How to Reconcile Two Pieces of PSLE Science Evidence That Seem to Disagree

Wait, What? Two Correct Observations Can Look Like a Contradiction

A learner reads that an object’s temperature fell during an investigation. A second graph shows that the object was still warmer than the surrounding air. “Those cannot both be true,” the learner says. “If it cooled, it should be cold.”

But cooling means the object’s temperature decreased. It does not mean the object instantly became colder than everything around it.

This is one of the most valuable PSLE Science distinctions: evidence may appear to disagree because the learner has compared different objects, different quantities, different times, different conditions or different levels of explanation.

Do not choose your favourite piece of evidence. First ask whether the two pieces are genuinely answering the same scientific question.

Quick Answer

When two pieces of PSLE Science evidence seem to disagree, place them side by side and align them. Check whether they refer to the same object, measured quantity, unit, time, condition and method. Keep observation separate from inference. Then decide whether the conflict is:

only apparent because the evidence describes different things;
resolvable once scale, timing or method is considered;
a real challenge to the proposed explanation; or
still undecidable because more evidence is needed.

Use this route:

STATE EVIDENCE A → STATE EVIDENCE B → ALIGN OBJECT, QUANTITY, TIME, CONDITION AND METHOD → SEPARATE OBSERVATION FROM INFERENCE → SELECT THE RELEVANT CONCEPT → EXPLAIN WHETHER BOTH CAN FIT ONE MECHANISM → REVISE THE CLAIM IF NEEDED → STATE WHAT THE EVIDENCE CAN AND CANNOT SUPPORT.

The Exact PSLE Science Learning Job This Guide Owns

This guide owns one learner job: how a Primary 5 or Primary 6 learner compares two pieces of scientific information that appear inconsistent, determines whether the conflict is real or only apparent, and revises the explanation when the evidence requires it.

It does not replace the individual skill of reading a table, graph or diagram. It does not replace the science concept involved. It also does not teach a generic debate method. Its job is narrower and more practical: preserve both observations long enough to understand what each one actually says.

This is not a licence to argue against every question. Most school evidence can be reconciled by careful reading. The skill matters because forcing data to agree—or discarding an inconvenient result—breaks scientific reasoning.

Why This Matters in the Current PSLE Science Frame

For examination from 2026, the revised PSLE Science paper assesses the 2023 Primary Science syllabus. The official assessment objectives include interpreting and analysing information, evaluating observations, information and methods, applying scientific facts, concepts and principles, and communicating explanations and reasoning.

Those jobs include handling evidence that is incomplete, indirect, measured at different stages or affected by method. A learner must be able to say not only what a result shows, but how it fits—or fails to fit—with the rest of the information.

Why Evidence Appears to Conflict

Apparent disagreement usually begins at one of seven boundaries.

Boundary	How the apparent conflict appears	Question to ask
Object	Evidence A is about Setup P; Evidence B is about Setup Q	Am I tracking the same scientific object?
Quantity	One source measures temperature; another describes heat transfer	Are these the same quantity or related but different ideas?
Time	One observation is from the start; another is from later	Can both be true at different times?
Condition	Measurements were made under different surroundings or treatments	Did the relevant condition change?
Scale or unit	Values look different because the axes or units differ	Have I placed them on a comparable scale?
Method	Two indicators respond differently or have different sensitivity	What did each method actually detect?
Reasoning level	An observation is compared with an explanation as though both were observations	Which statement is measured, and which is inferred?

The First Discipline: Preserve Both Pieces of Evidence

When learners feel a contradiction, they often do one of three unsafe things:

ignore the less familiar source;
change one observation so that it matches the expected answer; or
average two values even when they measure different things.

Instead, write each piece exactly as given. Do not add a cause yet.

Evidence A says ______ under ______.
Evidence B says ______ under ______.

The small phrase “under…” often reveals why both can be true.

The Eight-Check Reconciliation Method

1. Same Object?

Track the labels. A learner may accidentally pair Plant A’s condition with Plant B’s result, or compare the temperature of the water with the temperature of the surrounding air.

2. Same Quantity?

Temperature, heat, mass, water level, distance, speed, time, brightness and number of bubbles are not interchangeable. Two measurements can be related without being the same quantity.

3. Same Unit and Scale?

Convert units where necessary. Read graph axes carefully. A steep-looking line on a compressed scale may represent a smaller change than a gentler-looking line on another scale.

4. Same Time or Duration?

A value after five minutes should not be treated as simultaneous with a value after twenty minutes. A process can change direction or slow down over time.

5. Same Condition?

Check light, temperature, amount, surface area, material, position, source strength, distance and other relevant variables. Evidence obtained under different conditions may describe different responses rather than a contradiction.

6. Same Method?

A bulb, thermometer, balance, ruler and colour indicator reveal different features with different limits. One method may detect a change that another method cannot show clearly.

7. Observation or Inference?

“The bulb did not light” is an observation. “No electric current flowed” is an inference that depends on the circuit and detector. Do not call an inference conflicting evidence until you have separated it from the observation underneath.

8. Can One Mechanism Explain Both?

After alignment, select the relevant concept. Ask whether the process naturally produces both observations at different stages, places or scales. If not, the explanation or method may need revision.

Four Possible Outcomes After Reconciliation

Outcome	What it means	Appropriate response
Apparent conflict	The evidence refers to different objects, quantities, times or conditions	Clarify the distinction and keep both observations
Method-resolved conflict	Different tools, scales or indicator limits explain the difference	State what each method can detect and compare cautiously
Real challenge to explanation	The evidence is aligned and the proposed mechanism cannot explain both	Revise or reject the explanation
Insufficient evidence	More than one account remains possible	State the limit and identify what additional observation would help

Worked Example 1 — Cooling but Still Warmer

Original practice situation: Water begins at 70°C in a room at 28°C. Ten minutes later, it is 50°C.

Evidence A: the water cooled.

Evidence B: the water remained warmer than the room.

These statements do not conflict. “Cooled” describes a decrease from 70°C to 50°C. “Warmer than the room” compares 50°C with 28°C. One statement tracks change over time; the other compares two objects at one time.

The relevant mechanism is that heat is transferred from the warmer water to cooler surroundings. The water can lose heat and decrease in temperature while still remaining warmer than its surroundings.

Earliest weak link: the learner treated “cooling” as a final temperature category instead of a direction of change.

Worked Example 2 — A Taller Plant With Fewer Leaves

Original practice situation: After the same duration, Plant P is taller than Plant Q, but P has fewer leaves.

A learner says the observations disagree because the “better-growing” plant should be larger in every way.

Height and number of leaves are different measurements. A plant can have a longer stem and fewer leaves. Neither measure alone captures every aspect of growth or health.

The correct response is not to decide which observation is wrong. It is to state what each one shows and use the question’s conditions to determine which measurement is relevant to the claim being tested.

Worked Example 3 — Water Level Falls but the Mass Comparison Looks Small

Two differently shaped containers hold water. In the narrow container, the water level falls by 2 cm. In the wide container, it falls by 1 cm. A learner concludes that twice as much water was lost from the narrow container.

The level change and amount of water lost are related through container shape. A 2 cm fall in a narrow container may represent less volume than a 1 cm fall in a much wider container.

If a balance shows similar mass loss, the evidence need not conflict. Water level is not a direct universal substitute for amount. The reconciliation requires the cross-sectional shape or a direct mass/volume measurement.

Worked Example 4 — Bubble Count and Gas Production

An investigation counts visible bubbles released from a water plant. In one interval, fewer but larger bubbles are seen; in another, more but smaller bubbles are seen.

Bubble number alone and total gas volume are not identical measurements. Two intervals can differ in count without differing in total gas produced in the same direction.

This is not a reason to declare classroom bubble counting useless. It is a reminder that bubble count is an indirect indicator with limits. If the question asks for a comparison based on the supplied count, answer within that measure; do not silently turn every bubble into an equal gas volume.

Worked Example 5 — A Bigger Shadow From the Same Object

A learner sees Evidence A: the object is the same size in both setups. Evidence B: the shadow in Setup B is larger.

There is no necessary conflict. Object size is not the only factor affecting shadow size. The positions of the light source, object and screen matter. The same object can produce shadows of different sizes when the spatial arrangement changes.

The diagram must be read as a relationship among parts, not as a portrait of the object alone.

Worked Example 6 — A Bulb Lights Once but Not in a Repeat

A simple circuit test gives a lit bulb in Trial 1 and an unlit bulb in Trial 2 using the same labelled material.

Do not average “lit” and “not lit”. The results are categorical and inconsistent. Check whether the contact points, cell, bulb and wires were the same and working. Repeat with a known conductor to verify the circuit, then retest the material using consistent contact.

If the setup is reliable and the result remains inconsistent, the evidence does not support a confident simple conclusion. The correct scientific move is to report the variation and improve the method rather than force agreement.

Repeated Measurements: Variation Is Not Automatically a Contradiction

Repeated measurements often differ slightly. A ruler may be read from a slightly different angle. A moving object may not stop at exactly the same point. Living things vary. Surrounding conditions can change.

Small variation may be expected. Large or patterned disagreement deserves investigation.

Pattern across repeats	Possible interpretation	Next action
Values are close with no obvious pattern	Ordinary measurement or natural variation may be present	Use an appropriate summary only if the quantities and method are comparable
One value is far from the others	Possible reading, setup or recording issue	Check the trial; do not delete it without a reason
Values drift steadily	A condition may be changing over time	Check temperature, source strength, material state or timing
Results switch category	Contact, threshold or apparatus sensitivity may matter	Verify the detector and standardise the procedure

Do not use “take the average” as a reflex. Averaging is sensible only for comparable numerical measurements. It cannot repair mixed units, different variables, a broken method or contradictory categories.

Evidence Against a Claim Is Not the Same as Missing Evidence

Suppose a bulb does not light. That may be evidence against the claim that the tested setup produces enough current for visible light, provided the circuit and bulb are working. But it may not prove that absolutely no current exists if the indicator cannot reveal very small current.

Likewise, failing to observe bubbles during a short interval may reflect low activity, dissolved gas, unsuitable observation time or detector limits. “No visible effect” must be interpreted through the method.

When One Piece of Evidence Is Stronger

Reconciliation does not require pretending that all evidence is equally informative. A calibrated measurement under controlled conditions may carry more weight for a specific question than an unaided visual impression. A repeated pattern may be more reliable than one isolated observation.

However, “stronger” must be explained. Ask:

Is the evidence more directly related to the claim?
Was the measurement made under better-controlled conditions?
Is it more precise or sensitive for the required quantity?
Was it repeated consistently?
Does it rule out an alternative explanation?

How to Reconcile Evidence in Inquiry Questions

When results do not fit neatly, evaluate the investigation before rewriting the science.

Check the changed variable and measured variable.
Check whether other relevant variables were kept comparable.
Check the starting conditions.
Check the measuring instrument, unit and reading method.
Check whether observations were taken at the same times.
Check whether the indicator directly measures the target or acts as a proxy.
Repeat using a consistent method where appropriate.
Decide whether the explanation, the method or both need revision.

An unexpected result is not automatically a mistake. It may reveal a hidden variable, a method limit or an incomplete explanation.

How to Write a Reconciliation Answer

When the question asks you to explain apparently different observations, a useful reasoning shape is:

Evidence A refers to ______, while Evidence B refers to ______. They do not necessarily conflict because ______. Under the stated condition, the scientific mechanism allows ______, so the supported conclusion is ______.

This is not a universal phrase to memorise. Some questions need only one sentence; others require method evaluation. The structure earns its place only when there is a real distinction to clarify.

Multiple-Choice Questions: Do Not Let One Option Erase Evidence

An attractive option may explain Evidence A but ignore Evidence B. Another option may show how both can fit one mechanism.

Translate each option into a claim.
Check it against both pieces of evidence.
Ask whether it changes the object, time, quantity or condition.
Reject options that require one observation to be false when both can coexist.
Prefer the explanation that accounts for the full evidence without adding unsupported assumptions.

The Earliest-Weak-Link Diagnostic

Observable failure signature	Earliest weak link	Repair path
“One result must be wrong because the words sound opposite.”	The measured quantities were not defined.	State exactly what each observation measures.
“I use whichever source matches the concept I remember.”	Evidence selection replaced evidence evaluation.	Preserve both sources and align them before selecting the concept.
“I average the values to make them agree.”	Comparability was not checked.	Average only comparable numerical repeats; first check units, variables and method.
“The graph is correct because it looks scientific.”	Representation authority replaced reasoning.	Read axes, data origin, scale and conditions.
“An unexpected result must be a careless mistake.”	The preferred explanation was protected from evidence.	Check method and alternative mechanisms before judging the observation.
“I describe both results but never reconcile them.”	The mechanism step is missing.	Ask whether one causal process can produce both under different conditions or times.
“I cannot decide, so I invent a hidden detail.”	Evidence limits are not accepted.	State what remains uncertain and what additional evidence would discriminate.

Misconception Repair: “Contradictory Evidence Means Science Has Failed”

No. Disagreement can reveal that measurements concern different quantities, that a method needs improvement or that an explanation is incomplete. Scientific knowledge becomes stronger when it can be corrected by evidence.

Misconception Repair: “There Must Always Be One Correct Measurement”

Different methods may measure different aspects of a system. Height and leaf number can both be correct. Temperature and heat transfer are related but not identical. Bubble count and gas volume may move differently. The question is not which number feels more scientific; it is which measurement answers the claim.

Misconception Repair: “The Answer Key’s Explanation Makes Every Other Observation Irrelevant”

A model explanation should still respect the evidence. During practice, use answer keys to check the intended concept, but retain the habit of asking what each observation supports and whether any claim goes beyond the information.

Misconception Repair: “If Both Can Be True, Both Causes Are Proven”

No. Showing that two observations can coexist removes a false contradiction. It does not automatically prove a particular cause. Causal claims still require relevant conditions, fair comparison and a mechanism supported by the question.

A Practice Sequence for Reconciliation

Language contrasts: compare “cooled” with “cool”, “slower” with “small”, and “less” with “decreasing”.
Quantity contrasts: separate temperature from heat, level from amount, bubble count from volume, and height from overall growth.
Time alignment: practise pairing values from the same time point before explaining change.
Condition alignment: compare observations only after listing relevant setup differences.
Method limits: identify what a bulb, indicator, thermometer or ruler can and cannot reveal.
Mechanism reconciliation: explain how one process can produce two apparently different observations.
Revision: rewrite an explanation after adding a challenging result.
Delayed transfer: repeat days later in a different syllabus theme.

Unfamiliar Transfer Challenge

A sealed transparent chamber contains a strip that changes colour when a certain condition is reached. A sensor also records a steadily changing numerical value. During the first five minutes, the number changes but the strip does not change colour. Later, the strip changes suddenly.

Do these observations conflict? Not necessarily. The strip may respond only after a threshold is crossed, while the sensor shows smaller changes throughout. One method provides a continuous measurement; the other gives a category after a limit is reached.

You do not need to know the exact device. The transferable reasoning is to compare method sensitivity, scale and timing before declaring a contradiction.

Delayed Independent Return

Four days after studying this guide, choose a fresh question with two results or representations. Without notes, complete:

Evidence A, stated without explanation;
Evidence B, stated without explanation;
the object each refers to;
the quantity and unit each uses;
the time and condition for each;
whether either statement is an inference rather than a direct observation;
one mechanism that may explain both;
whether the conflict is apparent, method-based, real or undecidable;
the final bounded conclusion.

If you solve the familiar example but cannot align a new one, return to the first failed check rather than memorising the old conclusion.

The Answer-Checking Receipt

Did I state both pieces of evidence accurately?
Do they refer to the same object?
Do they measure the same quantity?
Are the units and scales comparable?
Were they observed at the same time or over the same duration?
Were the conditions the same?
Did the methods have different sensitivity or limits?
Did I separate observation from inference?
Can the relevant scientific mechanism explain both?
Does either piece genuinely challenge the explanation?
Have I revised the claim rather than forced the data?
Did I state what remains unknown?

Evidence and Model Limits

Not every evidence conflict can be settled from the information supplied. A small number of measurements may be insufficient. A diagram may not be to scale. An indirect indicator may miss small changes. Living systems may vary. A simplified Primary Science model may explain the dominant relationship without describing every real-world influence.

Do not manufacture certainty. A scientifically mature Primary learner can say, in age-appropriate language, that the evidence supports one conclusion under the tested conditions while another claim cannot yet be decided.

Useful Internal Routes

Parent and Tutor Teaching Guide

When a child says, “The question contradicts itself,” do not resolve it immediately. Ask the child to rewrite both observations without the words because, therefore or means. This removes premature inference.

Then ask in order:

“Are these about the same object?”
“Are they measuring the same thing?”
“Were they measured at the same time and condition?”
“Could one process produce both?”
“Which part, if any, challenges your original explanation?”

The child’s first failure identifies the teaching job. Mixing quantities needs vocabulary and concept repair. Mixing time points needs representation control. Protecting a preferred answer from an unexpected result needs evidence discipline. A broken setup needs inquiry-method repair.

Use near-miss pairs: one true conflict and one apparent conflict. Ask the learner to distinguish them and justify the decision. Later, return with a new topic and without announcing that it is an “evidence conflict” question.

Authoritative and Research References

The Quiet Ending

One observation may describe change.

Another may describe position, amount, condition or comparison.

They may challenge each other. They may also be looking at the same system from different windows.

Science does not ask you to make the windows agree. It asks you to identify what each one shows, test whether a single explanation can respect them both, and revise your conclusion when it cannot.

Small Group Tutorials