Wait, What? Two Correct Numbers Can Make the Wrong Comparison
A question gives two set-ups, each measured at the beginning and the end.
Set-up P starts at 30 and ends at 40. Set-up Q starts at 45 and ends at 48.
A learner sees 48 and says, “Q changed more because its final value is higher.”
The numbers are read correctly. The comparison is not.
Before comparing values, decide what the scientific question is asking you to compare.
Sometimes the right comparison is the same set-up before and after. Sometimes it is different set-ups at the same time or condition. Sometimes you need both, but in a deliberate order.
Many PSLE Science errors happen because a learner compares numbers that are individually relevant but belong to different comparison jobs.
Quick Answer
Use a before–after comparison when the question asks how one set-up changed over time or after a condition changed. Use a set-up–to–set-up comparison when the question asks how two different conditions affect the same measured outcome at a matched time or stage.
Before calculating or explaining, align four things:
- the scientific object or set-up;
- the measured quantity;
- the time or stage;
- the condition being compared.
READ THE QUESTION TARGET → NAME THE OBJECTS → ALIGN THE QUANTITY → ALIGN THE TIME OR STAGE → CHOOSE WITHIN-SET-UP OR BETWEEN-SET-UP COMPARISON → DESCRIBE THE DIFFERENCE OR CHANGE → SELECT THE SCIENTIFIC CONCEPT → EXPLAIN THE MECHANISM → CHECK THAT YOUR COMPARISON ANSWERS THE QUESTION.
The Exact PSLE Science Learning Job This Guide Owns
This guide owns one job: how a Primary 5 or Primary 6 learner chooses whether a PSLE Science question requires a within-set-up before–after comparison or a between-set-up comparison at a matched time or condition.
It does not replace fair-test reasoning, the general guide to comparing three or more set-ups, or the guide on different starting values. Those remain separate owners. This page teaches the earlier decision that makes all of them work:
which two observations actually belong together for this question?
Why This Matters in the Current PSLE Science Frame
For examination from 2026, PSLE Science assesses the 2023 Primary Science syllabus. SEAB’s assessment objectives include interpreting and analysing information, evaluating observations and methods, applying scientific facts and concepts, and communicating explanations and reasoning.
Comparison is therefore not just subtraction. It is evidence selection. The learner must decide which observations answer the scientific question before using them.
Two Comparison Families
| Comparison family | What stays the same? | What changes? | Typical question job |
|---|---|---|---|
| Within one set-up: before vs after | Same set-up/object | Time, stage or treatment state | How did this set-up change? |
| Between set-ups: P vs Q | Matched time/stage and measured quantity | Tested condition or set-up | How did the different condition affect the outcome? |
These two comparison families answer different questions.
Within-Set-Up Comparison: Follow One Thing Through Change
Suppose an original investigation records the mass of a wet cloth:
| Set-up | Initial mass | Mass after 30 min |
|---|---|---|
| P | 120 g | 105 g |
| Q | 150 g | 142 g |
If the question asks, “How much did the mass of cloth P change?”, the correct pair is P before and P after.
You do not need Q yet.
P changed by 15 g. That is a within-set-up comparison.
Between-Set-Up Comparison: Match the Observation Point
If the question asks, “Which cloth has the smaller mass after 30 minutes?”, compare P and Q at the same observation time:
- P after 30 min = 105 g.
- Q after 30 min = 142 g.
That is a between-set-up comparison.
But if the question asks which cloth lost more mass, final mass alone is not enough because the starting masses differ. You first calculate change within each set-up, then compare those changes.
The Two-Stage Comparison
Some questions require both comparison families:
- Compare each set-up with itself to calculate or identify its change.
- Then compare the changes between set-ups.
Using the cloth data:
- P: 120 g → 105 g, loss of 15 g.
- Q: 150 g → 142 g, loss of 8 g.
Now the between-set-up comparison is between 15 g loss and 8 g loss, not between 105 g and 142 g.
Sometimes you must compare within each set-up first so that the thing you compare between set-ups is scientifically equivalent.
Worked Example 1 — Temperature Before and After Heating
Two materials are heated for the same duration. Original data:
| Material | Initial temperature | Final temperature |
|---|---|---|
| A | 25°C | 38°C |
| B | 31°C | 41°C |
Question: Which material showed the greater temperature increase?
Wrong comparison: 38°C vs 41°C.
Correct route:
- A: increase of 13°C.
- B: increase of 10°C.
Therefore, A showed the greater measured temperature increase.
The question asks about change, not final state.
Worked Example 2 — Same Time, Different Conditions
Two identical containers of water are placed under different airflow conditions. Both start at the same mass. After one hour, P has 92 g and Q has 86 g.
If the question asks which condition produced greater water loss, either route works because the starting values are matched:
- compare final masses, knowing the starting masses were equal; or
- calculate each mass loss and compare the losses.
The scientifically safer habit is to keep the question target explicit: mass lost.
Worked Example 3 — Different Times Cannot Be Compared Casually
Set-up P is measured after 10 minutes. Set-up Q is measured after 30 minutes. Q has a larger change.
Can you conclude that Q’s condition caused the faster change?
Not from that comparison alone. The observation times are not matched. Q had more time for the process to occur.
A fair between-set-up comparison usually requires aligned time or another common observation point relevant to the question.
Worked Example 4 — Same Organism, Different Stages
An original sequence shows a plant before and after several days. If the question asks how the plant changed, compare the same plant or same measured group across the two stages.
If a second plant was grown under a different condition and the question asks which condition produced a greater increase, calculate or identify the change in each plant first, then compare the changes.
Do not compare Plant P’s starting height with Plant Q’s final height simply because those are the two largest-looking bars.
Worked Example 5 — A Circuit Before and After One Change
A circuit works in Diagram 1. In Diagram 2, one connection is changed and the bulb no longer lights.
The first useful comparison is within the system: what stayed the same and what changed between Diagram 1 and Diagram 2?
That comparison helps isolate the difference-making condition. It would be unhelpful to compare an unrelated component in Diagram 1 with another component in Diagram 2 simply because both are visible.
Worked Example 6 — Comparing Two Graph Lines
Two line graphs show P and Q across the same five time points.
If the question asks “Which set-up has the higher value at 15 minutes?”, compare the two lines vertically at 15 minutes.
If the question asks “Which set-up changed more from 5 to 15 minutes?”, compare each line with itself between those two times, then compare the changes.
Same graph. Different comparison job.
The Alignment Check
Before comparing two values, ask whether they match on these dimensions:
| Dimension | Question to ask |
|---|---|
| Object | Am I comparing the same object across time, or two different objects deliberately? |
| Quantity | Are both numbers measuring the same thing? |
| Unit | Are the units compatible? |
| Time/stage | Should the observations be matched at the same point? |
| Condition | What condition differs, and is that the one the question is testing? |
Comparison Is Not Always About the Largest Number
PSLE Science frequently asks about:
- greatest increase;
- greatest decrease;
- faster change;
- difference caused by a condition;
- same result under different conditions;
- change from an initial state;
- which evidence best supports an explanation.
The largest raw number may answer none of those.
The Reference Point Controls the Meaning
“Higher” needs a reference. “Changed more” needs a baseline. “Faster” needs comparable time. “More effective” needs a defined outcome.
Before using comparative language, finish the sentence:
Higher/lower/more/less compared with ______, for the quantity ______, at the time or stage ______.
Why Fair-Test Logic Still Matters
A perfectly aligned comparison can still fail to support a causal conclusion if several relevant conditions differ between the set-ups.
Comparison selection is one layer. Fair-test logic is another.
After choosing the correct pair of observations, ask whether the set-ups differ only in the intended factor or whether competing differences remain.
When the Same Set-Up Is Its Own Best Reference
Before–after designs are useful when the scientific question is about change in one system. The original state acts as a reference for the later state.
But even then, time itself can bring changes. If the question wants to establish the effect of a treatment, another set-up without that treatment may be needed as a comparison. That is why control set-ups exist.
When Another Set-Up Is the Better Reference
If the question asks whether one condition produces a different outcome from another condition, the evidence usually comes from matched set-ups that differ in the tested factor.
Comparing a set-up only with itself may show that change occurred, but not whether the tested factor was responsible.
This is the bridge from difference to causal evidence.
The Earliest-Weak-Link Diagnostic
| Failure signature | Earliest weak link | Repair |
|---|---|---|
| Chooses the highest final value when asked for greatest change | Final state was confused with change from baseline. | Compare each set-up with its own starting value first. |
| Compares P at 10 min with Q at 30 min | Observation point was not aligned. | Match the time or stage before comparing set-ups. |
| Compares temperature with mass | Quantity identity was lost. | Name the measured quantity before comparing values. |
| Uses before–after change to claim the treatment caused it | Change evidence was mistaken for controlled causal evidence. | Ask what suitable comparison set-up rules out time or other causes. |
| Uses two final values when starting values differ | Baseline difference was ignored. | Compute within-set-up changes before comparing. |
| Calculates every possible difference | Question target was not selected. | State what comparison would answer the question before calculating. |
Misconception Repair — “Same Time” Does Not Automatically Mean Fair Comparison
Matching the observation time is necessary in many investigations, but other relevant conditions must also be comparable. Do not stop checking once the clock matches.
Misconception Repair — “Before and After” Does Not Automatically Prove Cause
If something changed after a treatment, that sequence is evidence of change. To attribute the change specifically to the treatment, the design needs to deal with competing explanations.
At Primary level, use fair-comparison reasoning rather than advanced causal terminology.
Misconception Repair — The Nearest Numbers Are Not Necessarily the Right Pair
Tables often place values beside one another for convenience. Scientific comparability comes from meaning, not physical closeness on the page.
Question-Reading Protocol
- Write the target quantity. Height? Temperature? Mass? Number? Time?
- Identify the comparison word. Change, difference, greater, faster, before, after, compared with?
- Name the reference. Same set-up earlier, or another set-up now?
- Align time/stage. Are the observations comparable?
- Align conditions. Which condition is supposed to differ?
- Calculate only if needed. Use change rather than final value when the question asks for change.
- Select the concept. What scientific relationship explains the comparison?
- Check the claim. Does the evidence show difference only, or also support cause?
A Scratch-Box Method for Difficult Tables
Write four small labels beside the data:
WHO? → WHAT QUANTITY? → WHEN? → UNDER WHICH CONDITION?
If two values do not align on the dimensions needed by the question, they are probably not the correct comparison pair.
How This Appears in Multiple Choice
- Identify what is being compared.
- Reject options using mismatched times or quantities.
- Reject options using final value when change is requested.
- Check whether the comparison isolates the intended factor.
- Use the relevant concept to choose the scientifically supported option.
How This Appears in Open-Ended Answers
When comparing change, a useful reasoning shape is:
Set-up P changed from ______ to ______, while Set-up Q changed from ______ to ______ over the same ______. Therefore, ______ changed more/less. Under the stated controlled conditions, this is consistent with ______.
Use only the parts the question requires. This is not an official answer phrase.
Practice Sequence
- Take ten tables and mark every value as P-before, P-after, Q-before or Q-after.
- For each question, circle the exact pair of values needed before calculating.
- Practise questions that ask final state, amount of change and rate of change from the same data table.
- Practise matched-time versus mismatched-time comparisons.
- Practise set-ups with different starting values.
- Practise one example where a control set-up is needed to interpret a before–after change.
- Switch from tables to line graphs and diagrams.
- Return after several days with an unfamiliar mixed representation.
Unfamiliar Transfer Challenge
A mystery investigation gives these original values:
| Set-up | Start | After 20 min |
|---|---|---|
| M | 70 | 52 |
| N | 55 | 44 |
Which has the lower final value? N.
Which changed more? M, because M changed by 18 while N changed by 11.
Which condition caused the greater change? You cannot answer that merely from the numbers unless the set-ups form a valid scientific comparison and the relevant conditions are known.
Three questions. Three evidence jobs.
Delayed Independent Return
Four days later, use a fresh question and answer these before doing any calculation:
- What quantity is the question asking about?
- Is it asking for state, change or difference?
- What is the reference point?
- Do I compare one set-up across time or different set-ups at one time?
- Are the time/stage and units aligned?
- Do starting values matter?
- Does the comparison support only a difference, or a causal explanation too?
The Answer-Checking Receipt
- Did I name the measured quantity?
- Did I identify the correct reference point?
- Did I align time or stage?
- Did I avoid comparing final values when the question asks for change?
- Did I compare within each set-up first when starting values differ?
- Did I preserve fair-test logic?
- Did I separate evidence of change from evidence of cause?
- Did I stop calculating once the question was answered?
Evidence and Model Limits
Not every PSLE Science comparison requires arithmetic. Many can be made directly from diagrams, observations or graphs. The principle is the same: compare observations that answer the same scientific job.
Also, this guide does not turn every before–after question into a controlled experiment. Sometimes the task is simply to describe change. Read the question before applying causal language.
Useful Internal Routes
- How to Compare Change When Two Set-Ups Start at Different Values
- How to Compare Three or More PSLE Science Set-Ups
- How to Read More, Less, Faster and Higher by Finding the Comparison Reference
- How to Distinguish Evidence of a Difference From Evidence of a Cause
- How to Decode Variables and Fair Tests in PSLE Science Questions
- How to Use a Control Set-Up in PSLE Science
- Primary Science | Complete P1–P6 and PSLE Science Guide
Parent and Tutor Teaching Guide
Give the learner one four-cell table: P-before, P-after, Q-before, Q-after. Ask three different questions from the same table:
- What happened to P?
- Which set-up has the higher final value?
- Which set-up changed more?
Require the learner to point to the exact cells being compared before answering. This makes comparison selection visible.
If the learner chooses the wrong cells, do not immediately teach the science concept again. The earliest weak link may be evidence alignment, not content knowledge.
Then vary starting values, times and representations. Finish with a delayed return in which the learner must decide the comparison family without a prompt.
Authoritative and Research References
- Singapore Examinations and Assessment Board — PSLE Formats Examined in 2026.
- Singapore Examinations and Assessment Board — PSLE Science syllabus, for examination from 2026.
- Singapore Ministry of Education — Science Teaching and Learning Syllabus, Primary, 2023.
- Research on scientific inquiry and evidence-based comparison is used here as broader teaching evidence. It does not create a PSLE-specific marking rubric or compulsory wording.
The Quiet Ending
Comparison is a scientific decision before it is a mathematical one.
Choose the right observations. Align their meaning. Then compare.
Once the evidence pair is right, the Science has somewhere solid to stand.
