Many Primary 4 Science questions look like they are testing a topic—heat, light, magnets, materials, plants or another familiar chapter—but the hidden difficulty is often more basic: the student is comparing the wrong things.
A child may know the relevant Science and still reach a weak conclusion because they compare two final values with different starting points, compare two setups that differ in several important ways, or use a reference case that does not answer the scientific question being asked.
This preserved Hougang Primary 4 Science URL now owns that specific job: comparison architecture. It teaches how to choose what should be compared with what. The old duplicated 2019 tuition copy, obsolete schedules, location collisions and promotional claims have been removed.
This page is intentionally narrower than the existing Hougang Primary 4 pages on fair tests, variables, measurement quality, tables, graphs and diagrams. Those pages explain how to control, measure and read evidence. This page begins one step earlier: what comparison would actually answer the question?
A comparison needs a purpose
Two things can always be compared. That does not mean the comparison is scientifically useful.
Suppose a question asks whether the amount of light affects a plant-related outcome. Comparing Plant A on Monday with Plant B on Friday may produce numbers, but unless the comparison is structured around the light condition and other important factors, the result may not answer the question.
Teach the learner to state the relationship first:
I am comparing ______ because I want to find out whether ______ affects ______.
This sentence forces the comparison to serve the investigation instead of becoming an arbitrary pair of observations.
Four common comparison structures
Primary Science repeatedly uses a small number of comparison patterns. Recognising them reduces confusion.
1. Before versus after
The same system is observed before and after a change. The learner asks how much or in what way the state changed.
Examples include temperature before and after heating, length before and after growth, or position before and after a force acts.
2. With versus without
One setup contains the condition being investigated and another provides a reference without it.
This structure helps the learner ask what difference appears when the factor is present compared with when it is absent.
3. Lower versus higher level of one factor
Two or more setups contain different amounts, intensities or levels of the same relevant factor while other important conditions remain comparable.
4. Case A versus case B
Two objects, organisms or systems are compared because they differ in a scientifically relevant characteristic.
The difficulty is deciding which differences matter to the question and which are merely background differences.
A reference condition gives meaning to the changed condition
Imagine a plant grows 4 cm in a week under a new condition. Is that a lot? We cannot tell until we know what would normally happen under a suitable reference condition.
A reference case creates interpretive meaning. It helps answer:
- What would happen without the tested change?
- How much difference is associated with the changed condition?
- Is the observed effect larger, smaller or absent compared with the reference?
Primary students can use this logic even before they learn more formal experimental terminology.
Do not compare final values when starting values differ
This is one of the most common hidden comparison errors.
Suppose Plant A begins at 12 cm and ends at 17 cm. Plant B begins at 8 cm and ends at 15 cm. If the question is about growth during the period, comparing 17 cm with 15 cm is misleading. Plant A is taller at the end, but Plant B grew more.
The useful comparison is:
- Plant A: 17 − 12 = 5 cm growth
- Plant B: 15 − 8 = 7 cm growth
The scientific question determines whether final state or amount of change matters.
Matched starting conditions make differences easier to interpret
If two systems begin very differently, the final difference may reflect the starting state rather than the tested factor.
A fair comparison therefore often tries to match important starting conditions:
- similar initial size;
- similar starting temperature;
- same observation duration;
- same material amount;
- same apparatus arrangement except for the intended difference;
- same method of measuring the outcome.
The reason is not simply “because experiments must be fair”. The reason is that unmatched starting conditions create alternative explanations.
The comparison pair should differ in the relationship being tested
Suppose a learner wants to know whether material type affects water absorption. They compare a thick cotton cloth with a thin plastic sheet. If absorption differs, material type may matter—but thickness also changed.
The comparison becomes clearer when the cases are chosen so the scientifically relevant difference is isolated as far as practical.
Ask:
- Which difference between A and B is the question about?
- What other differences could also affect the outcome?
- Can we choose or design a better comparison pair?
This is comparison design, not merely variable naming.
Comparison direction matters
Students sometimes know that two values differ but express the relationship backwards.
For example, if Setup A takes less time than Setup B, then Setup A may be faster under the stated conditions. Saying “B is faster because its time is larger” reverses the relationship.
Teach the learner to verbalise the direction before explaining:
- A is higher than B.
- A is lower than B.
- A increased more than B.
- A changed while B remained approximately unchanged.
- A reached the same outcome in less time.
Once the relationship is stated accurately, the scientific explanation has a stable foundation.
Equal time matters when comparing change
If one system is observed for ten minutes and another for thirty minutes, a larger change in the second system may simply reflect the longer observation period.
When the question asks which condition causes faster change, the time window often needs to be comparable.
Ask:
- Were both cases given the same amount of time?
- Were measurements taken at comparable intervals?
- Does the question ask about final state or rate of change?
Time can be a comparison variable even when it is not drawn in the diagram.
The same outcome can still contain a useful comparison
Two setups may end with the same final result but reach it differently.
For example, both may reach the same temperature, but one may take less time. Both plants may end at the same height, but one may have started smaller. Both materials may become wet, but one may absorb more water.
The learner should ask which dimension the scientific question cares about:
- final state;
- amount of change;
- time taken;
- rate;
- presence or absence;
- another measured outcome.
“Same result” does not always mean “no difference”.
The different-outcome trap
The reverse is also true. Two outcomes can differ even when the tested factor is not the cause.
If Setup A and Setup B differ in several important conditions, the outcome difference is ambiguous. The student should resist the urge to attribute it immediately to the most visible changed factor.
A useful question is:
What else was different that could have produced this result?
This connects comparison architecture to fair testing without turning the page into another variables lesson.
Comparisons can be nested
Some questions contain several comparisons at once.
A table may show three temperatures over four time points. A diagram may compare two organisms before and after a change. A graph may contain two lines across several conditions.
Do not compare everything with everything. Use the question to choose the comparison path.
- Identify the outcome being asked about.
- Identify the relevant conditions.
- Select the two or more states that isolate the requested relationship.
- State the comparison verbally.
- Only then explain the Science.
This reduces cognitive overload in data-heavy questions.
The comparison matrix
When a question has several setups, a small comparison matrix can make the structure visible.
| Feature | Setup A | Setup B |
|---|---|---|
| Starting condition | ? | ? |
| Intended changed factor | ? | ? |
| Important conditions kept comparable | ? | ? |
| Measured outcome | ? | ? |
| Observed result | ? | ? |
The learner does not need to redraw a formal table in every exam question. The matrix is a training tool for seeing which similarities and differences matter.
Compare like with like
Meaningful comparisons usually require compatible quantities.
A child should not compare a temperature with a time value simply because both are numbers. Nor should a length measured in centimetres be compared directly with a volume in millilitres to claim one is “larger” in a scientific sense.
The quantities, units and meanings must match the relationship being tested.
This seems obvious when stated explicitly, but representation-heavy questions can hide the mismatch.
Comparison statements should be complete
“A is more” is incomplete. More what?
Better scientific comparison language includes:
- Setup A has a higher temperature than Setup B after the same period.
- Object A bends more than Object B under the same applied load.
- Plant A increased in height by a smaller amount than Plant B over the same number of days.
- Condition A produced no observable change while Condition B produced a clear change.
Precision in the comparison often improves the explanation automatically.
One comparison can support several possible explanations
A difference does not always point to one unique cause. If the comparison design leaves several important conditions unmatched, multiple explanations remain possible.
Teach the child to ask:
- What explanation does this comparison support?
- What alternative explanation is still possible?
- What new comparison would help distinguish them?
This transforms a weak investigation into a better next question.
Use a counterfactual reference
A helpful mental move is to ask:
What would I expect to happen if the tested factor were not changed?
That imagined reference helps the learner understand why a control or baseline is needed. It makes the comparison causal rather than merely descriptive.
Primary students do not need the word “counterfactual” to use the reasoning.
Comparison and classification are different jobs
Classification asks which characteristics place items into groups. Comparison asks how two or more cases are similar or different on a selected dimension.
A learner may compare two materials and then use the differences to classify them, but the operations should remain distinct.
This helps students avoid a common answer problem: giving a category label when the question asked for a difference, or listing differences when the question asked for the rule defining a group.
Comparison and explanation are different jobs too
“A is hotter than B” describes a relationship. It does not yet explain why.
Keep the stages separate:
- identify the comparison;
- state the observed difference;
- select the relevant scientific concept;
- explain the mechanism producing the difference.
Many weak Science answers jump from step one to step four and accidentally describe the wrong relationship.
Five Primary 4 comparison failure modes
1. Final-value fixation
The learner compares final states even though the question is about amount of change. Repair by always checking starting values.
2. Unmatched pair
The two cases differ in several important ways. Repair by identifying which alternative explanations survive.
3. Wrong comparison dimension
The student compares size when the question asks about time, or final value when it asks about rate. Repair by restating the measured outcome before comparing.
4. Direction reversal
The values are read correctly but the “more/less”, “faster/slower” relationship is reversed. Repair by verbalising the comparison before explaining.
5. Compare-everything overload
The learner tries to process every value or panel instead of selecting the pair relevant to the question. Repair by using the question to choose the comparison route.
A Phase 4 Primary 4 comparison lesson
- Question: state the relationship being investigated.
- Reference: identify the baseline or comparison condition.
- Match: check important starting conditions.
- Select: choose the two or more states that answer the question.
- Read: confirm units, values and observation duration.
- Compare: state the relationship precisely.
- Challenge: identify alternative explanations created by unmatched conditions.
- Explain: connect the comparison to the scientific concept.
- Redesign: improve the comparison if necessary.
- Transfer: use the same architecture in another Science topic.
The student learns that good comparisons are designed, not merely noticed.
Why small groups help with comparison design
Give three students four possible setups and ask which pair best answers one scientific question. Their choices reveal how they understand comparison.
The tutor can ask each student:
- Why did you choose this pair?
- What important condition matches?
- What relevant factor differs?
- What alternative explanation remains?
- Would another pair provide cleaner evidence?
The group becomes a design discussion rather than a worksheet-answering exercise.
What parents can practise at home
- When comparing growth, ask whether starting sizes were the same.
- When comparing speed, ask whether the distance or time window is comparable.
- When comparing two materials, ask what other properties differ.
- Ask “Compared with what?” whenever the child says “more” or “less”.
- Ask which pair of observations best answers the question.
- Ask what would happen without the changed condition.
- Ask whether final value or amount of change is the scientifically relevant comparison.
These short prompts teach comparison architecture without requiring parents to reteach the whole topic.
What evidence to bring when comparison is the suspected bottleneck
- a before-and-after question;
- a two-setup experiment;
- a graph with more than one line;
- a table with different starting values;
- the learner’s original comparison statement;
- teacher corrections;
- one question where the Science concept was known but the wrong values were compared;
- the child’s explanation of why they chose that comparison.
This reveals whether the failure occurs before the concept is applied.
How to tell whether comparison thinking is improving
- The student states the scientific relationship before choosing a pair.
- Reference conditions are identified more reliably.
- Starting values are checked before final values are compared.
- Amount of change is distinguished from final state.
- Observation duration is treated as a relevant condition.
- Comparison direction is stated correctly.
- Alternative explanations are noticed when setups differ in several ways.
- The learner selects only the evidence needed for the question.
- The same comparison architecture transfers across different Science topics.
These changes make later variables, data and evaluation work far more stable.
How this page fits the Hougang Science network
This eduKateSingapore page owns comparison architecture and reference conditions. It complements Hougang Primary 4 Science | Measurement, Units and Reliable Evidence, Primary 4 fair tests, variables and evidence, and Primary 4 tables, graphs and diagrams as evidence.
For the national subject map, continue to What Is Primary Science Education? | From Curiosity to Scientific Thinking, P3 to PSLE.
Official curriculum reference
The Ministry of Education’s Science Teaching & Learning Syllabus: Primary Three to Six develops scientific practices such as comparing, observing, analysing information, conducting inquiry and communicating evidence-based reasoning across the primary years.
A good Primary 4 comparison is not “A versus B” because both are on the page. It is “A versus B because this pair isolates the relationship the question is asking about.” Once the comparison is correct, the Science has a fair chance to work.