Choose your next learning step
Practise G3 Science predictions when two factors change: separate individual effects, use equations or data, and explain when the net result remains uncertain.
1. Orient · 2. Work through · 3. Compare · 4. Apply · 5. Practise and continue
G3 SEC Science prediction questions become harder when two things change at once. A learner may correctly remember the effect of each factor and still reach an unsupported final answer. If one change pushes an outcome upwards and another pushes it downwards, knowing the directions is not always enough to know the result.
This workshop teaches a calm sequence: identify the outcome, separate the changes, state the conditions behind each relationship, then decide what the evidence allows you to predict. Sometimes an equation gives a definite answer. Sometimes a supplied table does. Sometimes the correct conclusion is that the net effect cannot be determined from the information given.
You will practise with original numerical and qualitative cases from Physics, Chemistry and Biology. Use the cases that match your subjects. The aim is to explain a prediction with its basis and limits, rather than answer every “what happens?” question with a memorised increase or decrease.
Chapter index
Chapters 1–3
Chapters 4–6
Chapters 7–9
Suppose an outcome depends on two factors. In the stated situation, increasing factor A tends to increase the outcome, while increasing factor B tends to decrease it. If both increase, three results are possible: the outcome may rise, fall or remain unchanged. The result depends on the relative effects, not only their directions.
An everyday mathematical example makes this clear. Average speed equals distance divided by time. If a journey becomes longer and takes more time, you cannot determine the new average speed merely from those two increases. You need their sizes. A 20% increase in distance and a 10% increase in time differs from a 10% increase in distance and a 20% increase in time.
Scientific questions often have the same structure, although their relationships may be more complicated. Two influences may oppose one another. One factor may matter only in a certain range. A factor that was previously limiting may stop being limiting after another condition changes.
Your first job is to name the outcome precisely. “What happens to the experiment?” is too vague. Is the question about current, initial reaction rate, final product amount, temperature rise, photosynthesis rate or a measured signal? These are different quantities.
Next list the changes without interpreting them yet. For example: potential difference increases; resistance increases. Only after that should you attach the relevant relationship. This prevents the first familiar keyword from taking control of the answer.
Then consider what remains constant. Some scientific relationships support a simple prediction only under stated conditions. If those conditions also change, a one-factor rule may not settle the result. A careful answer can say which condition it depends on.
Do not interpret this as an instruction to answer “cannot tell” whenever there are two variables. When the question supplies a usable equation or enough measurements, you should use them. Uncertainty is a conclusion to justify, not a refuge from calculation.
Likewise, do not assume that two increases cancel because they look balanced in words. “More concentrated” and “cooler” are not numerical opposites. They operate through different features of the system. Without a quantitative relationship or relevant data, you cannot subtract one adjective from another.
A useful practice note has three lines: outcome; individual influences; evidence for combination. If the third line is empty, investigate whether a definite net prediction is actually supported. That small pause often prevents an overconfident answer.
There is no single G3 Science paper covering every learner in the same way. SEAB’s 2027 list identifies Physics K323, Chemistry K324 and Biology K325. The combined Science subjects are Physics/Chemistry K326, Physics/Biology K327 and Chemistry/Biology K328. Use the official school-candidate index to identify your actual route.
For the separate Physics, Chemistry and Biology syllabuses, the assessment comprises a one-hour multiple-choice paper, a 1 hour 45 minute structured/free-response paper and a 1 hour 50 minute practical paper, weighted 30%, 50% and 20%. Check the relevant Physics, Chemistry or Biology document.
For combined Science, the scheme requires Paper 1, Paper 5 and the two relevant subject papers. Multiple choice carries 20%, the two subject papers 32.5% each and practical 15%; their durations are one hour, 1 hour 15 minutes per subject paper and 1 hour 30 minutes for practical. See the combined Science syllabus.
These distinctions matter because a learner should not borrow another route’s practical weighting or paper duration. They do not change the central reading habit in this workshop: identify what changes and justify what follows.
“Competing changes” is our descriptive name for the practice problem. It is not an official section title or a promise about the questions in a future paper. The examples help you connect subject knowledge with unfamiliar information.
Some cases below use an explicitly stated simplified model. Treat that model as part of the exercise. Do not assume it describes every real device, organism or reaction. Scientific reasoning includes knowing when a model is being used and which conditions make it useful.
If you have not studied the underlying topic, return to your topic lesson first. A prediction workshop cannot replace knowledge of circuits, reaction rates or photosynthesis. It can show you how to use that knowledge once you have it.
The existing Science Hub remains the route for broader learning. Earlier Volume 0068 on controls and alternative explanations examines the design of comparisons. Here the job is different: judge the combined effect of specified changes.
MOE’s SEC explanation provides the wider examination setting. The worked answers below are teaching explanations, not official marking schemes. Follow your actual syllabus, teacher’s guidance and the instructions in the question.
CHAPTER 3 OF 9
3. Physics worked case: potential difference and resistance both increase
Consider a resistor described by the relationship I = V/R for the stated conditions. In State A, the potential difference across it is 6.0 V and its resistance is 12 Ω. In State B, the potential difference is 9.0 V and the resistance is 18 Ω. Predict the change in current.
State A gives I = 6.0/12 = 0.50 A. State B gives I = 9.0/18 = 0.50 A. The current remains unchanged in this exercise. The increased potential difference and increased resistance have the same multiplicative factor, 1.5.
A weak answer is “current increases because voltage increases”. It considers only one change. Another weak answer is “current decreases because resistance increases”. It considers the other change alone. A third weak answer is “two changes cancel” without showing why. The calculated ratio explains the result.
Now alter State B to 9.0 V and 24 Ω. Current becomes 9.0/24 = 0.375 A. The potential difference rises by a factor of 1.5, while resistance doubles. Current therefore falls by a factor of 1.5/2 = 0.75.
Alter State B again to 12 V and 18 Ω. Current becomes 12/18 A, approximately 0.667 A. The potential difference doubles while resistance rises by 1.5, so current increases by a factor of 2/1.5.
The direction labels are identical in all three versions: V increases and R increases. Yet the outcomes differ. This is a clean demonstration of why a qualitative list of influences does not always establish the net result.
You can compare states without calculating both currents separately. The ratio I_B/I_A equals (V_B/V_A)/(R_B/R_A), provided the given relationship applies to each state. A ratio above one means an increase; below one means a decrease; equal to one means no change.
Keep the potential difference attached to the relevant component. A source voltage is not automatically the voltage across every component in an arbitrary circuit. If the circuit arrangement matters, determine the component values from that arrangement before using the relationship.
Also respect the stated resistance in each condition. If a question gives resistance values for the two states, use them. Do not silently assume resistance is constant merely because you remember a simple classroom case. Conversely, if it explicitly says a resistance is constant, preserve that condition.
The syllabus relationship can be checked in the G3 Physics document. This exercise’s numbers and comparisons are original. The useful performance habit is to calculate the relation between the changed quantities before committing to a directional answer.
CHAPTER 4 OF 9
4. Chemistry worked case: two factors push the rate in opposite directions
Consider an original comparison in which the same stated reaction is carried out under two sets of conditions. In the second set, the solution is more concentrated but the temperature is lower. The question gives no numerical model of rate and no measurements. Can you confidently say that the initial rate is greater?
Under the usual relevant classroom conditions, increased concentration can increase the frequency of collisions between reacting particles, while lower temperature can reduce the frequency of successful collisions. These influences oppose one another. Their directions alone do not specify which effect is larger.
A defensible answer is that the net change in initial rate cannot be determined from the given information. Explain why: the concentration change tends to increase the rate, whereas the temperature change tends to reduce it, and their relative effects have not been established.
Do not write that the rate must remain the same. Opposing influences are not automatically equal. Do not write that the temperature effect always wins. The question supplies no basis for that ranking. Both answers pretend to know something that has not been given.
Now suppose the question supplies measured initial rates: Condition A gives 1.8 cm³/s and Condition B gives 2.4 cm³/s using the same defined measurement method. You can then state that the measured initial rate is higher in B. The data settle that comparison under those tested conditions.
The data still do not isolate the contribution of each factor. Because concentration and temperature changed together, the comparison does not show how much of the net difference was caused by either one alone. A net prediction and a causal attribution are different jobs.
To investigate each factor separately, you would need suitable comparisons in which the other relevant conditions are controlled. This connects with the earlier controls workshop, but the present question can be answered without designing an entire investigation.
Be precise about rate and amount. A higher initial rate does not automatically mean a larger final product amount. The final amount depends on the reacting quantities and other relevant conditions. If the question asks about final gas volume, do not answer only with a sentence about collision frequency.
For instance, an exercise may specify the same quantity of a limiting reactant in both runs and sufficient excess of the other reactant, with complete reaction and comparable gas-measurement conditions. It may then show different initial rates while the final product amount is unchanged. The rate describes how quickly the process proceeds; the final amount describes its eventual extent.
For topic scope and the treatment of reaction-rate factors, consult the 2027 G3 Chemistry syllabus. The performance step here is to keep the requested outcome separate from a neighbouring quantity that happens to use the same experiment.
Photosynthesis questions provide another setting in which a one-factor rule can be overextended. The G3 Biology syllabus includes the effects of light intensity, carbon dioxide concentration and temperature, including their roles as limiting factors. A prediction must therefore consider the conditions, not only the name of a factor.
Use this original simplified dataset. At lower carbon dioxide concentration, measured oxygen-production rates are 2, 3 and 3 arbitrary units at low, medium and high light intensity. At higher carbon dioxide concentration, the corresponding rates are 2, 5 and 7 units. Assume the exercise controls the other relevant conditions and uses the same measurement method throughout.
At lower carbon dioxide concentration, moving from medium to high light does not increase the measured rate in these data. At higher carbon dioxide concentration, the same light change increases it from five to seven. The effect of changing light depends on the starting conditions.
This does not mean that light becomes unimportant in all situations. It means that under the lower-carbon-dioxide conditions and tested range, extra light does not produce a further measured increase. The dataset supports that local conclusion.
Now compare two simultaneous changes. Move from medium light with higher carbon dioxide to high light with lower carbon dioxide. The measured rate changes from five to three. More light does not prevent a lower rate in this comparison because the carbon dioxide condition also changes.
A learner who sees only “light increases” may predict an increase. A learner who reads both changes but refuses to use the supplied table may answer “cannot determine”. Both miss information. The table gives the measured comparison directly.
If the destination were a condition not in the table, the situation would differ. Suppose the new carbon dioxide level lies outside the tested values and the question supplies no model. You should not invent an exact rate by assuming the rows extend unchanged.
Also distinguish the measured indicator from the underlying biological process. In this simplified exercise, oxygen production is the defined indicator. A real measurement can require attention to collection method, time interval and other processes. Follow the question’s supplied method instead of assuming every bubble count is a perfect measure.
The answer should say what the data show and, when required, use biology to explain it. “The rate falls from five to three in the supplied data despite increased light, because the carbon dioxide condition is also reduced” describes the comparison. Further explanation about limitation should remain consistent with the data and the stated conditions.
Keep the conclusion at the tested scale. A short exercise on one plant setup does not establish a universal quantitative rule for every species, temperature or environment. Accurate scientific prose can be specific without becoming evasive.
A table can support a net comparison even when it cannot isolate causes. That distinction is one of the most useful ideas in this workshop. You can know that the outcome changed without knowing how much each altered factor contributed.
Imagine a device tested under four original conditions. With setting A low and setting B low, its output is ten units. With A high and B low, output is fourteen. With A low and B high, output is seven. With both high, output is twelve. These are supplied exercise values, not claims about a real device.
The combined change from both low to both high raises output from ten to twelve. That is directly supported. But simply adding the separate changes, plus four and minus three, would predict eleven. The actual supplied combined value is twelve. The exercise therefore does not support simple additivity.
If the final cell were missing, you could not fill it confidently without an additional assumption or model. You might propose eleven under an explicitly stated additive model, but you must identify that assumption. A model-based estimate is different from a measured value.
This applies to biological responses, experimental measurements and many unfamiliar contexts. Do not assume that effects combine linearly merely because individual changes are easy to describe. The question may provide a graph, equation or table that specifies how to combine them.
Graphs require the same care. Read the curve for the relevant condition, then compare the correct horizontal position. A curve labelled “higher concentration” does not describe the lower-concentration case. A value at ten minutes does not answer a question about the initial rate.
Interpolation and extrapolation also differ. Estimating between nearby measured values may be reasonable under the question’s instructions. Extending far beyond the tested range requires stronger assumptions about the relationship. A straight-looking section is not evidence that the entire process remains straight indefinitely.
If uncertainty ranges are supplied, use them as instructed. Two displayed means may differ while their measurement uncertainty makes the strength of the comparison less clear. Do not invent uncertainty figures when none are supplied, but do not ignore them when they are.
An exact calculation from a model is exact only within that model and its inputs. It does not automatically make a real-world prediction exact. In an exercise, state the result clearly; in an evaluation, identify the conditions that make the model useful.
A good final sentence names its basis: “The supplied measurements show…”, “Using the stated equation…”, or “The net effect cannot be determined because…”. That phrasing helps the reader see why the answer deserves its level of confidence.
A clear prediction answer has a conclusion and a reason. The reason may be a calculation, a comparison in supplied data or an explanation of why the influences cannot yet be combined. You do not need to write everything you know about the topic.
For the resistor example, a compact complete answer is: “The current remains 0.50 A because both V and R increase by a factor of 1.5, so V/R is unchanged.” The outcome, numerical value and relationship are all visible.
For the chemistry example without rate data, a compact answer is: “The net change in initial rate cannot be determined: increased concentration tends to raise the rate, while lower temperature tends to reduce it, and their relative effects are not given.” This is a reasoned limit, not a guess.
For the photosynthesis dataset, you could write: “The measured rate falls from five to three units when both conditions change; the increase in light does not outweigh the change in carbon dioxide in this supplied comparison.” The exact phrasing should reflect the question and what the data establish.
Be careful with the word “outweigh”. It describes the net comparison but does not quantify each factor’s separate causal contribution. If the question asks for isolated effects, you need the appropriate controlled comparisons or model rather than only the combined result.
Avoid vague sentences such as “it depends on many factors”. They may be true in general, but they do not identify the factors relevant to the question. Name the competing influences and the missing information. Specific uncertainty is useful; general uncertainty is often unhelpful.
Avoid absolute phrases such as “always increases” when your relationship has a stated range or condition. A qualified answer can still be direct: “Within the range shown, the measured rate increases.” The qualification tells the reader where the conclusion applies.
If asked to suggest what additional information is needed, identify a practical target. For a numerical relationship, you may need the changed values. For a qualitative opposing-effects case, you may need measured rates under the combined conditions or a validated relationship. Do not merely request “more data” without saying which data.
If asked to design a test, then consider controls, measurement, repeats and a useful range of conditions. But do not turn every prediction answer into a full experimental plan. Answer the reader’s or examiner’s requested job first.
A useful final check is to circle the requested outcome in the question and underline the outcome in your answer. If one says final volume and the other says initial rate, you have answered a neighbouring question. Correct the target before refining your scientific vocabulary.
Precise answers often become shorter after this check because irrelevant facts can be removed.
Attempt these cases independently. For each one, state whether the evidence supports an increase, decrease, no change or an unresolved net effect. Give the basis.
Case A: A stated resistor has 4 V across it and resistance 8 Ω. In the second state, the values are 10 V and 20 Ω. Both values increase. What happens to current?
Current is 0.50 A in both states, so it is unchanged. The common scale factor is 2.5. A sentence about voltage alone or resistance alone would omit relevant information.
Case B: The first state is 4 V and 8 Ω. The second is 6 V and 16 Ω. What happens to current?
It changes from 0.50 A to 0.375 A, a decrease of 25%. The V factor is 1.5 and the R factor is two. Their ratio is 0.75. This is a calculable net result, not an unresolved opposing-effects case.
Case C: In an appropriate reaction-rate comparison, concentration is increased and temperature is lowered. No rates, magnitudes or quantitative relationship are supplied. What can you conclude about the new initial rate?
The net direction is not determined by the supplied information. Explain the opposing influences. Saying that they cancel would require evidence of equal effects, which has not been supplied.
Case D: The same two changes are made, but measured initial rates are 1.2 and 0.9 cm³/s before and after. What does the comparison show, and what does it not isolate?
It shows that the measured initial rate is lower in the second condition. It does not separately quantify the concentration and temperature contributions. The numerical difference alone cannot allocate responsibility between the factors.
Case E: A photosynthesis exercise supplies rate four at medium light/high carbon dioxide and rate six at high light/high carbon dioxide. Predict the change when moving between those exact two conditions.
The rate increases from four to six, a 50% increase relative to four. Carbon dioxide is unchanged in this comparison. Do not describe it as a simultaneous competing change.
Case F: The table gives rate four at medium light/high carbon dioxide but no value for high light/low carbon dioxide. Can you state that the new rate is six?
No. Six belongs to a different stated condition in Case E. Reusing it would lose the condition label. A measurement or suitable model is needed for the missing combination.
Case G: Two reaction curves reach the same final measured gas volume, but one is steeper near the start. Does the steeper curve prove more final product?
No. It indicates a greater initial rate in the displayed comparison; the equal final volumes show the same measured endpoint. Keep the rate and final amount separate.
Review your wrong answers by category: ignored change, wrong condition label, confused outcome, assumed cancellation or refused available data. Each category calls for a different repair. Recalculate only when the arithmetic was the problem; reread conditions when the model was the problem.
Begin with a relationship you understand well. Explain what happens when one factor changes while the others relevant to that relationship remain fixed. Then add a second change and ask whether the evidence allows the effects to be combined.
Use three kinds of practice deliberately. First, an equation case in which the net result is calculable. Second, a table case in which the exact combined condition is measured. Third, a qualitative case in which opposing influences are known but their relative sizes are not. Learning to distinguish these cases is the core progression.
Do not make every exercise an uncertainty exercise. That can teach a learner to avoid committing even when the evidence is sufficient. Equally, do not supply a definite answer to every case. That can teach false certainty. A balanced set rewards the right conclusion for the available information.
After a mistake, change only one feature. Supply the missing resistance value, add the combined-condition table cell or specify that temperature remains constant. Ask how the answer changes. This makes the role of information visible.
Then change the topic while preserving the reasoning structure. A voltage/resistance case can be followed by a distance/time model or a supplied unfamiliar biological table. The learner should recognise the need to compare magnitudes or conditions without pretending the underlying science is identical.
Parents and tutors can ask three calm questions: “What exactly are you predicting?” “What changed?” “What lets you combine those changes?” These questions reveal the reasoning without demanding a memorised model answer.
If a learner knows the factual rule but cannot apply it, ask them to draw a two-state table. If they cannot explain the factual rule, return to the topic lesson. Distinguishing the two difficulties keeps the support focused.
For continuity with earlier work, use Volume 0076 on rate versus amount when the predicted quantity is unclear. Use Volume 0084 on baselines when a measurement requires correction before comparison. Neither is replaced by this workshop.
Return through the Science Hub for subject learning and the G3 EMS Learner’s Guide for the larger sequence. For primary foundations, the existing PSLE Science revision guide supports explanation rather than model-answer copying.
Official sources checked on 13 October 2026, Singapore time: SEAB’s 2027 G3 index, the linked Physics, Chemistry, Biology and combined Science syllabuses, and MOE’s SEC explanation. All datasets here are original teaching data. The exercises are for reasoning practice; carry out practical work through your school’s supervised arrangements.
