The tallest plant at the end is not necessarily the one that grew the most. When two groups start at different heights, a comparison of final averages can produce an answer to the wrong question. G3 Mathematics helps us repair this mistake with paired changes, transparent averages and careful interpretation.
This learner’s guide continues a fictional fourteen-day seedling study. It takes the recorded numbers apart, calculates the right comparisons, checks variation, explains graph and denominator traps, and ends with original examination-style practice. Every value is for teaching, not a claim about a real biological experiment.
- Identify each seedling’s starting and final measurement.
- Calculate individual changes and mean increases.
- Compare mean, median, quartiles, range and standard deviation.
- Distinguish absolute, percentage and per-day growth.
- Read graphs and explain numerical findings without claiming causation.
Why the taller group may not be the faster-growing group
Two groups of seedlings are measured at the start and after fourteen days. At the end, Group B is taller on average. A student announces that its light treatment made it grow much faster. Another student checks the starting measurements first. Group B was already taller when the observation began. The final comparison is therefore answering a different question from the one the first student intended.
This guide teaches a simple mathematical discipline: if the question concerns change, compare change. When measurements refer to the same identified individual at two times, begin by pairing the start and end readings. Subtract to find each individual’s gain. Then decide whether a mean, median, range, standard deviation, proportion or graph is best suited to the question. Averages are useful only after we choose the correct underlying quantity.
We continue the fictional Biology investigation in Volume 1013. It distinguished daily light duration from light intensity and measured plant height from photosynthesis rate. Here, we treat the observations as a Mathematics problem. The computation has its own discipline, and the interpretation must remain faithful to the evidence. Nothing about this exercise proves the biological effect of a lighting treatment; it shows how to perform fair numerical comparisons when two groups did not begin at the same baseline.
Know the actual G3 Mathematics paper
According to the official SEAB 2027 G3 SEC school-candidate syllabus directory, updated on 26 October 2026, Mathematics is K310, replacing the earlier O-Level reference 4052. The K310 syllabus describes two written papers, each worth 90 marks and 50%, with two hours fifteen minutes allowed for each. Paper 1 contains about 26 short-answer questions. Paper 2 contains nine or ten questions of varying length, with a final question focused on applying mathematics to a real-world scenario. Students answer all questions in both papers.
SEAB expects essential working to be shown. Answers that leave out the important steps can lose marks even when the number happens to be right. Unless the question requires otherwise, non-exact numerical answers are generally given to three significant figures and angles in degrees to one decimal place. Appropriate formulas are supplied, but students still need to choose and apply them correctly. The official document, not a practice worksheet, is the source for examination formats and requirements.
The study of statistics includes reading and representing data, measures of central tendency and spread, and reasonable interpretation. This seedling context is our own extended example of those skills. It is not an SEAB question and is not a prediction about plant data appearing in a future Mathematics paper. Its value is that ordinary measurements force us to define the comparison before calculating anything.
The original data are paired observations
Each plant has a permanent identifier so that its two measurements can be matched. Group A contains five seedlings labelled A1 to A5. Their initial heights in centimetres are 4, 5, 6, 5 and 4. Their final heights after fourteen days are 8, 11, 12, 10 and 10. Group B contains five plants B1 to B5. Their initial heights are 6, 6, 7, 7 and 8; their final heights are 11, 13, 13, 14 and 13.
All the numbers are invented for this educational example. Group A received a specified six-hour daily lighting schedule and Group B an eight-hour schedule in the fictional study. Other relevant conditions were intended to be comparable but have not been established as perfectly controlled. The recorded height should be treated as a measured quantity, not as a direct measurement of photosynthesis rate.
This dataset is small enough to inspect by hand and rich enough to reveal a mistake. Group B’s initial values tend to be larger than Group A’s. Their final values also tend to be larger. To assess gain, we need the changes within individual plants rather than a simple ranking of final observations. The pairing matters because A1 at the end must be compared with A1 at the start, not with A3 or an unrelated plant.
The first calculation: final minus initial
For Group A, calculate each difference in centimetres: A1 is 8 − 4 = 4; A2 is 11 − 5 = 6; A3 is 12 − 6 = 6; A4 is 10 − 5 = 5; A5 is 10 − 4 = 6. The five increases are 4, 6, 6, 5 and 6 cm. For Group B: B1 is 11 − 6 = 5; B2 is 13 − 6 = 7; B3 is 13 − 7 = 6; B4 is 14 − 7 = 7; B5 is 13 − 8 = 5. The increases are 5, 7, 6, 7 and 5 cm.
A paired change is not the same as subtracting the height of one randomly selected final plant from another initial plant. In this context the identity of each seedling matters. The method also prevents an unhelpful leap: we do not need to claim that one group’s final height directly measures the effects of a particular biological process. We are calculating a clearly defined quantity from two readings.
When writing an exam response, name the quantity: “Increase in measured height = final height − initial height.” Put centimetres beside the result. A difference of 6 cm is not a rate of 6 cm per day; the six centimetres occurred over the full fourteen-day observation interval. If a rate is required, the period must be included in the calculation.
Average starting heights show the baseline difference
The Group A initial total is 4 + 5 + 6 + 5 + 4 = 24 cm. Its mean initial height is 24 ÷ 5 = 4.8 cm. The Group B initial total is 6 + 6 + 7 + 7 + 8 = 34 cm, giving a mean of 34 ÷ 5 = 6.8 cm. Before the experiment had produced any measured change, Group B was already 2.0 cm taller on average.
The Group A final total is 8 + 11 + 12 + 10 + 10 = 51 cm, giving a mean final height of 10.2 cm. Group B totals 11 + 13 + 13 + 14 + 13 = 64 cm, for a mean final height of 12.8 cm. The final mean difference is therefore 2.6 cm. That is a correct descriptive number. But only 0.6 cm of the difference remains when we compare the groups’ average gains.
This is a powerful warning against an attractive but wrong conclusion. Saying “B finished 2.6 cm taller on average” is mathematically correct. Saying “B grew 2.6 cm more on average” is false for this dataset. The starting comparison explains why the two statements differ. If you are comparing improvement, a good first question is always: did the groups start at the same point?
Mean gains, calculated two ways
The five gains in Group A total 4 + 6 + 6 + 5 + 6 = 27 cm, giving a mean increase of 27 ÷ 5 = 5.4 cm. In Group B, the gains total 5 + 7 + 6 + 7 + 5 = 30 cm, giving a mean increase of 30 ÷ 5 = 6.0 cm. The difference between mean increases is 0.6 cm.
There is also a useful algebraic identity for fully paired observations in the same group. Mean(final − initial) equals mean(final) − mean(initial). For Group A, 10.2 − 4.8 = 5.4 cm. For Group B, 12.8 − 6.8 = 6.0 cm. This works because each group contains the same five identified plants at both measurement times. It becomes unsafe if plants are missing or the two means refer to different sets of individuals.
Do not confuse equality of two correctly calculated means with a licence to ignore identity. The per-plant changes remain important when we want to examine variation or identify outliers. If the final table has different individuals or missing values, the shortcut may not represent actual paired changes. For assessment, show the definition of the mean and the original values relevant to the question, so the marker can see what your average measures.
The difference between descriptive fact and causal claim
The descriptive result is clear: Group B’s five plants had an average increase of 6.0 cm, while Group A averaged 5.4 cm. The question of whether the longer daily light exposure caused that 0.6 cm difference belongs to the experimental design. In a Mathematics answer, we should not invent a cause simply because the two groups had different treatments.
The distinction is important in real-world modelling questions. A numerical conclusion can be supported without a scientific causal conclusion being established. For instance, we can compare the two sample means, calculate a difference and examine its size. We cannot use these operations alone to prove what would happen to every plant, or what would occur under a different lamp, water supply or growing season.
A well-written interpretation is: “In the supplied five-plant records, Group B has a mean measured height increase 0.6 cm greater than Group A.” If the question requests a limitation, add that the samples are small, initial sizes differ, and uncontrolled factors could affect a biological conclusion. There is no need to water down the actual arithmetic. State what it proves within the dataset, then state what it does not prove outside that dataset.
Median increase and why it can disagree with the mean
Sort Group A’s gains: 4, 5, 6, 6, 6. Its median gain is the middle value, 6 cm. Sort Group B’s gains: 5, 5, 6, 7, 7. Its median is also 6 cm. The means differ, but the medians are equal. That is not a contradiction. The mean incorporates every value in a way that responds to their magnitudes; the median gives the middle observation once the data are ordered.
If a student reports only “B did better”, they hide that different measures can answer different questions. B’s mean gain is larger. The middle gain is the same in both groups. A clear comparison names the measure and the exact figures: “The mean increase is greater in B by 0.6 cm, but the median increase is 6 cm for both groups.”
Neither measure automatically determines which treatment to recommend. Choosing a measure depends on the task and how the values are distributed. If a dataset contains a strongly unusual measurement, the mean may be influenced considerably while the median changes less. If a question asks for average total change across all observed plants, the mean can be useful. Examine the problem and select the summary that fits its meaning.
The range reveals something that averages omit
For Group A, the smallest increase is 4 cm and the largest is 6 cm. Its range is 6 − 4 = 2 cm. Group B has a minimum increase of 5 cm and a maximum of 7 cm, so its range is also 7 − 5 = 2 cm. Two groups can have different means and the same range. Here, that is exactly what happens.
The range is easy to calculate, but it uses only the two extremes. It does not show how most observations sit between them. Group A’s gain values are concentrated toward 6 cm; Group B has two 5 cm observations and two 7 cm observations, with one at 6 cm. The equal ranges do not mean the distributions are identical.
If a student sees the same range and claims “all plants grew equally consistently”, the response has gone too far. A more accurate statement is that the distance between the highest and lowest recorded gain is the same in both groups. We would need other measures to compare spread more completely. In a question that specifies range, give the range correctly and stop unless an interpretation is requested. In a longer data comparison, range can be one useful feature but should not be the only one.
Quartiles and the interquartile range
With five sorted gains, the median is the third value. For an illustrative hand calculation using the method that excludes the median from the lower and upper halves, Group A’s lower half is 4, 5 and upper half 6, 6. The lower quartile is therefore 4.5 cm, and the upper quartile 6 cm. Its interquartile range is 6 − 4.5 = 1.5 cm. Group B’s lower half is 5, 5 and upper half 7, 7, giving lower quartile 5, upper quartile 7 and interquartile range 2 cm.
Quartile conventions can differ for tiny datasets and calculator implementations, so follow the method required by the question or its supplied data representation. We are using a clearly stated convention for this exercise rather than claiming every method must return the same numerical quartiles for five observations. The conceptual point is stable: the interquartile range describes the spread of the central part of an ordered dataset.
Under this convention, Group B has a wider interquartile range, although the groups have equal ranges and medians. A thoughtful interpretation reads the measures together. It would be incorrect to infer a universal biological treatment effect from those summaries. The numerical question asks how these five observed values compare; the scientific question asks why differences arose. Keep the boundary visible.
Standard deviation: a more complete measure of spread
The standard deviation uses the deviations of all data values from their mean. For Group A’s gains 4, 6, 6, 5 and 6, the mean is 5.4. The squared deviations are 1.96, 0.36, 0.36, 0.16 and 0.36 square centimetres, totalling 3.20. For the five recorded values, the standard deviation using the usual population-data divisor of five is the square root of 3.20 ÷ 5, which is 0.8 cm.
Group B’s gains are 5, 7, 6, 7 and 5, with mean 6. Their squared deviations are 1, 1, 0, 1 and 1, totalling 4. The corresponding standard deviation is the square root of 4 ÷ 5, approximately 0.894 cm to three significant figures. Under that calculation, the B gains are slightly more spread around their mean than the A gains.
Do not treat the word “population” in this formula as proof that five seedlings represent the entire population of plants. We are computing the standard deviation of the five supplied observations, using the formula appropriate to the set of values. An inference about all seedlings would need further reasoning and an appropriate study design. For a focused explanation of what standard deviation means, see the distinct Volume 0351 statistics guide; this volume applies that tool specifically to paired growth records and baseline fairness.
Why one graph can make the difference look enormous
Imagine a bar chart showing the two mean gains, 5.4 and 6.0 cm. If the vertical scale begins at 5.0 cm, Group B’s bar looks many times taller than Group A’s. The actual difference remains 0.6 cm. When a graph uses a truncated axis, the visual impression of the bar heights may exaggerate the comparison, especially if the viewer assumes the bar length begins at zero.
A line graph can create a different mistake. Connecting a point at day 0 to a point at day 14 with a straight line suggests a linear visual trend between the two measurements, but we did not measure each intermediate day. The drawn line is an interpolation or simple visual connection, not direct evidence of steady daily growth. A pair of end measurements cannot reveal pauses, bursts or changes in rate between them.
Good graph reading begins with axis labels, units and scale. Is the plotted value final height, gain, daily average rate or percentage gain? How many data points support each representation? What is the baseline? The PSLE Science guide on comparing graphs with different scales introduces the same caution in a simpler setting. At G3, students should apply it while explaining numerical conclusions and model assumptions.
Absolute gain versus percentage growth
An increase of 6 cm means something different for a seedling that began at 3 cm than for one that began at 30 cm. Percentage growth from a defined baseline is calculated as increase divided by initial value, multiplied by 100. In this dataset, the mean gain divided by the mean initial height gives a group-level ratio: Group A, 5.4 ÷ 4.8 × 100 = 112.5%; Group B, 6.0 ÷ 6.8 × 100 ≈ 88.2%.
This reverses the apparent ranking of growth compared with the absolute average gains: B has the larger mean gain, but A has the larger gain relative to the mean initial height. Neither ratio is inherently more “correct” without a defined comparison question. The denominator changes the meaning. Also, the ratio of group mean gain to group mean starting height is not necessarily equal to the mean of individual percentage gains. Those operations weight the observations differently.
A careful response would say precisely what was calculated: “The group mean gain is about 112.5% of Group A’s mean initial height.” It would not say “Every A seedling grew by 112.5%”. To calculate each plant’s percentage growth, use that individual plant’s starting height as the denominator and then summarise those five percentages if requested. The order of operations and the identity of the baseline matter.
Compute the average change per day with care
All measurements in the case are fourteen days apart. Dividing each group mean gain by fourteen gives an average change over the period, expressed in centimetres per day. For Group A, 5.4 ÷ 14 is approximately 0.386 cm per day. For Group B, 6.0 ÷ 14 is approximately 0.429 cm per day. Their difference is about 0.0429 cm per day, based on the two group means.
These are average rates over a fourteen-day interval. They do not show the rate on any particular day. If the plants grew slowly at first and faster later, the end-to-end average would still be the same. Do not use a straight-line interpretation without intermediate measurements to support it. A Mathematics question may ask for the gradient between two known endpoints; that gradient is valid as an average rate of change, but not as a claim that the process was uniform every day.
The unit deserves attention. A gain of 5.4 cm has units of length. Divide by fourteen days and the unit becomes centimetres per day. If a student writes “5.4 cm/day” without dividing by time, they have changed the meaning of the number. Check each new unit when building a rate, a proportion or a graph label.
When pairing stops working because a record is missing
Suppose plant B3 is not measured on day 14 because the label was lost. Its starting height is recorded, but its final height cannot be matched confidently. The correct response is to acknowledge that its paired gain is unavailable. Entering zero centimetres invents an observation. Pairing the remaining initial reading with a different plant’s final height also invents an identity. Averages made with fabricated or mismatched values are not rescued by neat arithmetic.
One possible revised analysis uses only the four seedlings with complete pairs, reports that the sample size is now four, and explains why this differs from the original planned five. That decision can introduce bias if missingness is related to plant condition or treatment. It should not be described as proof that the remaining plants are representative. Other approaches may be required in more advanced statistical work, but our exam-level lesson is to preserve the meaning of the data rather than silently fill blanks.
A common shortcut, mean(final observed) minus mean(initial all), can be misleading when the sets of individuals differ. The arithmetic identity for the mean of paired changes holds cleanly when the same identified observations are included at both times. Check membership before applying a formula that once worked perfectly on complete data. This is one of the most useful real-world habits to carry from the case.
A model with unequal group sizes
Now imagine that Group A has five complete records and Group B has only four. You cannot combine their mean gains by simply taking the ordinary mean of the two group means if you want the overall mean across individual plants. The groups contribute different numbers of observations. The appropriate combined mean uses their totals, or a weighted mean with weights equal to sample counts.
For a transparent illustration, suppose A’s five gains still total 27 cm, and the four available B gains total 24 cm. The combined mean across all nine observed plants is (27 + 24) ÷ 9 = 51 ÷ 9, approximately 5.67 cm. The unweighted average of the two group means, (5.4 + 6.0) ÷ 2 = 5.7 cm, is a different quantity because it gives the two groups equal weight despite their different sizes.
Neither figure automatically expresses a treatment effect. A combined mean answers the question “What was the average across these recorded plants?” Comparing treatment groups is a separate task. When a table contains unequal totals, always ask what is being averaged and what each observation or group should count for. The number of entries is part of the meaning of the statistic.
Why a bigger sample is not the same as a fairer comparison
Suppose a class repeats the project with forty seedlings, but every longer-light plant is placed beside a warm window and every shorter-light plant in a cooler corner. The sample size is larger, yet the light schedule is now systematically entangled with location. The measured difference could reflect multiple factors. Mathematics can describe its size accurately, but the data design does not by itself isolate the proposed cause.
By contrast, a smaller but carefully assigned comparison may be more interpretable within its limits. Good data collection defines the units, records baselines and treats the two groups comparably apart from the intended difference. Random allocation can reduce systematic initial differences, though random assignment does not guarantee identical groups. Repeated measurements make the history of each plant visible and support calculation of changes.
This is an important lesson for real-world Paper 2 questions. An appealing dataset can still require a critical interpretation. The mathematical response should respect the supplied model, show valid working, and note relevant limitations when asked. It should not use statistical terms as a licence to invent biological conclusions. Sample size, design, representation and measurement quality all affect what the numbers can reasonably tell us.
A budget-and-measurement extension without changing the owner
Suppose a class has S$60 for simple non-electrical measuring materials and labels. Each labelled plant record costs a hypothetical S$2 for consumables, and there is a fixed S$12 record-keeping cost. If n plants are included, the illustrative cost model is C = 12 + 2n dollars. The budget restriction is 12 + 2n ≤ 60. Therefore n ≤ 24. The maximum whole number of plants allowed by this simplified budget is 24.
That result is a Mathematics interpretation of a linear inequality. It is not a Biology recommendation that 24 plants are scientifically sufficient, nor does it include growing equipment or safe lighting costs. The question defines a mathematical model with selected costs. A correct examination answer solves the inequality and explains the integer constraint rather than presenting 24 as a universal experimental design rule.
The extension shows why the same setting can produce distinct questions without repeating the same intent. Volume 1010 focused on electrical-energy kWh and tariff budgets. Our current focus is baseline-adjusted paired data, so the small budgeting exercise is merely a short application of mathematical constraints. The central lesson remains data identity, appropriate summaries and transparent comparisons.
One integrated Paper 2-style real-world problem
A school must publish one clear figure from the supplied plant observations. The proposed options are: mean final height, mean gain, mean gain relative to mean initial height, and median gain. The task is to choose a measure that directly describes average absolute growth during the fourteen-day period, calculate it for both groups, and explain why a different measure might produce a different impression.
The suitable primary measure is the mean paired gain, because growth is defined as measured height change from the same identified plants. Show final minus initial for each, total the changes, then divide by five. The results are A = 5.4 cm and B = 6.0 cm. If the school also reports starting and final means, readers can see that the groups were unequal at baseline. If a question requests typical middle growth, the median is 6 cm in both groups. If it requests relative change, the denominator must be defined clearly.
The last part asks whether the numerical evidence proves that extra light duration raised photosynthesis rate. The answer is no. Height change is not a direct rate measurement, and the small observational comparison does not by itself establish causation. The Mathematics student needs to interpret the numbers properly, not solve an entire Biology examination. A complete response shows method, values, units and an explanation proportionate to the evidence.
How to show essential working clearly
Good mathematical working is a short trail another person can follow. For a paired change, write the difference and label the plant or group. For a mean, show the total and the count. For a percentage, show the denominator. For a standard deviation, state what the data values are and use an appropriate formula or calculator process. For a rate, write the interval and resulting unit.
Avoid two common extremes. At one extreme, a student writes only a calculator answer such as “6.0”. The marker cannot know whether it is centimetres of gain, centimetres of final height, or centimetres per day. At the other extreme, the student fills the page with repeated commentary that does not affect the result. The strongest solution is concise, complete and connected to the task.
When reporting a non-exact value, observe the question’s accuracy instruction and the syllabus’s general convention. For example, 6.0 cm may be an exact mean of the listed integer-centimetre changes, whereas 0.428571… cm/day is non-exact and may be reported as 0.429 cm/day to three significant figures. Do not round intermediate results unnecessarily in a long chain, as this can distort the final answer. A unit check and a sentence of interpretation often reveal mistakes faster than redoing the arithmetic blindly.
Five original questions for independent practice
Question one: State the mean initial height of each group and the difference. Question two: Find each plant’s gain and calculate the mean gain for each group. Question three: Explain why comparing the final mean heights alone could mislead someone who wants to compare growth. Question four: Calculate the median and range of the gain records, then give one comparison that is true even though the means differ. Question five: The school displays the phrase “Group B’s light treatment caused 2.6 cm more growth.” Identify two separate problems with that sentence.
Attempt the questions without looking up the answers. Give each quantity a clear label and unit. If you are unsure which mean is intended, write the mathematical definition before substituting values. A calculator can assist with the operations, but it cannot decide what the question means. For a numerical comparison, identify whether the figures describe starting values, final values, paired gains or percentage changes.
After completing the set, test the last answer especially carefully. There is a numerical error because 2.6 cm is the final-height difference, not the difference in mean gains. There is also an evidence error because the numbers alone do not establish that the lighting schedule caused the difference. Recognising both is a stronger response than merely saying the statement is “misleading”.
Model answers and a compact error clinic
For question one, the means are 4.8 cm for A and 6.8 cm for B, so B starts 2.0 cm taller on average. For question two, A gains are 4, 6, 6, 5 and 6 cm, mean 5.4 cm. B gains are 5, 7, 6, 7 and 5 cm, mean 6.0 cm. The mean-gain difference is 0.6 cm.
For question three, Group B’s plants began taller, so their higher final average includes an initial advantage. Final height cannot be substituted for change. For question four, both groups have median gain 6 cm and range 2 cm, despite the different mean gains. For question five, the reported 2.6 cm is the gap between final mean heights rather than the gain difference, and the word “caused” exceeds what the data alone establish.
The main errors in this topic usually belong to a small number of categories: subtracting the wrong individual’s readings; using final values when the question asks for changes; overlooking an unequal baseline; confusing a percentage denominator; treating missingness as zero; and drawing a causal conclusion from a group contrast. Record which mistake occurred and repair that method. Doing another ten worksheets without correcting the cause is unlikely to produce the same improvement.
Two weeks from arithmetic confidence to examination judgment
Begin by practising five pairs of before-and-after readings with no statistical software. Calculate each change and label its unit. On the next day, find group starting, final and change means, and test the identity mean(final − initial) = mean(final) − mean(initial) when the pairs are complete. On day three, contrast mean and median. On day four, calculate range, quartiles under a stated convention and a standard deviation. On day five, compare absolute and relative change with named denominators.
On day six, draw two different graphs of the same data and explain how scale can alter the impression without changing the numbers. On day seven, answer an unseen word problem in which a group began with a higher baseline. The goal is to see whether you choose the correct variable before touching the calculator. In the second week, introduce a missing record, unequal group sizes and a sample-size limitation.
Finish with a full timed G3 Mathematics practice section and an error log. Categorise errors into selection, arithmetic, interpretation, unit, graph or working. Focus the next short practice session on the two most common categories. The strongest improvement often comes from learning to make one deliberate decision at the start: what exactly is the quantity being compared?
Questions parents and students ask
Why not simply compare the tallest plant? One extreme value does not describe the typical response of the group or account for starting height. A question asking for average change needs paired gain data.
Is Group B definitely better? Its mean absolute gain is larger in this fictional sample, but the median gains are equal and its biological advantage is not proved. “Better” requires a specified criterion.
Why calculate both centimetres and percentages? They answer different questions. Absolute gain reports added length. Relative gain reports change in relation to a defined starting value, and the denominator is essential.
Does calculating standard deviation prove the results are reliable? No. Standard deviation describes numerical spread in the recorded values. Validity also depends on the design, measurements and the interpretation.
Do students need to learn Biology for the Mathematics exam? No. The plant story supplies a realistic setting, but the mathematical skills are independent: paired change, averages, spread, graphs and interpretation of assumptions.
How can a parent check an answer quickly? Ask, “What did you subtract? What did you divide by? What unit do you have? Does the conclusion follow from that number?” Those questions reveal many errors without requiring a long lecture.
Where the route continues
To review the biological meaning of the same dataset, return to Volume 1013 on light, photosynthesis and seedling growth. To practise modelled quantities and careful financial interpretation, read Volume 1010. To improve the language used when sharing claims with a school audience, read Volume 1011. The EMS lighting evidence panel demonstrates how different disciplines can share evidence without confusing their standards.
For a coherent secondary-school route, use the Secondary Mathematics Sengkang capability map and the G3 Mathematics foundations family guide. The PSLE graph-scale comparison guide remains a useful starting point for younger students moving towards more demanding data interpretation.
The formal examination details come from the SEAB 2027 G3 school-candidate directory and official Mathematics K310 syllabus. Our calculations are original teaching data, not exam papers, official predictions or claims about actual school outcomes. In the next stage of this learning sequence, the same evidence can be developed into an English situational report that explains what the group actually found, what remains unknown and what action a school might reasonably consider.
