Quick answer: use a yield ledger, then test every constraint
An impressive gross total can hide a small useful amount. A high percentage can hide a very small starting batch. A correct calculation can answer the wrong question if a table mixes total output, target fraction, retained amount and rejected material. This learner’s guide teaches how to read such records as mathematics rather than as advertising.
Everything in the worked setting is fictional. Imagine a paper-only educational model of a biological manufacturing line. Its ‘units’ are arbitrary counters used for exam-style calculations; they are not doses, chemical quantities, concentrations, medical specifications or production instructions. This volume makes no judgement about the safety, purity, efficacy or clinical use of any medical product.
The quick method is to choose the right starting amount, express every stage as a fraction that is kept, multiply retained fractions in sequence, and solve backwards only after checking the target’s meaning. The last step is to test whether the result is feasible when batches must be whole numbers. The purpose is to turn a long table into a traceable model that a reader can verify.
The earlier Volume 0908 owns the biological gene-to-protein explanation. Volume 0905 owns concentration and weighted-mixture comparisons. Here the specific mathematical intent is sequential yield accounting, retained fractions, inverse targets and discrete capacity decisions. No medical or biological claim follows from the arithmetic.
The official 2027 G3 Mathematics SEC syllabus is K310, with the earlier reference code 4052. Both Mathematics papers are 2 hours 15 minutes, 90 marks and 50% of the subject grade. Paper 2 ends with a real-world application problem, and essential working matters. This invented modelling task practises percentage, ratio, algebraic equations, data interpretation and decision reasoning; it is not a claim that SEAB will set a biotechnology question.
Thirty-two worked decisions from first percentage to final feasible plan
1. First identify what the number counts
A record says that a fictional batch contains 800 units. Before calculating, ask whether the units represent gross output, target component, material retained after inspection or approved educational tokens. These quantities are different even when they share the word ‘units’.
Suppose 800 means total tokens and 65% are classed as target tokens. The target amount is 800×0.65 = 520. It would be wrong to present 800 as the target amount merely because it is the largest value in the record.
Make a small key beneath the problem: G for gross, f for target fraction and T for target. Write T = Gf before inserting numbers. Symbols are useful when they preserve meaning rather than merely shortening a sentence.
For practice, explain why ‘800 units collected’ cannot answer ‘How many target units were retained?’ until at least two additional definitions are supplied: the fraction of target units and the retention process.
2. Percent means out of one hundred
A 65% fraction is 65/100, which is 0.65. Multiplying by 65 rather than 0.65 creates an answer one hundred times too large. This error often survives because the student forgot to check whether the result can exceed the initial total.
If the gross quantity is 800, a portion of 65% must lie between zero and 800. The answer 52,000 violates this basic boundary and should be rejected without needing an elaborate proof.
Convert the percentage into a decimal or fraction before calculating. Then perform a reasonableness check: a part of a non-negative total cannot be larger than that total when the fraction is at most 100%.
Practise changing 65% to 0.65, 6.5% to 0.065 and 0.65% to 0.0065. Apply each to 800 and explain why the results differ by factors of ten.
3. Percentage points are not percent change
If a target fraction rises from 60% to 70%, the change is 10 percentage points. Relative to the original 60%, the percentage increase is (70−60)/60×100%, or about 16.7%.
A report that says ‘improvement of 10%’ when it means ten percentage points may mislead the reader. The two metrics use different denominators and answer different questions.
State both terms only when needed. A clear answer could be ‘The target fraction increased by 10 percentage points, which is a relative increase of about 16.7% from the original fraction.’
Try a change from 80% to 84%. Calculate the 4-percentage-point difference and the 5% relative increase. Explain why the latter is not 4%.
4. Two stages multiply; they do not add
Suppose 80% of fictional material remains after the first review, and 75% of that remainder survives a second review. The overall retained fraction is 0.80×0.75 = 0.60, or 60% of the original.
It is incorrect to subtract 20% plus 25% from the starting amount, because the 25% second-stage rejection is applied to the already reduced material. The base changes after the first stage.
Write stage one output = initial×0.80. Stage two output = stage one×0.75. Substitution shows final = initial×0.60. This makes the order and denominator of each percentage visible.
For an initial 500 educational units, calculate 400 after stage one and 300 after stage two. Check that the total loss is 200, or 40% of the original, not 45%.
5. Retained fractions explain loss precisely
A stated loss of 12% corresponds to a retained fraction of 88%, or 0.88. The retained amount can therefore be found directly by multiplying the starting value by 0.88.
The wording ‘12% lost’ is not the same as ‘12% remains’. The two conditions produce very different answers, especially in a multi-stage problem.
Translate each sentence into a positive retained fraction before multiplying. For a 1,000-unit fictional batch with a 12% loss, the answer is 880 units retained. Then calculate the 120-unit difference as a check.
Create a two-column table showing percentage lost and percentage retained for losses of 5%, 12%, 37% and 100%. Explain what happens to the retained amount at the extremes.
6. Why stage order matters to interpretation
Mathematically, multiplying two independent fixed fractions is commutative: 0.8×0.75 = 0.75×0.8. But a real problem may attach different meanings, capacities or selection conditions to stages.
Do not infer that the stages can be physically rearranged merely because their product is unchanged under a simplified model. The mathematical calculation describes fractions applied in sequence, not the feasibility of changing a process.
If a question states that the second stage can handle at most 400 units, the quantity reaching it matters. A sequence starting from 600 with 80% retained produces 480, exceeding that stated limit.
Write both the numeric overall fraction and a separate sentence about capacity. Demonstrate why the same final theoretical fraction does not prove that every process order is feasible.
7. Three stages make a visible chain
An invented sequence retains 90%, then 80%, then 50% of what enters each successive stage. The overall fraction is 0.9×0.8×0.5 = 0.36. From 1,000 starting units, 360 remain.
Adding the retention percentages, averaging them or treating the last 50% as a fraction of the original will give the wrong answer. Each stage has its own incoming base.
Build a small ledger: 1,000 at entry; 900 after the first stage; 720 after the second; 360 after the third. The ledger should agree with the single multiplication expression.
Change the last stage to 75%. Predict whether the output grows or falls, then calculate 540 units. State why the comparison isolates the effect of one altered fraction in the fictional model.
8. The algebraic reverse target
Suppose the model must deliver at least 480 final units after two stages retaining 80% and 75%. Let x be the required starting amount. Then 0.8×0.75×x ≥ 480, so 0.6x ≥ 480 and x ≥ 800.
A student who simply adds a 20% and 25% allowance to 480 obtains 696, which would produce only 417.6 under the stated sequential fractions. Forward checking exposes the mistake.
When solving backwards, divide by the product of retained fractions rather than adding a guessed margin. Write the inequality if the requirement is a minimum, not an exact equality.
Verify x=800 by forward substitution. Then test x=799 to see why it falls below the threshold in this idealised continuous calculation. Explain how whole-number batch rules might change the choice further.
9. Minimum means at least
If a question asks for at least 480 units, the inequality is final output ≥ 480, not final output = 480 as the only acceptable possibility. A larger feasible final amount is allowed.
Students can obtain the algebraic boundary and still misinterpret it by describing every higher result as invalid. Understanding the inequality is as important as isolating the variable.
Solve the boundary first, then describe the feasible region in words. In the example, any sufficiently large starting amount meets the pure yield inequality, subject to any additional limits given.
Make a number line with 800 as the boundary in the simplified model. Mark 790, 800 and 830 as failing, meeting or exceeding the requirement. Then introduce a maximum-capacity condition to see why the feasible set may shrink.
10. Whole batches demand ceiling logic
If one whole fictional batch produces 72 final units and a target requires 500 units, then 500÷72 ≈ 6.94. Seven whole batches are needed because six produce only 432 units.
Rounding the quotient to the nearest integer can be unsafe. If the answer were 6.49, rounding to six would still fail a minimum target. This is a ceiling decision, not ordinary approximation.
After finding a quotient, check the two neighbouring integers. For this example, six is insufficient and seven yields 504. Present seven with the explanation that batches cannot be divided.
Change the target to 504 and 505. Show that the first requires seven batches and the second eight. This boundary exercise reveals why a small change in a target can alter a discrete operational decision.
11. Maximum capacity creates another gate
Suppose a fictional storage room can accept at most seven final batches, while the target needs eight. Even if an algebraic yield calculation suggests eight will meet the quantity requirement, the proposed plan violates capacity.
An exam answer about feasibility must combine every relevant constraint. Satisfying the output threshold does not automatically satisfy space, budget or time limitations.
Write each condition separately: batches ≥ required minimum; batches ≤ available capacity. If the lower bound is eight and the upper bound is seven, no feasible integer meets both.
Practice interpreting ‘no feasible solution’. It does not mean the equations were solved incorrectly; it means the stated constraints cannot all be satisfied simultaneously.
12. Do not confuse input and final capacity
A stage described as accepting at most 600 incoming units is not necessarily a limit on the final output. If 75% remain, the corresponding downstream amount is at most 450 under the given model.
Students who attach a capacity limit to the wrong stage may accept a plan that exceeds the input limit. The phrase ‘per stage’ requires careful attention to where the number is measured.
Draw boxes for entry, stage one, stage two and final output. Put the capacity notation beside the correct box or arrow. Then substitute the incoming amount at that point.
For a 700-unit starting load with 90% passing the first stage, 630 enter the second. A stated second-stage maximum of 600 is breached even if final output later falls below 600.
13. A rate needs a time denominator
A fictional line processes 240 units in six hours, giving an average processing rate of 40 units per hour. A second line handling 200 units in four hours averages 50 per hour.
The second line has a higher rate although it processed fewer units in total. ‘Faster’ and ‘larger total output’ are different descriptions and must not be used interchangeably.
Use rate = quantity ÷ time, keeping time units consistent. If one period is in minutes and the other in hours, convert before comparing.
Practise by changing the six-hour period to 360 minutes. The first rate becomes 2/3 of a unit per minute, which equals 40 units per hour. Explain the conversion rather than moving decimal points by habit.
14. Average rate is not instantaneous rate
If a model reports 240 units over six hours, the average rate is 40 per hour. It does not follow that exactly 40 were processed during each individual hour.
A table that contains only beginning and ending totals cannot show the path between them. Some intervals may be faster and others slower. A claim of perfectly constant output needs additional evidence.
Where a question provides interval data, calculate each interval separately. If it gives only an overall total and duration, describe the average, not a measured hour-by-hour rate.
Construct hourly increments of 20, 30, 60, 70, 40 and 20 units. They total 240, yet the hourly rates differ. Explain why the same overall average hides the variation.
15. Idle time changes the interpretation
An eight-hour shift might include two hours when the fictional line is inactive. The gross rate based on scheduled time differs from the operating rate based on active time.
If 360 units are processed in six active hours across an eight-hour shift, the active rate is 60 per hour while the whole-shift average is 45 per hour. Neither figure is inherently wrong; the denominators differ.
Name the denominator in the answer. Use ‘per scheduled hour’ or ‘per active hour’ rather than leaving ‘per hour’ ambiguous where both possibilities appear.
For a planning question, identify which rate is relevant to estimating end-of-shift output. Explain whether the question assumes the same proportions of active and idle time in future shifts.
16. Weighted totals need compatible units
A table may give two batches measured in tokens and one entry measured in hundreds of tokens. Adding 500, 6 and 700 directly creates a false total unless the notation is converted.
Convert all quantities to a single agreed unit before calculating. Six hundreds of tokens means 600 tokens; the compatible total is 500+600+700 = 1,800.
Show the unit column beside every intermediate value. When a large number seems to change unexpectedly, inspect whether a prefix, multiplier or ‘per batch’ description was overlooked.
Practice with entries 0.8 thousand, 250 units and 3 hundreds. Convert each to units, then add 800+250+300 = 1,350. Explain the source of any common factor-of-ten or factor-of-one-thousand errors.
17. An average of percentages can mislead
One fictional batch of 100 units retains 90%, while a batch of 900 retains 50%. The unweighted average of 90% and 50% is 70%, but the pooled retained fraction is (90+450)/1,000 = 54%.
The difference comes from unequal starting sizes. The larger batch contributes more to the total material and should influence the combined percentage accordingly.
To pool, convert each batch percentage into a retained amount, add the retained amounts and divide by the combined starting total. Keep the batch denominators visible.
Change the second batch to 100 units. Recalculate the pooled fraction and explain why the unweighted average now works: both initial batch sizes are equal.
18. Mean, median and range describe different things
Five fictional batch outputs are 60, 70, 70, 80 and 220 units. Their total is 500, so the mean is 100; the median is 70 and the range is 160. These values summarise different aspects of the same data.
A headline saying ‘typical batch output is 100’ may misrepresent the middle of this skewed set, depending on the intended meaning of typical. The unusual 220 raises the mean.
Show all three calculations, then state which summary answers the question. If the task asks for spread, the mean alone is insufficient; if it asks total capacity, the median alone may not help.
Replace 220 with 90. Recalculate the mean, median and range. Explain which statistic changes most and what the exercise shows about extreme values.
19. Missing rows are not failed batches
A table with four measured batches and one blank record contains a missing value, not necessarily a recorded zero. Substituting zero changes the mean and can create a false performance judgement.
If the four recorded outputs are 60, 70, 80 and 90, their observed mean is 75. The full five-batch mean is unknown unless the missing fifth value is supplied or bounded.
State the denominator and the nature of missing information. Say ’75 among the four recorded batches’, not ’75 for all five’. Do not invent a cause for the blank entry.
Practice comparing a blank, a true zero and a rejected record. Each has a different evidential meaning and may require a different calculation under the question’s instructions.
20. Uncertainty is not a hidden safety allowance
A fictional output of 80 ± 5 tokens represents an illustrative interval from 75 to 85 under the stated assumption. It cannot automatically be replaced by 85 when checking a minimum requirement.
For a guaranteed-minimum style planning question, the lower bound is the relevant conservative value if the task requires all allowed outcomes to meet the threshold. For a different question, another interpretation may apply.
State the assumption about what the interval means before calculating. If seven batches each may yield as little as 75, the lower-bound combined result is 525.
Compare a target of 530 with the seven-batch lower bound. Explain why the point estimate of 560 does not by itself satisfy a worst-case guarantee and why the problem’s wording decides the test.
21. Rounding belongs at the end
If a calculation involves several fractional stages, rounding each intermediate value too aggressively changes the final result. With integer final-batch requirements, premature rounding can change the answer by one full batch.
Keep sufficient precision through intermediate steps, and round only where the prompt instructs or where the final interpreted unit requires it. A displayed rounded estimate should be distinguished from an exact algebraic boundary.
Suppose each batch yields 72.4 fictional units under an idealised model and the target is 500. The quotient is about 6.906; whole batches still require seven. The final decision comes from the inequality and discreteness.
Write the unrounded quotient beside the rounded result. Check whether the rounded value actually meets the target. This habit prevents a neat-looking answer from defeating the problem’s constraints.
22. The inverse plan can fail for other reasons
A calculation may show that ten starting batches are required for a desired output. The problem might also impose a maximum of eight available batches. There is then no feasible plan under the existing model.
Students should resist ‘fixing’ an infeasible result by lowering the target silently or claiming an unavailable extra batch. The purpose of a constraint question is often to identify precisely this conflict.
State which inequality fails. Then, only if the task asks for alternatives, propose a mathematically defined change such as revising the target or increasing stated capacity.
Practice writing the conclusion in a complete sentence: ‘The minimum required is ten whole batches, but only eight are available, so the target is infeasible under the given conditions.’
23. Compare two plans on the same basis
Plan A begins with 1,000 fictional units and retains 60%; plan B begins with 800 and retains 75%. Both end with 600. Plan B has a higher retained fraction but not a larger final output.
The word ‘better’ needs a criterion. If the aim is maximum final output, they tie. If it is a higher fraction retained, B wins. If it is lower resource use, the question needs appropriate resource data.
Create a table with starting amount, retained fraction and final amount. Add cost or time only if the problem supplies them. Do not award a plan an advantage based on an invented measure.
Practice stating two different, non-contradictory comparisons from the same pair. This teaches the discipline of keeping the performance criterion visible.
24. Cost per final unit needs the final denominator
Suppose a fictional plan costs $900 to generate 600 final educational units, while another costs $800 to generate 400. Costs per final unit are $1.50 and $2.00 respectively.
The cheaper plan in total is not the cheaper plan per final unit. Conversely, a lower unit cost does not prove greater affordability if the required upfront cash is restricted.
Divide cost by the quantity specified by the question, and name the denominator. A gross-output unit cost may differ from a final-output unit cost after losses.
Change the target to only 350 units and impose a budget of $850. Reconsider the feasible plan without assuming that the lower per-unit cost alone decides the result.
25. A minimum budget and a maximum budget are different
Financial feasibility questions usually give a maximum spending limit and a minimum output requirement. These produce inequalities pointing in opposite directions: cost ≤ budget and output ≥ target.
A student who writes both inequalities with the same sign may arrive at a mathematically neat but logically impossible recommendation. Translate the words before using algebra.
For a cost of $120 per whole batch and budget of $750, at most six batches are affordable. If the output target requires seven batches, the plan fails the budget gate.
Show the two calculations separately, then compare the integer limits. Explain why the contradiction does not disappear merely because the fractional quotient of budget divided by cost is 6.25.
26. Find the bottleneck rather than inventing gains
An invented two-stage line can process at most 500 units at stage one but only 300 at stage two. The smaller capacity limits what can proceed through both stages under otherwise comparable assumptions.
Summing stage capacities gives 800, which describes neither the throughput of the sequential line nor its maximum final output. Sequential stages do not add their per-period capacity in that way.
Sketch a flow with arrows connecting the stages. Ask how much reaches each stage and which capacity binds. A correct model exposes where the conflict occurs.
Practice changing stage two’s capacity from 300 to 450 while leaving stage one fixed. State why the line’s capacity rises but does not become 950.
27. A scenario with two product classes
A fictional batch includes target and non-target tokens. If 700 total units contain 420 target units, the target fraction is 420÷700 = 60%. If 70% of those target units are retained, the final target amount is 294.
Applying the final 70% to all 700 units produces 490, which answers a different question. The first stage is classification; the second is retention of the designated target group.
Use a two-level ledger: gross → target → retained target. Label every percentage with the quantity to which it applies. This prevents accidentally moving to a different denominator halfway through the problem.
Create a new example with 800 gross units, 75% target and 80% retention of target. Calculate 480 final target units, then verify 800×0.75×0.80 gives the same answer.
28. Reverse a two-stage model with integer batches
A whole initial batch contains 200 fictional units, 75% are target, and 80% of target remain after review. Each whole batch therefore gives 200×0.75×0.80 = 120 retained target units.
If the final requirement is at least 650, divide 650 by 120 to get about 5.42. Six initial batches are required, giving 720 retained target units under the model.
The answer must state what constitutes a batch. Rounding 5.42 down to five would produce only 600, which fails. Rounding the number of retained units rather than the number of batches is also a different operation.
Test both neighbouring integer choices. Five fails and six succeeds. Then explain why the result cannot be used as a real product-release calculation without additional evidence and definitions.
29. Averages require a complete observation window
A seven-day table with only five reported days may describe 350 units over five observed days, an observed average of 70. It cannot automatically imply a seven-day total of 490.
Projecting from the observed average adds an assumption about the missing days. This may be reasonable in a labelled model but is not a fact recorded in the table.
Separate the measured statement from the projection. Write ‘The five observed days total 350; if their average were maintained for seven days, the projected total would be 490.’
List two reasons the projection could differ from reality, such as different active hours or different batch sizes. Name the uncertainty without inventing a precise probability.
30. An algebraic condition can determine feasibility
Suppose x whole batches cost $90 each and yield 60 final educational units each. The requirements are at least 450 output units and no more than $720 spent.
The output condition 60x ≥ 450 implies x ≥ 7.5, so x must be at least eight. The budget condition 90x ≤ 720 implies x ≤ 8. Thus the only feasible whole-batch solution is x = 8.
Show both inequalities and why x is restricted to an integer. An algebraic interval is not sufficient as a final answer if fractional batches are not permitted.
Now lower the budget to $700. The maximum affordable whole batches becomes seven, while the output condition still needs eight. State the resulting infeasibility clearly.
31. Use tables to check an algebraic answer
After solving a long equation, a small table of candidate integer values can catch mistakes. In a two-condition problem, test output and cost for each plausible number of batches.
For x = 7 in the preceding example, output is 420 and cost is $630; output fails. For x = 8, output is 480 and cost is $720; both pass. For x = 9, output is 540 but cost is $810; budget fails.
The table is not a substitute for the inequality solution; it is a verification step that explains the feasible boundary to a reader.
Practice including the headings ‘x’, ‘output’, ‘cost’, ‘meets target?’ and ‘within budget?’ in one compact table. Explain why the row satisfying every condition is the final decision.
32. Interpret numbers without making medical claims
A problem may use a biotechnology-themed setting because it supplies a meaningful sequence of percentages. The mathematical conclusion is about the invented quantities, not about the purity, safety or efficacy of an actual medicine.
The correct ending states the number of whole fictional batches required under the assumptions. It must not turn ‘480 target tokens’ into a clinical dose or claim that the product has been approved.
Use ‘under the stated model’ where the assumptions matter. This signals both mathematical accuracy and appropriate boundaries of inference.
Write an answer with a valid numeric result and one sentence identifying its scope. Then remove any sentence that makes an unsupported assertion about real manufacturing or treatment.
33. How to diagnose a recurring error
If the answer is repeatedly off by a factor of 100, the weak link is probably percentage conversion. If the reverse target is too low, the student may be adding losses instead of multiplying retained fractions.
If all the calculations are correct but the recommended batch count misses the budget, the error is constraint interpretation. Different mistakes need different repairs, rather than more random practice questions.
Keep a three-column correction record: erroneous move, reason it fails, new check. For example, ’rounded 6.9 to six’ becomes ‘minimum requires rounding up; verify output at six and seven’.
After each correction, attempt a similar question with changed numbers. The goal is transfer: knowing why the method works, not remembering the old answer.
34. Mastery check: solve then explain
The strongest exam response sets up named quantities, gives the necessary workings, handles percentages and units consistently, applies whole-number constraints, and ends with a sentence answering the actual question.
Long calculations are not a replacement for interpretation. The examination expects the student to show essential working; an isolated final integer does not communicate which constraints were checked.
Try a final scenario: 250 starting tokens per whole batch, target fraction 60%, retention of target 80%, minimum final target 700 units, maximum budget $1,100 at $180 per batch. Each yields 120; six batches yield 720 and cost $1,080, so six satisfy both requirements.
Explain why five batches fail output, seven batches exceed budget, and why the result applies only to the educational model. That complete explanation is a more valuable demonstration than saying simply ‘six’.
Final reverse-target problem and further reading
**Worked final answer.** Let x be the number of whole starting batches. Each delivers 250×0.60×0.80 = 120 final target tokens in the invented model. The output requirement gives 120x ≥ 700, so x ≥ 5.833…, and whole batches require x ≥ 6. The budget condition gives 180x ≤ 1100, so x ≤ 6.111…, and whole batches require x ≤ 6. Therefore x = 6 is the only feasible number. It produces 720 target tokens and costs $1,080. No medical or quality conclusion is implied.
**Keep the subject owners distinct.** For the gene-to-protein science, use Biology Volume 0908. For a different mathematical problem involving concentration units and weighted mixtures, use Mathematics Volume 0905. The earlier PSLE Mathematics word-problem transition guide helps learners who can calculate but struggle to translate English conditions into relationships.
**Official source.** Consult the 2027 SEAB G3 syllabus index (last updated 26 October 2026) and the 2027 G3 Mathematics syllabus K310. Mathematical steps in this invented context illustrate official skills; they do not describe an actual SEC paper question.
