4°C and A is 6°C. That gives a lower reading for B, but whether it is ‘better’ depends on the specific objective and conditions. Some systems require cooling, others require controlled warming, and the source provides no independent performance standard. A sound comparison states the readings and leaves the ranking unclaimed unless the task specifies a criterion. Practise replacing value judgments with named measurements.

Evidence-check 17: Different audiences, same evidence

For a young visitor, say that ice can absorb energy while melting at nearly the same temperature and that the example’s samples differed. For a teacher, state the energy equations, useful budget, observation intervals and missing controls. Both versions should identify the case as fictional and reject universal safety claims. Compare the two drafts line by line; if the public version asserts a new guarantee, it is not a successful simplification.

Evidence-check 18: A small correction has large consequences

Suppose the observation card is corrected from twelve to ten minutes for sample B. Against A’s eight minutes, the duration difference becomes two minutes rather than four. The relative increase becomes 25 percent rather than 50 percent. Every caption or committee note referring to these values must change. The modelled four-pack energy result remains unaffected because it belongs to a separate card. This demonstrates how a correction can have a local rather than universal effect.

Evidence-check 19: Name the model boundary

The calculation assumes all 170,000 J is useful energy available for the specified material journey. If the energy source loses some heat to the surroundings, fewer complete packs might be feasible. The example does not quantify losses, so do not invent a new count. State the condition under which the answer of four holds. Identify what new information, such as measured useful transfer, would be needed for a revised numerical result.

Evidence-check 20: The final claim-by-claim audit

Review the original three claims. ‘Twice as efficient’ fails both numerical ratio and undefined-efficiency checks. ‘Four packs proved by experiment’ mislabels an ideal model result as observation. ‘Keep medicines cold all afternoon’ makes an unsupported safety and duration claim. A valid replacement explains melting, states the separate modelled four-pack maximum, reports the eight-versus-twelve-minute observation where relevant, and avoids a new promise. This is the complete EMS evidence discipline in one exercise.

Small Group Tutorials

How to Perform in the new G3 SEC Examinations | Learner’s Guide Vol 0425 | EMS Cold-Pack Evidence Audit — Melting Energy, Feasible Counts and Honest Visitor Claims