A light may work while the argument for installing it fails
A school committee is considering a small educational status-indicator panel. It has two quotations, some electrical readings and a draft notice. Everyone likes the idea. The concern is whether the measurements really support the words in the notice and whether the line drawn through the price quotations can be trusted at a different quantity. These are not obscure problems. They are the kinds of reasoning errors that appear whenever a neat number is asked to carry more meaning than it can bear.
The purpose of this EMS case is to bring English, Mathematics and Science into one careful judgement. EMS is a practice method here, not an official combined G3 SEC subject. The actual assessment papers remain separate. A student learns to handle each subject’s concepts and then translate correctly between observations, models, claims and decisions.
The sequence begins with Vol 0322, English meaning in context, continues through Vol 0323, Physics I–V graphs, and takes its cost mathematics from Vol 0324, two-point line models. It does not repeat their complete lessons. It tests whether the learner can use them responsibly.
An accurate SEC examination boundary
The official SEAB directory for 2027 G3 syllabuses lists English Language K300, Mathematics K310 and Physics K323. The K300 syllabus specifies writing, comprehension, listening and oral communication; K310 Mathematics includes interpreting graphs and linear relations; K323 Physics includes the current–potential-difference characteristics of a constant-temperature metallic conductor, a filament lamp and a semiconductor diode.
English K300 has Writing Paper 1 and Comprehension Paper 2, each 1 hour 50 minutes, with separate Listening and Oral Communication components. K310 has two papers, each 2 hours 15 minutes. K323 uses multiple-choice, structured/free-response and practical assessment. The combined-science specifications K326 and K327 are not automatically identical to standalone K323. The official syllabus for the actual subject must decide the examination boundary.
All school descriptions and numbers below are created for teaching. The case is not an SEAB question, not a prediction and not a plan for pupils to install real equipment. It trains the habit of making a defensible conclusion from a mixed set of information.
The imaginary brief and four evidence packets
The proposed display has twelve indicator positions and printed labels. A staff member requests evidence of the electrical pattern, one typical measured operating point, the complete likely cost and a modest notice for visitors. Authorised adults would handle any real design review. No pupil should connect a component to mains or construct unapproved wiring to complete this exercise. Read and calculate from the supplied data.
Packet A is a metallic reference resistor measured at approximately constant temperature: (0 V,0 A), (1 V,0.05 A), (2 V,0.10 A), (3 V,0.15 A), (4 V,0.20 A). For matched non-zero readings, V/I = 20 Ω. Its I–V graph with current vertical is straight through the origin. This supports an ohmic model over the observed range, not a guarantee about arbitrary temperatures.
Packet B is a filament lamp with points (1 V,0.10 A), (2 V,0.16 A), (3 V,0.20 A), (4 V,0.23 A). Corresponding V/I ratios are 10 Ω, 12.5 Ω, 15 Ω and about 17.4 Ω. This is consistent with a filament whose resistance rises as it heats. The graph becomes less steep even though its current continues to rise.
Packet C is a fictional light-emitting diode measured in its forward direction at (1.4 V,0.1 mA), (1.6 V,0.4 mA), (1.8 V,3 mA), (2.0 V,12 mA) and (2.1 V,20 mA). The sharp nonlinear increase means a single constant-resistance line is not justified. These illustrative values are not the specified turn-on voltage, colour, maximum current or ratings of any real commercial LED.
Packet D is a separate low-voltage supervised observation: source potential difference 5.00 V, LED potential difference 1.98 V, path current 0.0100 A. Neglecting small wiring drops, a series resistor has 3.02 V across it, with indicated V/I about 302 Ω, consistent with a nominal 300 Ω part within ordinary measurement and tolerance considerations. The LED’s instantaneous electrical power is 0.0198 W; the resistor’s is 0.0302 W. These point calculations do not establish a complete safety rating.
The committee also has two fictional quotations: $38.80 for eight positions and $49.20 for twelve. The cost gradient is (49.20 − 38.80)/(12 − 8) = $2.60 per position. The line C = 2.60n + 18 matches both totals, where n is position count and C cost in dollars. At sixteen positions, the predicted cost is $59.60. A later illustrative quotation of $63.20 is $3.60 above this prediction. The residual calls for investigation, not accusation.
Why the three graphs cannot share one convenient law
An ordinary resistor at unchanged temperature behaves differently from a filament lamp whose temperature can rise and from a diode whose conduction depends strongly on polarity. The learner who extracts V/I from a single measurement and treats it as constant for every component has calculated a number but misunderstood what the graph says.
For the fictional diode, the point (2.0 V,0.012 A) gives V/I of approximately 167 Ω. At (2.1 V,0.020 A), the ratio is 105 Ω. The difference is evidence against one fixed-resistance description over that part of the forward characteristic. It does not imply the component is broken or that the lower numerical ratio measures its efficiency.
In school work, a diagram of a diode may also show negligible reverse current over the plotted negative-voltage interval. Describe the displayed asymmetry; do not invent its reverse breakdown threshold. Any real LED circuit needs appropriate component ratings, current limitation and competent supervision. A current–voltage plot is not a construction instruction.
How to check a budget without calling a discrepancy a crime
The line C = 2.60n + 18 gives an algebraically precise prediction. It does not give a legally fixed future price. The $18 intercept might represent a panel, setup charge or an accounting allocation, but the two totals cannot identify that component’s nature by themselves. Only an itemised quote can do that.
The sixteen-position prediction is outside the observed range of eight to twelve. Extrapolation assumes the same terms continue. If the later quote says $63.20, the positive residual of $3.60 could come from a different pack size, added installation work, a new delivery charge or changed prices. A useful writer says what disagrees and what should be checked.
If a purchaser has a maximum available budget of $55 and needs twelve positions, the fitted line suggests $49.20 leaves $5.80. Yet this is only meaningful if installation, safety review and all required materials are included. A seemingly comfortable margin can disappear once omitted costs are restored. Mathematics and honest language belong together.
Four kinds of sentences in a defensible report
An observation is something the evidence explicitly records: ‘At 2.0 V, the illustrative diode current was 12 mA.’ A calculation is a transformation: ‘This gives about 167 Ω as the V/I operating ratio.’ An interpretation draws a limited inference: ‘The ratio changes with voltage, so the behaviour is not ohmic across these sample points.’ A recommendation tells a responsible reader what to do next: ‘Check the part specification and installation requirements before any purchase.’
These four sentences look similar on a page but carry different authority. When a report says ‘proves safe’, ask which observation could establish every relevant hazard, environmental condition and long-term operating limit. The packet does not contain those observations. The honest expression ‘needs technical review’ is not evasive; it accurately preserves responsibility.
English is also about audience. A visitor needs one clear line explaining a display; a purchaser needs a transparent cost breakdown; a Physics teacher needs units and a mechanism. The same evidence can serve different readers if the writer changes the explanation without changing the facts.
Forty-six worked EMS audits
These original tasks move from basic ratios and units to more advanced distinctions between model, measurement and commitment. They are not copies of examination questions. Try each one on paper and give the explanation in your own words before comparing it with the answer.
01. Does increasing voltage alone demonstrate Ohm’s law?
No. Ohmic behaviour requires the current to be directly proportional to potential difference while relevant physical conditions remain unchanged. Compare matched ratios or the straightness and origin of the plotted I–V graph. A rising current is not enough: a curved filament-lamp graph rises too.
02. Find the reference resistor’s resistance at 2 V.
The matched current is 0.10 A, so R = 2/0.10 = 20 Ω. Check a second point such as 4 V,0.20 A and recover the same ratio. Its measured constant ratio supports the reference model, but it does not guarantee that the resistor never changes with temperature.
03. What does the gradient of the reference I–V line mean?
With current on the vertical axis and voltage on the horizontal, gradient is I/V = 0.05 A/V. For a straight through-origin ohmic characteristic, resistance is the reciprocal, 20 Ω. If you say ‘gradient equals resistance’ without reading the axes, you have inverted the relationship.
04. Identify the unusual feature in the filament readings.
Voltage rises by one volt each step, but current additions fall from 0.06 A to 0.04 A to 0.03 A. The graph becomes less steep. The current itself still rises. The pattern is consistent with a hotter metal filament having greater resistance.
05. Why is the 4 V lamp’s resistance not 10 Ω?
At 4 V the current is 0.23 A, so its operating ratio is 4/0.23, approximately 17.4 Ω. Ten ohms belonged to the 1 V,0.10 A point. Carrying a previous point’s ratio into a new condition is the mistake the characteristic curve is meant to prevent.
06. Convert a diode reading of 12 mA.
Twelve milliamperes equals 0.012 amperes. A power or resistance equation with volts and SI units needs this conversion unless every unit is explicitly adjusted. Treating 12 mA as 12 A would inflate the calculated current by a factor of one thousand.
07. Calculate the diode’s V/I at 2.0 V.
Use 2.0/0.012 ≈ 167 Ω. This is the ratio for one operating point, not the constant component resistance across its whole characteristic. A diode’s curved response makes the distinction between one point and an entire law essential.
08. Calculate the diode’s V/I at 2.1 V.
Use 2.1/0.020 = 105 Ω. Since 105 differs from about 167, the two points are not described by a constant V/I value. Do not call this change a negative resistance or a failure of arithmetic; it reflects nonlinear behaviour.
09. Can the two diode points be joined by a straight line?
They can define a line between themselves, because any two distinct coordinate points can. That mathematical fact does not establish that the other diode readings follow the line. Compare the complete data set, especially the lower-voltage measurements.
10. What can a reverse-bias diode graph show?
It can show very small current in one polarity over the supplied plotted interval. State that observed direction-dependent response. Avoid absolute words such as ‘impossible forever’ or invented breakdown numbers that are not provided by the question.
11. Does an electrical curve tell us the LED’s colour?
No. Wavelength, optical output and component characteristics need additional information. A hypothetical current–voltage graph cannot by itself identify an LED as red, green or yellow. The writer should not smuggle a visual claim into a technical table.
12. Does it prove the device will last five years?
No. A short trial has no lifetime test, warranty or environmental assessment. The correct conclusion concerns the measured point and trend. A maintenance plan and rated-life evidence would be separate checks before a durability promise.
13. What is the voltage across the series resistor?
Subtract the measured LED drop from the supply: 5.00 − 1.98 = 3.02 V, neglecting small wiring drops. The calculation is simple, but its meaning is component-specific. Calling 3.02 V the LED voltage would be wrong.
14. Estimate the resistor value from the trial.
R = 3.02/0.0100 = 302 Ω. This is reasonably consistent with a nominal 300 Ω resistor, given that readings and parts have tolerances. It would be unsound to claim the component has exactly 302 Ω under all conditions.
15. Calculate the LED electrical power.
P = VI = 1.98 × 0.0100 = 0.0198 W at the measured point. This is neither annual energy consumption nor an optical brightness measure. To calculate energy you would also need a defined operating duration and an appropriate power assumption.
16. Why is 0.0198 W not proof of safety?
Electrical power at one point omits many considerations: ratings, thermal environment, mounting, supply changes, current limits and supervision. Safe equipment selection is the responsibility of qualified reviewers. A low number in one calculation is not a complete assessment.
17. Should a student test this panel from household mains?
No. The case is a written exercise using fictional data, not permission for a construction trial. Any practical classroom work must use authorised low-voltage setups, protective components and competent supervision. Do not improvise dangerous connections to verify a graph.
18. Reconstruct the cost gradient.
The change in quotation is $49.20 − $38.80 = $10.40 for four extra positions. The candidate gradient is $2.60 per position. The unit explains the number. It is not the total price or a guaranteed rate for every future order.
19. Reconstruct the cost intercept.
For eight positions, 38.80 = 2.60 × 8 + c, so c = 18. This mathematical intercept suggests a fixed charge. Ask for an itemised quotation before saying whether it is board printing, installation or something else.
20. Is the cost model direct proportion?
No. C = 2.60n + 18 is linear with a non-zero intercept. Doubling the position count does not double the entire cost because the fixed $18 remains. ‘Linear’ and ‘directly proportional’ are different claims.
21. Calculate the model’s twelve-position total.
The variable part is 2.60 × 12 = $31.20, plus the fixed $18, giving $49.20. A student who reports only $31.20 may have done one multiplication correctly while omitting the most important condition of the cost model.
22. Estimate cost at sixteen positions.
C = 2.60 × 16 + 18 = $59.60. The number is exact for the proposed line, but it is an extrapolation because sixteen exceeds both observed order quantities. Preserve that limitation in a recommendation.
23. Calculate the new quote’s residual.
Observed minus predicted equals $63.20 − $59.60 = +$3.60. The positive sign means the later quote lies above the model. It does not by itself determine why, and it cannot establish intent or wrongdoing.
24. What can explain the price discrepancy?
An extra delivery fee, different pack sizes, a changed product, another installation condition or supplier pricing could all matter. The responsible next action is to request conditions and line items, not to announce an accusation from two numbers.
25. What does C(0) = 18 mean?
It is the line’s y-intercept at zero positions. It might model a fixed setup fee, but a supplier may not actually accept a zero-unit order. Mathematical extension and real commercial policy need not coincide outside the quoted domain.
26. Estimate cost for ten positions.
C = 2.60 × 10 + 18 = $44.00. Ten falls between eight and twelve, so the estimate is interpolation. It still assumes the two price points describe an appropriate rule for quantities in between.
27. Does a budget margin guarantee feasibility?
No. A proposed total below the available budget is useful only if the estimate includes all required work. Missing permits, parts, labour or safety assessment can change the real cost. Affordability and safe feasibility are separate tests.
28. Compare two quotations that include different work.
A lower-looking price may exclude a panel, installation, warranty or testing. Before declaring one quote better, put both on a like-for-like scope. A numerical ranking cannot replace reading the conditions in the small print.
29. What is a model residual rather than a calculation error?
A residual compares an observation with a model prediction at the same input. It may reveal variation, a changed assumption or omitted costs even when both arithmetic steps are correct. Investigate what differs; do not erase the inconvenient observation.
30. Why do two perfect cost points not validate a universal rule?
Two distinct points always define a straight line, so perfect agreement with those construction points is inevitable. A third independent quotation is the actual test of continued linearity. This is the same mathematical caution developed in Vol 0324.
31. Write a factual sentence about the sample diode.
‘The supplied example shows a rise from 12 mA at 2.0 V to 20 mA at 2.1 V.’ This sentence reports the data without claiming the diode will behave identically in every product or environment.
32. Write an interpretation about that sample.
‘The changing V/I ratio indicates the forward response is not described by a single constant resistance across these points.’ This sentence makes a limited inference that can be checked against the numbers.
33. Write a recommendation based on the sample.
‘Verify the intended LED’s manufacturer ratings and technical design before any real installation.’ The recommendation is useful because it identifies a specific missing check rather than claiming the demonstration itself certifies safety.
34. Rewrite ‘the display is guaranteed to be safe’.
A safer sentence is: ‘The supplied calculations describe a sample operating point; installation should proceed only after an appropriate technical and safety review.’ The revision does not pretend to know what the evidence has not established.
35. Rewrite ‘every additional station costs $2.60’.
Use: ‘The two initial quotations fit a candidate model with $2.60 variable cost per position, subject to the same pricing conditions.’ This keeps the estimate without turning it into a promise about future orders.
36. What if a chart has no numerical scales?
It may support qualitative conclusions about a straight or curved response, but not a precise resistance or current value. Do not import numerical values remembered from another textbook. That would make an invented measurement look official.
37. Can a current in milliamperes be used directly as amperes?
Only if you convert appropriately. The conversion factor matters in calculations of V/I and VI. A current of 20 mA is 0.020 A; wrong units can make a plausible-looking calculation physically absurd.
38. A teacher asks whether current is ‘energy per second’.
Current is charge per second, I = Q/t. Power is energy transfer per second; electrical power can be calculated as P = VI. A correct report uses the right physical noun and unit, not a loose everyday synonym.
39. What if the LED trial voltage differs from another example?
Do not assume the measurements came from identical components, test conditions or sources. The forward characteristic is an example set and the 1.98 V operating point is from a separate supplied trial. Keep datasets distinct unless the task explicitly connects them.
40. What does an ammeter measure in a series path?
It records the current through that path. A voltmeter across an LED measures the potential difference across the LED, not automatically the entire supply. Match the quantity to the instrument’s location before using a formula.
41. What would an auditor do with a suspiciously low total?
Recalculate the variable subtotal, add the fixed component, then check whether installation and accessories are included. If the written total still differs, identify the exact stage at which the calculation or assumption changed.
42. Should a public school notice list all the numbers?
Not necessarily. It should state the exhibit’s purpose and any limits needed to prevent misleading claims. The technical report should retain the numerical evidence. Different readers need different detail, but both deserve accuracy.
43. Can the committee call the future quote ‘certain’?
No. It can call it a forecast under a named pricing assumption. A new quote and relevant scope details are needed before committing money. Being precise about the future is more useful than sounding assured.
44. Why is this not a new EMS SEC paper?
Because the 2027 official system specifies subject-level assessments. This activity is a cross-subject study exercise that strengthens separate examinable skills. The source of examination format remains the SEAB syllabus directory, not a fictional school story.
45. What does a complete approval recommendation contain?
It identifies supported data, the model and residual, missing technical specifications, and the person or process responsible for the next checks. A recommendation that ends with ‘buy it’ without conditions is incomplete for the evidence given.
46. What is the final self-test?
Ask whether each number has the right unit, each inference a matching observation, each model a stated domain, and each action a clear owner. If those links survive, the proposal is usable even if some decisions must remain open.
A complete, readable report for the hypothetical committee
The sample metallic conductor readings are consistent with a fixed 20 Ω resistance over the tested range, whereas the filament-lamp and forward-diode readings show changing operating resistance. In particular, the forward-diode example does not support a universal turn-on voltage or a fixed-resistance model. The separate low-voltage trial records a 1.98 V diode drop at 0.0100 A, giving approximately 0.0198 W of instantaneous electrical power at that point. These calculations are not a full electrical safety assessment.
The two initial quotations fit C = 2.60n + 18, where n is the number of positions and C is cost in dollars. This predicts $49.20 for twelve positions and $59.60 for sixteen. A later $63.20 quotation for sixteen is $3.60 higher than the model prediction. The committee should obtain an itemised explanation before treating the initial rate as binding.
The display merits further technical and commercial review rather than immediate approval from the limited information supplied. Verify manufacturer component ratings, competent low-voltage installation arrangements, the full scope of work, maintenance responsibility and complete costs. This is a stronger recommendation than an absolute promise because every claim can be traced to evidence or to a named next step.
Check the direction of every transfer
When a learner moves a number from a Physics table into a cost report, ask what the number measures and whether it belongs there. An ampere is a flow of charge per second, whereas a dollar is a monetary cost; neither is a count of devices or a rate of maintenance failure. Units help keep the translation honest even when all values are printed in one attractive spreadsheet.
Likewise, the phrase ‘increase in current’ does not automatically mean ‘increase in brightness by the same percentage’. A diode’s optical performance is not supplied by these electrical numbers alone. Do not use a related scientific measurement as a substitute for the one a decision requires.
A fair test for a new quotation
A disciplined committee would request a third quotation with the same definition of a position and the same inclusions as the original two. It would compare the new amount with C = 2.60n + 18 and record the residual. If the cost structure changes beyond a quantity threshold, a piecewise model may be more appropriate than one line.
The additional quotation should be treated as data, not as a competitor to a cherished prediction. The original model may have been useful for a preliminary estimate even if the newer information reveals its limits. Revising a model because evidence has improved is good reasoning.
A safety statement needs its own evidence
The low-voltage trial contains one measured current and some component voltage readings. It does not describe every transient, every wiring fault, thermal operating limit, physical enclosure or maintenance condition. An official installation judgement therefore needs the appropriate documentation and expertise.
Students can practise calculating and explaining these values without turning the lesson into practical electrical advice. The report should say where science class analysis ends and a competent authorised review begins. Keeping that boundary is part of responsible learning.
Separate certainty from confidence in a public paragraph
A model can be mathematically certain about its own arithmetic and uncertain as a description of future reality. Under C = 2.60n + 18, the value at sixteen is exactly $59.60. Whether a supplier will quote that amount is a different question, and the additional $63.20 quotation provides a reason to reconsider the assumption.
Students often think qualifications make a sentence weak. In a well-written report, a phrase like ‘under the initial pricing model’ makes the claim stronger because it tells readers exactly when the result applies.
Prepare for transfer to a different real-world case
Suppose a future school exercise replaces an LED with a water filter, attendance survey or science-garden sign. The domain changes, but the checks do not: define quantities, examine units, test the mathematical model against more data, separate observation from mechanism and identify who should act.
This is the connection between school examinations and good everyday judgement. The learner is not memorising a single display’s answer. They are learning how to prevent an unsupported inference when information crosses from one discipline into another.
Rehearse an explanation without notes
Close the example and explain why the metallic conductor, filament lamp and diode cannot be treated as the same fixed resistor. Then reconstruct the budget model from the two quotations, calculate the third residual, and name one missing technical and one missing commercial check.
If you can do those tasks without copying the final paragraph, you have transferred the learning. If you can only recite the wording, return to the measurements and rebuild the steps independently. Practice should test ability, not familiarity with the page.
Return to the separate learning routes
The English owner develops accurate reporting and comprehension. The Mathematics owner develops graph interpretation, formulas and testing models. The Science owner develops electrical reasoning and valid investigation. Students who need earlier foundations can use the PSLE Learning Guide and the Secondary 1 Mathematics foundation reference.
An informed recommendation is one whose facts, calculations, limits and next responsibilities are visible. That is a useful result even when the correct final decision is to request better evidence.
