Wait, What? The Best Comparison Is Not Always the Set-Up That Looks Most “Normal”
When a PSLE Science investigation shows several set-ups, students often look for one that feels like the “control” and compare everything with it. That can work sometimes. It can also produce the wrong comparison.
A useful reference set-up is not chosen by appearance, alphabet order, position on the page or the idea that “nothing should happen” there. It is chosen because the comparison answers the scientific question while keeping other relevant conditions aligned.
CHOOSE THE REFERENCE THAT MAKES ONE SCIENTIFIC DIFFERENCE INTERPRETABLE.
Quick Answer
Before choosing a reference set-up, use this chain:
READ THE SCIENTIFIC QUESTION → IDENTIFY THE OUTCOME BEING COMPARED → IDENTIFY THE CONDITION OR RELATIONSHIP UNDER TEST → FIND SET-UPS WITH COMPARABLE STARTING STATE, TIME AND MEASUREMENT → CHOOSE THE PAIR THAT ISOLATES THE RELEVANT DIFFERENCE AS FAR AS THE METHOD ALLOWS → STATE WHAT THAT COMPARISON CAN SHOW → STATE WHAT IT CANNOT SHOW.
Do not ask, “Which one is the control?” first. Ask, “Which comparison gives the evidence I need?”
The Exact PSLE Science Learning Job This Guide Owns
This guide owns one learner job: choosing the most informative reference set-up from several possible comparisons in a PSLE Science investigation or data question.
It does not own the general definition of a control set-up, controlled variables, fair tests or indirect comparison. Those already have separate learning jobs. Here, the question is narrower and more practical: when several set-ups are available, which one should serve as the reference for the claim you are trying to test?
Why This Matters in Current PSLE Science
For examination from 2026, PSLE Science assesses attainment in the 2023 Primary Science syllabus. The official assessment frame includes applying scientific concepts and principles, interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning. Choosing a valid comparison is therefore part of understanding what evidence can support a claim.
This guide does not claim that every PSLE Science investigation must contain a specially labelled control set-up. Some questions ask for direct comparisons between treatment conditions, some use a reference or baseline, and some do not need a separate control at all.
Start With the Claim, Not the Apparatus
Imagine four set-ups, A, B, C and D. They differ in light, temperature and water supply. If the question asks about the effect of light, the most useful reference is not simply Set-Up A. You need a pair in which light differs while the other relevant conditions are comparable.
If the question instead asks about temperature, the correct comparison may be a different pair. The same set of apparatus can therefore contain more than one scientifically useful reference relationship.
The Five-Part Reference Test
| Check | Question to ask | Why it matters |
|---|---|---|
| Outcome | Am I comparing the same measured or observed quantity? | Different outcomes cannot answer one comparison directly. |
| Time | Were the results taken after comparable durations or at comparable stages? | A later measurement can create a false difference. |
| Starting state | Were the set-ups reasonably comparable before the tested condition acted? | A pre-existing difference can be mistaken for an effect. |
| Relevant condition | Does the pair differ in the condition needed for the claim? | If the condition does not differ, the comparison cannot test its effect. |
| Other influential conditions | Are other relevant factors aligned as far as the method requires? | Several simultaneous differences weaken causal interpretation. |
Worked Example 1 — Three Set-Ups, Two Possible Comparisons
An original practice scenario uses three identical containers with equal amounts of water. A is placed in moving air, B in still air, and C in moving air but with a lid. The outcome is the mass of water remaining after the same time.
If the learning question is about moving versus still air, compare A with B because the lid condition is the same. If the question is about covered versus uncovered under moving air, compare A with C.
There is no single permanent “control” for every possible question. The informative reference depends on the relationship being tested.
Worked Example 2 — The Most Familiar Set-Up Is a Bad Reference
Suppose Set-Up P is at room conditions, Q has a different temperature, and R has both a different temperature and a different exposed surface area. A learner compares R with P to explain the temperature effect because P looks like the ordinary baseline.
That pair changes two scientifically relevant conditions. Q versus P is the cleaner reference comparison for temperature if the other relevant conditions are comparable.
The lesson: reference quality comes from alignment, not familiarity.
Worked Example 3 — A Reference Can Be an Initial Value
Sometimes the reference is not another physical set-up. A question may compare the same specimen before and after a condition is applied. The starting measurement becomes the reference for the change.
In that case, check whether the same scientific quantity is measured before and after, whether the specimen identity is preserved, and whether the treatment itself creates other changes that affect interpretation.
Worked Example 4 — A Reference Does Not Mean “Nothing Happens”
Imagine two seeds growing under different conditions. The reference group also grows. Its purpose is not to remain unchanged. Its purpose is to provide a comparison under a different defined condition.
A reference can have active processes and changing measurements. What matters is what difference the comparison lets you interpret.
When Several Reference Set-Ups Are Legitimately Useful
A complex investigation can support several pairwise questions. One pair may isolate temperature. Another may isolate light. A third may compare a treatment with an untreated condition. Do not force every scientific question to share one reference if the evidence structure does not work that way.
Instead, write the comparison in a compact scientific form:
SET-UP X versus SET-UP Y → same measured outcome and comparable background conditions → relevant difference = ___ → therefore this pair can test ___.
Reference Set-Up Versus Controlled Condition
A reference set-up is something you compare against. A controlled condition is a condition intentionally kept comparable across the set-ups so it does not create an alternative explanation.
Students often mix these ideas because both support fair comparison. Keep the jobs separate: one is the comparison case; the other is what stays aligned.
Reference Set-Up Versus Reference Value
A reference value can also be a known or starting measurement used for checking or comparison. It does not have to be a complete experimental set-up. Read the question carefully before assuming that “reference” means another container, organism or apparatus arrangement.
Observable Failure Signatures
| Failure signature | Earliest weak link | Repair |
|---|---|---|
| Always chooses Set-Up A as the control | Page position is being used as scientific meaning | Rewrite the claim first, then select the comparison. |
| Chooses the set-up where “nothing happens” | Reference confused with inactivity | Ask what scientific contrast the reference provides. |
| Compares two set-ups with different outcomes or time points | Measurement alignment failed | Match quantity and time before interpretation. |
| Uses a pair with two important differences to claim one cause | Alternative explanation not controlled | Find a cleaner pair or narrow the conclusion. |
| Says every investigation must have one control | Memorised rule replaced method reading | Identify the evidence structure of this investigation. |
| Cannot explain why the chosen reference is useful | Reference selected by pattern recognition | State the one difference the comparison makes interpretable. |
Earliest-Weak-Link Diagnosis
- What exact scientific question or claim am I testing?
- What outcome must be compared?
- Which set-ups contain that outcome measured in a comparable way?
- Which relevant condition should differ?
- Which other relevant conditions should stay comparable?
- Are starting states and time points aligned?
- Which pair now gives the cleanest evidence?
- What alternative explanation still remains?
Misconception Repair — “The Control Is the One With No Variable”
Every real set-up has conditions. The useful distinction is whether a tested condition differs across the comparison, not whether one set-up somehow has “no variables”. Practise naming the actual condition in each set-up.
Misconception Repair — “Everything Must Be Identical Except One Thing”
A fair comparison requires relevant conditions that could influence the measured outcome to be controlled as appropriate. Decorative or scientifically irrelevant sameness does not strengthen the evidence. Learn to justify why a condition matters instead of chanting “everything the same”.
Misconception Repair — “The Biggest Difference Gives the Best Evidence”
A dramatic difference between two set-ups is not automatically the cleanest evidence. If several relevant conditions differ, the cause of the difference may be unclear. Evidence quality and effect size are different questions.
Practice Protocol: Claim → Pair → Difference → Limit
- Write one claim the investigation could test.
- List every available set-up.
- Cross out pairs that do not measure the same outcome.
- Cross out pairs that use mismatched times or stages.
- Mark the relevant condition that differs in each remaining pair.
- Check other influential conditions.
- Choose the most informative reference pair.
- Write one sentence explaining what the pair supports.
- Write one sentence explaining what it cannot prove.
Unfamiliar Transfer Challenge
Create four imaginary set-ups that vary in two conditions, such as temperature and exposed area. Give all four the same measured outcome. Then write two different scientific questions and choose a different reference pair for each.
If one pair is unsuitable, explain the exact extra difference that weakens it. This is more powerful than simply labelling one set-up “control”.
Delayed Independent Return
Several days later, use a new multi-set-up investigation. Before looking at any answer, write the claim, identify the outcome, choose the reference pair and justify the choice. Then test whether changing the claim would change the best reference.
Reference-Choice Receipt
- I start from the scientific claim, not the position of the set-up.
- I compare the same measured outcome.
- I align time and starting state where relevant.
- I identify the condition that makes the comparison informative.
- I check other conditions that could affect the outcome.
- I do not assume a reference means “nothing happens”.
- I can state what the comparison supports and what it cannot establish.
Parent and Tutor Teaching Guide
Do not begin by asking, “Which is the control?” Ask the learner, “What are you trying to find out?” Then make the learner point to the two set-ups that create the cleanest comparison for that question.
Next ask, “If I changed the scientific question, would you still use the same reference?” This reveals whether the learner understands comparison structure or has memorised one set-up as permanently special.
When the learner chooses a poor pair, do not supply the correct pair immediately. Ask which other relevant condition differs. That small question often exposes the weak link more effectively than reteaching the whole fair-test chapter.
Useful Internal Routes
- PSLE Science Learning Guide
- How to use a control set-up without treating it as the set-up where nothing happens
- Controlled variable versus control set-up
- Indirect comparison through a common reference
- Which conditions actually need to stay the same
Authoritative References
- Singapore Examinations and Assessment Board — PSLE Science syllabus, for examination from 2026
- Ministry of Education, Singapore — Science Teaching & Learning Syllabus, Primary, 2023
- National Research Council — A Framework for K–12 Science Education, used as broader science-practice context rather than PSLE marking policy.
Evidence and Boundary Note
This guide teaches comparison design and evidence reasoning. It does not claim that every PSLE Science question uses the terms “reference set-up” or “control set-up”, and it does not impose one universal experimental template. The learner should always follow the actual information, conditions and scientific job in the question.
The Quiet Return
A good reference is not the set-up that looks simplest.
It is the comparison that makes the scientific difference readable.
