The immediate answer: follow the population, not just the survivor
When a G3 Biology question shows more resistant bacteria in a later population, the reliable explanation is a sequence: variation already exists or arises through mutation; an environmental selection pressure changes which individuals survive and reproduce; the survivors pass on inherited characteristics; over generations, the proportion with a helpful characteristic can increase. A change in the fraction is a population result. It does not mean that each bacterium deliberately became resistant because it needed to.
Begin with what the source actually reports. Is it the number of resistant organisms, the proportion resistant, a drawing, or a claim by a narrator? Each needs a different answer. A classroom diagram with fifty resistant colonies cannot establish that a medicine works in people. A percentage growing from ten to twenty does not automatically mean the resistant population doubled. State the measured quantity, explain the mechanism appropriate to the question, and finish with the limit of the evidence.
This volume continues the series after Volume 0882 on inhibition-zone evidence and Volume 0885 on the exhibit audit. Those owners examine a clear region around a disc and the danger of medical overclaims. Our new problem is different: how to reason from repeated observations of a changing population. All data below are invented for examination practice. This is not a guide for culturing bacteria or selecting treatment.
Five things to decide before writing
First identify the population and the time point. A count from one fictional sample is not the same as a count from all organisms in an environment. Second identify the trait: resistance is not a vague claim of superior health; it is a characteristic defined against the particular conditions described. Third identify the selection pressure, if the question supplies one. Fourth ask whether the data support survival, reproduction and inherited transmission separately. The mechanism can explain a pattern, but the data may not prove every link. Fifth identify exactly what the examination command word requests: describe a trend, explain a process, compare quantities, or evaluate a conclusion.
An efficient answer is often shorter than a confident-sounding paragraph. For ‘describe’, use numbers and direction without importing mechanism. For ‘explain’, connect cause and sequence. For ‘suggest’, mark uncertainty rather than presenting a possibility as fact. For ‘evaluate’, balance what the results support against a material limitation. These distinctions help you collect marks without filling space with scientific vocabulary that is only loosely connected to the task.
Where this belongs in the 2027 SEC
The Singapore–Cambridge Secondary Education Certificate starts in 2027 under Full Subject-Based Banding. G3 is a subject level, not a separate certificate or a new fourth stream. SEAB’s 2027 school-candidate G3 index, last updated on 26 October 2026, lists Biology as K325, English Language as K300 and Mathematics as K310. Combined Science involving Biology is listed under K327 or K328 according to the subject combination. Students should follow the syllabus for the subject actually registered through their school.
The 2027 G3 Biology K325 assessment includes a 40-mark multiple-choice Paper 1 lasting one hour, an 80-mark structured and free-response Paper 2 lasting one hour forty-five minutes, and a 40-mark practical Paper 3 lasting one hour fifty minutes. This is a learning exercise, not a reproduction of an official SEAB question and not a claim that this exact dataset will be examined. Biology assessment values applying concepts, interpreting unfamiliar information and drawing supported conclusions. You can verify the current codes and links through SEAB’s official 2027 G3 syllabus index and the SEC overview.
The core distinction: resistance in an organism versus resistance in a population
A bacterium may have a heritable characteristic that permits survival under a specified selective condition. In a population, some individuals may possess the characteristic and others may not. The population’s composition is expressed in numbers or proportions. If the environment changes, those with a survival advantage under that particular condition may leave more descendants. The distribution of the trait can change. The original meaning of ‘resistance’ must stay attached to the source’s stated condition, not migrate into a claim that the organisms resist every substance.
An examination trap arises when a pupil says ‘the bacteria adapt by becoming stronger when the antibiotic arrives’. This makes the change purposeful and individual. A clearer explanation is that variation exists among organisms and selection changes the relative success of those variants. Mutations occur without planning for the current need; a mutation may be advantageous, neutral or disadvantageous in a given environment. Do not claim that all resistance results from a brand-new mutation during the observation window when the prompt has not established that.
Another trap is to write ‘the weak bacteria die, so the survivors reproduce’ as though the surviving group must increase in absolute count immediately. A population can decline sharply in total even while the percentage resistant increases. The resistant fraction may rise because the non-resistant group becomes much smaller. Number and proportion must be treated separately; the next volume owns the detailed Mathematics.
A completely fictional population record
Imagine that a researcher describes two sealed, hypothetical records. At an earlier point the entire illustrative population contains 900 organisms without the defined resistant characteristic and 100 organisms with it. Later, the record shows 180 without the characteristic and 120 with it. No real laboratory procedure is described or required; the counts are designed only to train reading and explanation.
At the first point, the total is 1,000 and the resistant proportion is 100 out of 1,000, or 10%. Later, the total is 300 and the resistant proportion is 120 out of 300, or 40%. A student can state that the percentage rose by 30 percentage points. Yet this was not a fourfold increase in the number resistant: that count rose from 100 to 120, an increase of 20%. The total population fell from 1,000 to 300. These four statements are compatible. They answer different questions.
The Biology explanation must be conditional. A selective environmental change could favour organisms with the resistant characteristic, helping explain why their proportion became higher. But two counts alone do not prove how inheritance operated, whether immigration occurred, or which precise mechanism caused the change. If the question supplies experimental details excluding those alternatives, a stronger causal explanation may be possible. Without such details, say what the data indicate, then explain a plausible selection mechanism while respecting the evidence boundary.
How natural selection differs from wishful adaptation
Natural selection is not a statement of intention. An organism does not forecast a future problem and request a helpful mutation. Variation can arise, and features can be inherited. Where environmental conditions affect survival and reproduction, individuals may contribute different numbers of descendants to later generations. Over many generations, a favourable inherited variant can become more frequent. The words ‘variation’, ‘selection pressure’, ‘survival’, ‘reproduction’ and ‘inheritance’ describe distinct links rather than synonyms.
Be careful about the word ‘generation’. A later measurement is not automatically a later generation. A population sampled at two times could include many of the same individuals, especially when no reproduction record is supplied. If an examiner asks for an explanation over generations, mention reproduction and inheritance; if the source merely reports two snapshots, present the evolutionary explanation as an interpretation compatible with the observations. This distinction makes your answer both scientifically stronger and less likely to overclaim.
Selection also depends on context. A characteristic that is advantageous when a particular pressure is present need not be advantageous in every environment. The question may ask what happens after conditions change again. Do not mechanically predict that resistance must keep increasing forever. Look for evidence about growth, competition, inheritance and environmental conditions before describing a likely trajectory.
Separating mutation, selection and inheritance
Mutation is a change in genetic material. Natural selection is differential survival and reproductive success associated with inherited variation under given conditions. Inheritance means that genetic information can pass to descendants. These are related processes, but one should not be substituted for another in a written explanation. A mutation does not necessarily become widespread. If no descendants inherit it, its population frequency may not rise. Equally, existing variation may be selected without a new mutation occurring at the moment the pressure begins.
Try the following correction. A rushed response says, ‘The antibiotic teaches every bacterium to make resistance genes.’ That incorrectly assigns purposeful learning to bacteria and universal change to every individual. Replace it with, ‘The population may contain individuals with a heritable resistance characteristic. Under the stated selective condition, those individuals may be more likely to survive and reproduce, increasing the proportion of resistant descendants.’ This is precise without pretending to know how every organism acquired the trait.
At G3, a response must match the amount of information in the stem. Use a general process explanation if the question does not supply a particular molecular mechanism. Do not add an invented gene name, an unsupported chemical pathway or a claim about treatment success. Scientific vocabulary earns its place by explaining what happens in the scenario.
Four levels of evidence and what each can support
The first level is observation: a table shows recorded counts. The second is description: a mathematically accurate account of their difference. The third is mechanism: a biological explanation that could produce the observed pattern under stated assumptions. The fourth is an applied decision: whether the evidence should guide real-world action. Each step requires more support. A school assessment may legitimately ask learners to move from the first to the third; it does not turn their fictional example into clinical advice.
A simple evidence ladder prevents a common error: ‘The resistant percentage is 40%, therefore the condition is untreatable.’ The percentage does not establish treatment outcomes or whether every individual in the population behaves identically. Nor does it identify the patient, the illness or the tested option. In a Biology answer, stick to the defined organism, trait, selection condition and population. In everyday life, clinical choices belong with qualified healthcare professionals who have actual case information.
Another version of the mistake is to confuse a useful model with an observed reality. A teacher may present coloured dots to illustrate selection. Those dots make the reasoning visible; they are not themselves measured microorganisms. State ‘in the model’ whenever a conclusion depends on the model’s assumptions. The habit serves you well in Mathematics and English too, where a diagram or persuasive caption may hide a missing assumption.
Understanding a trend without hiding the denominator
Many students see a percentage change and immediately describe improvement or deterioration. A proportion describes a relationship between part and whole. Its denominator may change. In our fictional record, the percentage resistant increases while the total number of organisms falls. The description should mention both if the question concerns the population as a whole. One number cannot substitute for the other.
Before forming a claim, draw a simple two-row table labelled ‘earlier’ and ‘later’. Add columns for number resistant, number not resistant and total. If the stem supplies percentages only, ask whether the total population or sample sizes are also given. A percentage difference between two samples of different size cannot reveal the difference in actual counts unless the sample totals are known. This is an elementary fact with important consequences for scientific interpretation.
A concise evidence sentence might be: ‘The resistant proportion was greater in the later record, rising from 10% to 40%, while the overall population fell from 1,000 to 300.’ That sentence describes the data without suggesting that resistance was deliberately developed, that the later conditions were safer or that a patient’s outcome can be inferred. The next mathematical guide explores how to compute such sentences correctly.
Reading a graph that is designed to persuade
Suppose a bar graph starts its vertical axis at 30%, placing 35% and 40% at dramatically different heights. A reader may feel a huge jump even though the numerical difference is five percentage points. The graph is not necessarily false, but its scale can exaggerate a visual impression. Read the numbers before interpreting the apparent height. Note the axis minimum and any break; compare intervals as numbers and units.
If two bar charts compare different samples, examine whether the label ‘resistant isolates’ refers to tested organisms, all organisms, or samples from a selection process. A subset chosen because it already has a characteristic is not a random picture of a wider population. Similarly, a table showing absolute resistant counts may not be enough to describe frequency without total population values. Think about what was counted and how it was chosen.
Exam answers should not invent a standardised sampling procedure that the question does not state. Say, for example, ‘the displayed sample has a higher resistant fraction’, not ‘the entire country has more resistance’, unless the data genuinely represent the whole population. Strong examination reading keeps the scope of the conclusion as small as the supporting evidence requires.
Three kinds of change that sound similar
An absolute change in resistant count is the later number minus the earlier number. A percentage change in that count divides the change by the earlier count and multiplies by 100%. A percentage-point change compares two percentages by subtraction. In the illustrative record, count change is plus 20, relative count change is plus 20%, and fraction change is plus 30 percentage points. Saying only ‘it increased by 30%’ would be ambiguous and potentially wrong.
Biology students do not need to turn every explanation into a lengthy mathematical derivation. They do need to use the language that matches the data. ‘A greater proportion’ is different from ‘more individuals’, and ‘became prevalent’ is different from ‘was detected’. If the question asks for an explanation, bring the biology after the numerical description. Use a short link: ‘This pattern is consistent with selection favouring individuals with the inherited characteristic under the given conditions.’
Keep measurement uncertainty visible. If the graph gives estimates, do not write decimal precision unsupported by the diagram. If values are rounded, avoid claiming the precise difference is exact. An answer that makes sensible qualification sounds confident because it states precisely what is known, not because it contains sweeping statements.
What would count as stronger evidence?
A well-designed explanation asks how an alternative account might be distinguished. Two snapshots may reflect selection, but they could also reflect the arrival of new individuals or different sampling locations. To decide between explanations, you would need additional evidence appropriate to the scenario: information about when measurements were made, how populations were identified, whether reproduction occurred and whether the same definitions were used at each point. You need not propose an unsupervised laboratory experiment; an evidence plan can be written on paper.
For a school question, a suitable response could ask for repeated records under comparable conditions and a clearly specified denominator. Where the stem explicitly includes controls, explain the role of those controls. Where it does not, identify the missing information rather than inventing it. Repeated measurement can reduce the risk that an unusual observation dominates the conclusion, but repeating the same flawed sampling method does not guarantee validity.
Explain why an improvement matters. More independent observations help separate a consistent pattern from an isolated anomaly. Known sample totals permit valid proportion comparisons. Identical definitions of resistance across records reduce ambiguity. Documentation of population movement and reproduction strengthens a causal account. These are not decorative checklist items; each repairs a particular logical weakness.
A four-sentence structured-response method
Sentence one describes the initial variation: ‘The starting population contains individuals with and without the stated resistant characteristic.’ Sentence two identifies pressure: ‘The described environment makes survival less likely for some individuals than for others.’ Sentence three explains differential contribution: ‘Individuals with the favourable inherited characteristic are more likely to survive and leave descendants under those conditions.’ Sentence four connects to the population result: ‘Over generations, the proportion of descendants with the characteristic may therefore increase.’
Adapt that pattern to the question. If no reproductive record is given, keep the genetic conclusion conditional rather than asserting that the four steps have all been observed. If the prompt supplies a specific gene variant, you may name it accurately. If the examiner asks only to describe the graph, stop after the first numerical statement; do not waste time rehearsing all four sentences. The method exists to organise reasoning, not to make every answer four sentences long.
Check your final verb. ‘May explain’ signals a possible mechanism; ‘demonstrates’ claims a stronger standard of evidence. A model showing selection does not prove an entire real-world population changed for that reason. Choose language as carefully as you choose terminology.
Avoid treating every change as natural selection
Population numbers can change for reasons unrelated to a change in heritable trait frequency. Changes in food supply, space, temperature, disturbance or the addition or removal of individuals can alter counts. A shift in sampling method can create an apparent trend without any population change at all. If the question asks you to infer selection, first see whether it tells you about inherited variation and an environmental advantage.
Consider two populations with identical resistance proportions but different sizes. The larger population contains more resistant individuals, yet the resistant fraction is the same. Selection cannot be inferred simply because one record has more resistant organisms. Conversely, a small subgroup may have a very high percentage without representing the wider population. These are contexts where looking only at one bar or headline creates error.
Scientific understanding does not require cynicism toward data. It requires constructive discipline: state the result the data support, identify the biological process if the question supplies enough information, and say what additional evidence would distinguish competing explanations. In examinations, clear limits usually strengthen the answer rather than weaken it.
Why an inherited advantage is not a moral label
The terms ‘fittest’, ‘strongest’ and ‘weakest’ can mislead when borrowed from everyday conversation. A trait is advantageous only relative to the environment and the biological outcome under discussion. The individuals favoured by selection may not be larger, faster or healthier in every respect. A bacterium with a resistance characteristic is not ‘better’ in a general sense; the characteristic may affect survival under a specified pressure.
The word ‘successful’ is also context-dependent. In a population-change question, it usually refers to leaving descendants, not to achieving a human goal or making an intentional choice. A learner who says ‘the bacteria wanted to win’ misses the mechanism. Use neutral language: ‘individuals with this characteristic were more likely to reproduce under these conditions.’
This habit links Biology to English. Persuasive phrases such as ‘survival of the toughest’ may seem vivid but substitute a moral or dramatic story for a specific explanation. Examination science prefers a chain of biological causes, with each claim grounded in the stem.
When the time interval is too short for a generational conclusion
A question may show measurements taken only minutes apart. Without information about reproduction, it would be inappropriate to describe an evolutionary increase over generations as a proven event. The change could reflect differential survival within the original population. The mechanism is not the same as a later increase caused by reproduction and inheritance. Carefully distinguish selection among existing individuals from an intergenerational change in trait frequency.
Imagine a later diagram in which some organisms are visibly absent. It may support a statement about the group remaining in that illustration. It does not automatically prove that the remaining organisms reproduced, passed on genetic information or became more frequent in a future generation. If the question says descendants were counted after several rounds of reproduction, you can discuss inheritance with a firmer basis.
A strong candidate reads the timeline as a constraint. Time is part of the evidence, not scenery. If the test gives no timescale at all, avoid making one up.
Interpreting a control group without making it magical
A controlled comparison can strengthen an explanation when groups differ in a specified factor while other relevant conditions are reasonably held constant. In a fictional example, one population record may be labelled ‘pressure present’ and another ‘pressure absent’. The comparison can help assess whether a change is associated with the pressure. It does not automatically eliminate every possible source of bias.
Ask what ‘similar starting populations’ means. Do they have the same total size and starting fraction? Is the measured characteristic defined identically? Are the observations made after comparable intervals? If a control group begins with a different fraction, that matters. If sample sizes differ, that matters too. A valid conclusion compares like with like as far as the data allow.
In Paper 3-style investigation questions, a student may be expected to identify variables, controls and limitations. Here we practise those decisions as text-based reasoning rather than giving any instructions for performing work with live microorganisms. Laboratory tasks belong under school supervision and official safety procedures.
Turning a description into a causal explanation
Suppose an examiner asks ‘Explain why the resistant fraction increased in the hypothetical population.’ A low-quality answer repeats the graph: ‘Because there were more resistant bacteria later.’ That is not a cause; it restates the observation. Another weak answer says, ‘Because the bacteria adapted.’ It skips the mechanism. A stronger response identifies existing heritable variation, differential survival under the given pressure, and transmission of the characteristic to descendants where reproduction is stated.
Now suppose the command is ‘Describe how the proportion changed.’ The full natural-selection account would overshoot the task. Write the two proportions, direction and, where appropriate, percentage-point difference. A careful candidate does not use an explanation when a description is requested or an inference when a calculation is requested. The command word determines how deep the answer should go.
One practical technique is to circle the command word and underline the variable. Then draft a single line in the margin: ‘I must explain why the fraction changed, not merely calculate the count.’ That short plan often saves more time than rushing into a paragraph.
Looking for a counterexample
Universal claims are fragile. ‘Resistance always increases when an antibiotic is present’ can be challenged if the population lacks the trait, the pressure is insufficient to favour it, the interval is brief, or the recorded denominator shifts. An examiner may not require an exhaustive catalogue of exceptions. One relevant counterexample connected to the source is enough to show that the original claim overstates the evidence.
Try a different claim: ‘Any rise in the resistant fraction means that the resistant count increased.’ Our illustrative record offers one counterexample pattern in principle, but in that dataset the count also rose slightly. Construct a new hypothetical record: 20 resistant of 200 initially (10%), and 12 resistant of 40 later (30%). The resistant proportion triples while the resistant count falls from 20 to 12. This directly disproves the universal claim.
Good counterexamples are not distractions. They test the logical shape of the claim. If the statement says ‘always’, one coherent case where it fails may be decisive.
Exam practise A: describe only
Prompt: ‘In a fictional sampling sheet, resistant counts are 24 of 240 at Time A and 21 of 70 at Time B. Describe the change in resistant proportion.’ Work: Time A is 24/240 = 10%; Time B is 21/70 = 30%. The fraction rises by 20 percentage points. The absolute number resistant decreases by three, so an answer claiming the resistant count increased would be wrong.
Model response: ‘The fraction recorded as resistant increased from 10% to 30%, a rise of 20 percentage points, although the resistant count fell from 24 to 21.’ The first clause answers the command. The second reveals an important contrast without adding an unsupported biological mechanism. On a short-mark question, the examiner may not need the second clause; use judgment about what is requested.
Exam practise B: explain a selection scenario
Prompt: ‘A hypothetical population contains an inherited variation. Under the stated environmental condition, individuals with variant R survive more often and leave descendants. Explain the later increase in R.’ A reliable response begins with variation before the pressure, not with the claim that pressure purposely created R. Then it states that the condition gives a survival or reproductive advantage to individuals with R, followed by inheritance in descendants.
Model response: ‘The starting population contains heritable variation. Under the specified pressure, organisms with R are more likely to survive and reproduce than those without it. Their descendants can inherit R. Over successive generations, R therefore forms a greater proportion of the population.’ Note that this conclusion is supported by the fictional prompt’s explicit statements about inheritance and reproduction. If those were absent, the certainty should be reduced.
Exam practise C: repair a misleading inference
Prompt: ‘The percentage resistant increased from 15% to 45%, so the total number resistant must have tripled.’ Explain the flaw. The percentage of a total depends on the total’s size. If that total changes, the number resistant can increase by less than three times, remain constant or decline. We need the group sizes before comparing absolute counts.
Model response: ‘The conclusion is not justified. The percentages refer to fractions of potentially different population totals. Without the earlier and later totals, we cannot calculate or compare the absolute numbers of resistant organisms.’ This answer uses no laboratory assumptions. It is a concise mathematical restriction applied in a biological context.
Exam practise D: distinguish time points from generations
Prompt: ‘A chart gives observations at 9:00 and 9:10. A student writes that several generations inherited resistance during those ten minutes.’ The chart only names observation times. It has not stated how many generations occurred or whether reproduction took place. The student has imported a story that the source does not provide.
Model response: ‘The chart does not establish that multiple generations occurred. It reports two observations over ten minutes, and additional information on reproduction and inheritance would be needed before concluding that the later fraction changed across generations.’ If the source explicitly defines generation intervals, use those facts; never assume them from general knowledge.
Exam practise E: separate model from clinical reality
Prompt: ‘A school poster uses an invented dataset and says a medicine cures 90% of infections.’ The dataset records only a hypothetical trait fraction, not patient recovery. The word ‘cures’ refers to an outcome not observed, and ‘infections’ covers many causes. The claim should not be repaired by quietly replacing 90% with 80%; the variable itself is wrong.
Model response: ‘The poster’s claim is unsupported. The invented population data describe the prevalence of a characteristic in the model, not clinical treatment outcomes. It should state the classroom observation and identify it as a model without making a medical promise.’ The role of this answer is critical reading, not healthcare guidance.
Exam practise F: recognise sampling selection
Prompt: ‘A display reports that 90% of the specimens chosen from a container labelled “resistant examples” had the trait, and concludes that 90% of all organisms in the original environment had it.’ The selection procedure is already biased toward the trait. The selected group is not representative of the original whole. The impressive percentage does not repair the sampling problem.
Model response: ‘The conclusion cannot be extended to the whole population because the sample was deliberately selected from a group already labelled resistant. A representative sampling design and the corresponding original population denominator would be needed.’ The wording is careful about what is missing without pretending that a classroom table can prove a real-world prevalence.
Exam practise G: decide which variable to control
Prompt: ‘Two fictional groups are compared to examine a specified environmental pressure. They differ in starting trait frequency and observation duration as well as the pressure.’ A response should identify at least one confounding difference and explain why it matters. If the groups start at different trait frequencies, a later difference may reflect the starting composition rather than the pressure. If durations differ, there may be different opportunities for change.
Model response: ‘Start the groups with comparable initial proportions and observe them over comparable intervals so differences at the endpoint can be more plausibly associated with the specified factor.’ This is a paper-based design answer; do not turn it into instructions for growing organisms.
Exam practise H: evaluate a strong-sounding conclusion
Prompt: ‘A report states: “Our two charts prove that environmental pressure caused all resistance in the population.”’ The word ‘prove’ overstates the support; two charts might show correlation, while the mechanism and history remain unspecified. ‘All’ also claims a universal origin, ignoring existing variation and other possibilities.
Model response: ‘The charts may be consistent with selection, but they do not alone establish the origin of every resistant individual or exclude other changes in the sampled population. The claim should be restricted to the observed proportions and the mechanism presented as a supported interpretation only if the study conditions warrant it.’
Exam practise I: translate the graph axis
Prompt: ‘The chart is labelled “resistant organisms per 100 examined”, but a pupil writes “number of resistant organisms in the whole habitat”.’ The axis defines a rate within a sample. It does not give the total habitat population. If two samples of different sizes are converted to the same per-100 measure, the axis permits comparison of sample fractions, not total organism counts.
Model response: ‘The graph presents resistant cases per 100 examined and therefore describes the observed sample proportion. It does not provide the number of resistant organisms in the whole habitat.’ Check the axis before writing any result sentence.
Exam practise J: write an answer with deliberate limits
Prompt: ‘The later fraction was higher. Suggest one explanation and one limitation.’ A compact answer can offer a possible selection process and one missing fact. The explanation should match the source’s given pressure. The limitation should matter to the conclusion, such as the absence of evidence about reproduction or comparability of samples.
Model response: ‘Selection may have favoured individuals already possessing a heritable resistance characteristic under the stated conditions, increasing their representation in the later sample. However, the data do not show whether the two samples were drawn from the same underlying population, so the change cannot be attributed confidently to selection alone.’
A practical revision method for this guide
Read one data table and identify three separate quantities: resistant count, whole-group count and resistant percentage. Next, write a purely descriptive sentence. Then write a mechanism sentence only after marking which facts the question supplies. Finally, cross out any unsupported universal word, such as always, all, guaranteed or cured. This sequence builds fluency without encouraging memorised paragraphs detached from evidence.
Across a week, vary the task. On one day practise translating between counts and percentages. On another repair a deliberately misleading poster. On a third explain the same mechanism with different initial conditions. On the fourth identify missing data. The variation matters: in an unfamiliar SEC question, you must recognise the reasoning structure rather than recall one diagram.
Ask a peer or teacher to mark which words in your response are observations and which are explanations. If they cannot tell, improve the connection. A strong learner can say not only what their answer is, but exactly where in the prompt its factual support comes from.
Common weak answers and a better replacement
‘The bacteria learn resistance’ becomes ‘selection changes the frequency of a heritable trait among descendants under the specified pressure’. ‘More resistant bacteria means the percentage increased’ becomes ‘compare the resistant count with the total count at each time’. ‘The graph proves the medicine will fail’ becomes ‘the graph describes an illustrative population characteristic, not a clinical outcome’.
‘The trait improved because the environment wanted it to’ becomes ‘individuals with a favourable inherited characteristic were more likely to leave descendants under the stated conditions’. ‘The 40% resistant group is bigger than the 10% group’ becomes ‘the resistant proportion is larger, but absolute numbers require group totals’. Each repair replaces a dramatic phrase with a measurable relationship or a clear causal step.
Do not memorise these as magic sentences. Learn what is wrong in the weak version. Once you can name the error, you can repair a new example rather than only one you have seen before.
A harder synthesis: a shrinking population with a rising resistant share
Consider three hypothetical snapshots. At A, 50 of 500 carry the defined trait. At B, 30 of 150 do. At C, 40 of 100 do. The fractions are 10%, 20% and 40%. The resistant count drops from 50 to 30 and then rises to 40, while the total falls throughout. If a poster says ‘resistant organisms quadrupled from A to C’, it confuses fraction and count. The resistant count at C is actually lower than at A.
A well-written response might say: ‘The recorded proportion carrying the trait rises from 10% to 40% across A–C, even though the overall sample size falls from 500 to 100 and the absolute resistant count falls from 50 to 40. These results do not by themselves identify the reason for the change.’ That is not evasive; it is complete evidence handling. The biological question of why remains dependent on conditions supplied elsewhere.
This is the point where the series moves into Mathematics. The next volume shows how to calculate and reverse such records, use weighted fractions when samples are combined, and test whether a claimed change is possible. The biological explanation here remains separate from the mathematical owner.
How to transfer the skill beyond bacteria
Population selection questions are not confined to microbes. A similar reasoning pattern may concern inherited traits in another organism, a change in proportions between generations or a sample that does not represent its wider environment. The transferable skill is to distinguish a changing individual from a changing population and to decide what evidence would connect a trait to differential survival and reproduction.
Some problems are not about natural selection at all. A graph of the number of people choosing a club, for instance, describes choices and may have no inherited mechanism. Do not import evolutionary vocabulary just because a fraction changes. The context decides which explanation is legitimate. A science term used in an unrelated setting is not evidence of deeper understanding.
Your answer should therefore begin with the actual subject: identify the variable, its meaning and the evidence before choosing an explanatory model. This habit is useful in G3 Biology, Mathematics statistics and English comprehension.
Where to go next and what to keep
For related science reasoning, revisit Volume 0882, interpreting inhibition evidence. For mathematical control of measurement and geometry in that earlier case, use Volume 0883. To see how English prevents a scientific overclaim, use Volume 0884. These links are narrow companions, not alternative versions of the present topic.
Students arriving from primary school can begin with the PSLE Learning Guide, the Foundation Science explanation route, or subject-specific study methods. For the official distinctions between SEC, subject levels and syllabuses, consult eduKate Sengkang’s 2027 G3 EMS overview alongside SEAB. The public information helps you choose the right subject syllabus; the exercises here help you think inside it.
The principle to carry into the examination is simple. Describe numbers as numbers. Explain selection as an interaction between inherited variation and an environment. State only what the source allows. Clarity protects both accuracy and marks.
Official examination references and scope
The syllabus and format details above refer to SEAB’s 2027 G3 school-candidate listing, updated 26 October 2026: official G3 subjects and codes. See also SEAB’s SEC overview. Always check the version corresponding to your examination year; future-year schedules or syllabuses may differ. None of the fictional numerical scenarios is an official question or a report of a real test.
This lesson is educational, not clinical guidance. Antibiotics are relevant to bacterial infections but not viral infections; treatment decisions should be made with qualified healthcare professionals. Do not use classroom diagrams or fictional data to choose, stop or alter treatment. The purpose here is to become a more accurate scientific reader and writer.
