How to Combine Evidence From Text, Diagrams and Data in One PSLE Science Answer

Wait, What? More Evidence Can Produce a Worse Answer

A PSLE Science question may give you a paragraph, a labelled diagram and a table of results. Many learners react by copying one fact from each source into three separate sentences.

The result can be long, accurate in parts—and scientifically unfinished.

The problem is not that the learner missed the evidence. The problem is that the evidence never met inside one explanation.

A strong answer does not collect clues. It gives each clue a distinct reasoning job, connects the clues through the relevant scientific mechanism, and uses the combined evidence to answer the exact question.

Text may establish the starting conditions. A diagram may reveal how the parts are arranged. A table may show the measured outcome. None of those sources has to carry the whole answer alone. The scientific work is to make them cooperate.

Quick Answer

To combine evidence from text, diagrams and data in one PSLE Science answer, first identify what each source contributes. Then align the sources so that they refer to the same object, condition and time. Select the relevant science concept, explain the mechanism once, use each piece of evidence where it strengthens a different part of the reasoning, and check that no source has been ignored, repeated or contradicted.

Use this route:

READ THE QUESTION JOB → LABEL EACH INFORMATION SOURCE → EXTRACT ITS RELEVANT CONTRIBUTION → ALIGN OBJECT, CONDITION AND TIME → SELECT THE SCIENTIFIC CONCEPT → EXPLAIN THE CAUSAL MECHANISM → CONNECT THE EVIDENCE TO THE CONDITION → STATE THE OUTCOME → CHECK EVERY SOURCE AND LIMIT.

The Exact PSLE Science Learning Job This Guide Owns

This guide owns one learner job: how a Primary 5 or Primary 6 learner synthesises information supplied in different forms—words, labels, diagrams, tables, graphs or observations—into one coherent PSLE Science answer.

It does not replace the separate skill of reading a graph, decoding a diagram or identifying an observation. Those pages remain the owners of those individual jobs. This guide owns what happens next: several evidence sources must be made to support one scientific explanation without double-counting, contradiction or loose copying.

It also does not teach a compulsory answer template. The structure must follow the scientific task. Some questions need two evidence sources; others need three. Some sources supply conditions rather than conclusions. The goal is not to mention everything. The goal is to use everything that is relevant.

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 applying scientific facts, concepts and principles; interpreting and analysing information; evaluating observations, information and methods; and communicating explanations and reasoning.

Information may be presented in more than one representation because scientific reasoning rarely lives in one sentence. The learner may have to use a condition from the text, a relationship from a diagram and a pattern from data before an explanation becomes defensible.

Why Questions Use Several Representations

Different representations are good at showing different things.

Information source	What it often shows well	What it may not show by itself
Written text	Conditions, instructions, timing, what was kept the same, what changed, what is being asked	Spatial arrangement, exact pattern across many values, hidden relationships between parts
Labelled diagram	Parts, positions, connections, direction, sequence, relative arrangement	Whether the drawing is to scale, how strongly something changed, whether a difference is reliable
Table	Exact recorded values, categories, repeated measurements, comparisons	The mechanism that produced the values
Graph	Patterns, changes over time, trends, peaks, plateaus and comparisons	The exact cause of the pattern unless conditions and science support it
Observation statement	What was seen, measured or detected	Why it happened

A learner who expects one source to do every job will either overclaim from that source or ignore the rest. A learner who understands representation roles can build an answer with fewer words and stronger connections.

The Evidence-Job Principle

Before writing, ask what job each piece of evidence performs. Common jobs include:

Identity: names the object, organism, material, part or variable being discussed.
Starting condition: shows what was the same before the change.
Changed condition: reveals the factor that differs between setups.
Arrangement: shows how parts are connected, positioned or ordered.
Measurement: gives a value, count, time, temperature, mass, distance or category.
Comparison: establishes more, less, faster, slower, higher, lower, present or absent.
Pattern: shows a trend or sequence across several data points.
Boundary: shows what happened only under particular conditions or within a measured range.
Mechanism clue: helps select the scientific relationship that explains the outcome.

Two sources can support the same conclusion without doing the same job. That is synthesis. Copying the same comparison from the graph and table in different words is repetition, not synthesis.

The Three-Layer Structure of a Synthesised Answer

Layer 1 — Case Evidence

What do the text, diagram and data establish about this particular situation?

Layer 2 — Scientific Mechanism

Which fact, concept, principle or causal relationship explains how the stated condition can produce the measured outcome?

Layer 3 — Question-Specific Conclusion

What exact outcome follows, and how directly does it answer the command word?

Evidence tells you what happened here. Science explains why that relationship is possible. The conclusion returns both to the question.

A Practical Seven-Step Synthesis Protocol

Step 1 — Read the Question Before the Sources

Decide whether you must compare, explain, predict, evaluate, identify, conclude or suggest an improvement. The same table can support different answers depending on the learner job.

Step 2 — Mark the Sources

Use small labels such as T for text, D for diagram and R for results. This prevents the loudest source—often the graph—from taking over before you have read the conditions.

Step 3 — Extract One Relevant Contribution From Each Source

Write a short note, not a sentence for the final answer. For example: “T: same amount and start temperature”, “D: one container covered”, “R: less mass remained in uncovered container”.

Step 4 — Align the Scientific Object

Check that all notes refer to the same container, plant, circuit, interval or measured variable. Learners often combine the condition of Setup A with the result of Setup B.

Step 5 — Select the Mechanism

Name the concept only after the evidence is organised. Ask what changed, what process responds to that change, and what outcome the process should produce.

Step 6 — Build One Causal Chain

Do not write a separate explanation for every source. Join them:

Because the stated condition differed in this way, the relevant process occurred differently; this produced the measured pattern, so the required outcome follows.

This is a reasoning shape, not a compulsory phrase.

Step 7 — Audit the Evidence

Did I use every relevant source?
Did two sentences merely repeat the same evidence?
Did I confuse a measurement with the process it indicates?
Did any source contradict my explanation?
Did I answer beyond the observed conditions?

The Alignment Check: Object, Condition, Time and Quantity

Evidence cannot be combined merely because it appears on the same page. It must be aligned.

Alignment question	Common error	Repair
Same object?	Using Plant A’s light condition with Plant B’s growth result	Name the object beside every note
Same condition?	Comparing a covered container at one temperature with an uncovered container at another	Check all relevant controlled variables
Same time?	Comparing a 5-minute value with a 20-minute value as though they were simultaneous	Align time points or compare rates carefully
Same measured quantity?	Combining water level and mass as though they are interchangeable	State what each measurement actually represents
Same unit or scale?	Treating centimetres and millimetres as direct numerical equals	Convert or compare on a common basis
Same direction?	Assuming a smaller time always means a slower process	Explain what an increase or decrease means in context

Worked Example 1 — Text, Diagram and Mass Data in an Evaporation Investigation

Original practice situation: The text states that two identical shallow containers hold equal masses of water at the same starting temperature. A diagram shows Container P uncovered and Container Q covered with a sheet containing one small opening. A table shows the mass of water remaining after the same duration.

Suppose less water remains in P.

Text contribution: the starting amount, temperature, container and duration are comparable.

Diagram contribution: a larger water surface in P is directly exposed to the surrounding air.

Table contribution: P has a greater decrease in water mass during the stated interval.

Scientific mechanism: water at the exposed surface can change into water vapour and leave the liquid. With more of the surface exposed under the stated comparable conditions, more water can evaporate during the interval.

Synthesised answer: “Both containers began with the same mass of water and were observed for the same duration, but P had a larger exposed water surface. More water evaporated from P into the surrounding air, so its water mass decreased more and less water remained.”

The answer does not say “the table proves evaporation” by itself. It combines the controlled conditions, arrangement and measured outcome with the relevant process.

Worked Example 2 — Circuit Arrangement and Results

Original practice situation: The text states that identical fresh cells and identical bulbs are used. A diagram shows two circuits. In Circuit X, the components form one unbroken path. In Circuit Y, one connection is open. A results table records that the bulb lights only in X.

Text contribution: differences in bulb and cell condition are reduced.

Diagram contribution: X has a complete path while Y has a gap.

Result contribution: visible light is observed only in the complete setup.

Mechanism: the simple circuit must provide a complete path through which electric current can flow for the bulb to produce light.

Synthesised answer: “Since identical working components were used, the important difference is the open connection in Y. X forms a complete circuit, allowing electric current to flow through the bulb, so it lights; Y has a gap, so the path is incomplete and its bulb does not light.”

Notice that “the bulb lights” is evidence about the circuit’s output. It is not a direct measurement of every electrical quantity.

Worked Example 3 — A Plant, a Diagram and a Growth Graph

Original practice situation: The text states that two similar young plants receive equal amounts of water and are kept for the same number of days. A diagram shows Plant A beside a bright window and Plant B inside a box with small air openings. A graph shows the increase in height of each plant.

The graph alone shows a difference in height change. It does not, by itself, explain the difference. The text supplies comparable conditions. The diagram supplies the light condition. The science concept supplies the relationship between light availability and the plant process relevant to making food.

A careful answer connects only what the evidence supports. It should not claim that every difference in growth is always caused by light, nor should it assume that height alone measures all aspects of plant health.

Synthesised answer: “The plants received the same amount of water for the same duration, but A received more light. Light is needed for the plant to make food, which supports growth. The graph shows that A increased more in height under these conditions.”

The final clause returns to the measured evidence. The answer does not replace the graph with a memorised plant paragraph.

Worked Example 4 — A Force Setup and Distance Data

Original practice situation: The text states that the same toy car starts from the same line on the same surface. A diagram shows an elastic band pulled back by different distances before release. A table records the distance travelled by the car.

Text contribution: the car, surface and starting position are comparable.

Diagram contribution: the elastic band is stretched by different amounts.

Table contribution: greater pull-back distance is associated with a greater travel distance over the tested range.

Mechanism: stretching the elastic band more stores more potential energy in the band; when released, more energy can be transferred to the car’s motion, leading it to travel farther under the comparable test conditions.

The learner must still respect the range. The data support the relationship for the tested pull-back distances, not for every possible stretch, surface or car.

Worked Example 5 — Shadow Size Across a Diagram and Table

Original practice situation: A diagram shows a lamp, an opaque object and a screen. The text states that the object and screen remain fixed while the lamp is moved to several positions. A table records shadow width.

A weak answer says, “The table shows the shadow gets bigger.”

A stronger answer identifies the changed spatial relationship, explains that the opaque object blocks light travelling from the source, and uses the table to show how the resulting shadow width changes with lamp position.

The diagram supplies geometry; the table supplies measurement; the light concept supplies mechanism.

Do Not Double-Count the Same Evidence

Suppose a table gives values of 12 cm and 18 cm, while a graph plots the same values. Writing “B is 18 cm” and “the graph also shows B is higher” may be one piece of evidence shown twice, not two independent supports.

Ask whether the second source adds a new job:

Does it reveal a trend not obvious from one pair of values?
Does it show arrangement or direction?
Does it establish a controlled condition?
Does it confirm the same result using a genuinely different observation?

Repeated presentation can improve readability, but it does not automatically strengthen the scientific conclusion.

Corroboration Is Different From Repetition

Corroboration occurs when different observations support the same explanation through distinct routes. For example, a mass decrease and a collected condensate may both support movement of water from one place to another, but they are not the same measurement.

At Primary level, the learner does not need formal evidence theory. Keep the distinction:

Same fact shown twice = repetition. Different facts that fit the same mechanism = possible corroboration.

When the Sources Seem to Disagree

Do not force disagreeing evidence into a smooth paragraph. Stop and check:

Are the sources about the same object?
Are the measurements from the same time?
Are they measuring the same quantity?
Does one source describe a starting condition while another shows a later outcome?
Could the diagram be schematic rather than drawn to scale?
Could the method or instrument have a limit?
Does the explanation need revision?

A graph showing a decrease in water level and a table showing little change in mass may not be a true contradiction if the containers have different shapes, the scales differ or the observations refer to different intervals. Reconciliation begins with alignment, not with choosing the source you prefer.

How Much Evidence Belongs in the Final Answer?

Use enough evidence to make the mechanism and conclusion clear. Do not narrate the whole page.

Question job	Evidence usually needed	Common excess
Explain a difference	Relevant differing condition + mechanism + outcome	Copying every controlled variable
Conclude from data	Relevant comparison or pattern + bounded conclusion	Explaining a cause the data do not establish
Predict	Changed condition + mechanism + expected outcome	Repeating all previous data values
Evaluate an investigation	Method feature + effect on evidence quality + repair	Writing generic “repeat for accuracy” without identifying what repeatability changes
Support a claim	One or more distinct pieces of evidence connected to the claim	Listing evidence without explaining relevance

How to Write the Answer Without Sounding Mechanical

Good scientific prose can be simple. It often moves in this order:

Name the relevant condition or comparison.
Explain the scientific mechanism.
Connect the mechanism to the observed result.
State the outcome requested.

Useful connectors include because, therefore, so, under these conditions, compared with and as a result. They help only when the reasoning on both sides is correct.

Do not decorate an answer with “hence”, “thus” or scientific vocabulary to hide a missing bridge. The causal connection must be real.

Multiple-Choice Questions: Use Every Source Before Eliminating

An option may match the graph but contradict the diagram. Another may use the correct concept but ignore a condition stated in the text.

Read the option as a complete claim.
Check it against the text.
Check it against the diagram.
Check it against the data.
Check whether its scientific mechanism fits all three.
Reject it if any load-bearing source contradicts it.

An option does not become correct because one source supports part of it.

Inquiry Questions: Sources Can Describe Different Parts of the Investigation

In an investigation question, text may describe the procedure, a diagram may show the apparatus and a table may give the observations. Treat them as one experiment.

Use the procedure to identify what was changed, measured and kept comparable.
Use the diagram to check whether the intended arrangement is actually possible.
Use the results to decide what relationship the evidence supports.
Use scientific knowledge to explain the relationship.
Use method limits to decide how far the conclusion can go.

Do not invent a variable because it is common in another experiment. Variable roles come from the investigation shown.

The Earliest-Weak-Link Diagnostic

Observable failure signature	Earliest weak link	Repair path
“I copy one fact from every source but the answer has no explanation.”	Evidence extraction happened without mechanism selection.	Ask what one scientific relationship connects the facts.
“My answer uses the graph and ignores the written conditions.”	The loudest representation captured attention.	Label each source before interpreting any one of them.
“I combine Setup A’s condition with Setup B’s result.”	The scientific object was not kept clear.	Write the object label beside every extracted fact.
“I mention the same comparison three times.”	Repetition was mistaken for evidence strength.	Assign a distinct reasoning job to each sentence.
“The data and diagram seem inconsistent, so I choose the one I understand.”	Alignment was skipped.	Check object, time, quantity, scale and condition before deciding there is a conflict.
“My final paragraph is much longer than the question needs.”	Selection failed after extraction.	Keep only evidence that carries the requested conclusion.
“I state a trend as a cause.”	Pattern recognition was confused with causal explanation.	Use the method and relevant science to justify cause; otherwise state only the observed relationship.

Misconception Repair: “Every Picture Detail Is Evidence”

No. Some diagram features are labels or visual aids. Unless the question indicates otherwise, a schematic diagram may not be drawn to scale. Colour, shape and spacing should not be treated as measurements merely because they are visible.

Misconception Repair: “The Table Tells Me Why”

A table records values or categories. It can show a comparison or pattern. The cause requires conditions, method and scientific mechanism. Data do not explain themselves.

Misconception Repair: “Using More Evidence Always Makes the Answer Better”

Irrelevant evidence makes an answer harder to follow. Evidence earns its place by supporting the exact claim. A concise answer using two well-connected facts can be stronger than a page of copied detail.

Misconception Repair: “I Must Use a Fixed Number of Points”

No universal number applies to every question. Use the question’s demand, the evidence and the scientific mechanism. Do not invent a marking rule from one school worksheet or one model answer.

A Retrieval and Practice Sequence

Single-source fluency: practise extracting one useful fact from text, one from a diagram and one from a table.
Evidence-job labelling: name each fact’s job—condition, arrangement, measurement, comparison or pattern.
Two-source synthesis: combine text and diagram, then text and data, then diagram and data.
Three-source synthesis: build one explanation using all three without repetition.
Contradiction check: include one practice case where a source limits or challenges the preferred explanation.
Representation switch: convert a table into a verbal comparison, or a diagram into a relationship statement, without adding information.
Unfamiliar transfer: repeat in a different Primary Science theme.
Delayed return: solve a fresh multi-source question several days later without the checklist.

After each attempt, compare the reasoning jobs rather than memorising the model wording. Ask: Which source established the condition? Which showed the outcome? Which sentence carried the mechanism?

Unfamiliar Transfer Challenge

A sealed container holds a movable coloured marker. The text says the same container is placed first in cool water and later in warm water for equal durations. A diagram shows the marker connected to a flexible membrane. A graph shows the marker position rising during the warm-water interval and falling during the cool-water interval.

You are not expected to know a named device. Work from the representations:

The text supplies the changed surrounding temperature and equal duration.
The diagram shows which parts can respond and move.
The graph supplies the direction of marker change over time.
The scientific task is to propose a mechanism consistent with the relevant Primary Science ideas and to state only what the evidence supports.

A weak answer merely says “warm makes it go up”. A stronger answer identifies what in the system changes, how that change affects the membrane and why the marker moves. It also avoids claiming a precise quantity that was not measured.

Delayed Independent Return

Four days after studying this guide, take a fresh question containing at least two representations. Without notes, produce this receipt:

Question job: what must I do?
Text contribution: one relevant condition or fact.
Diagram contribution: one relevant arrangement or relationship.
Data contribution: one measured comparison or pattern.
Scientific mechanism: how the condition leads to the outcome.
Final conclusion: the exact answer under the stated conditions.
Limit: one claim the evidence does not support.

If the final answer cannot be reconstructed without the labels, return to the weakest step. Do not simply reread the completed paragraph.

The Answer-Checking Receipt

Did I answer the command word?
Did I identify every relevant information source?
What distinct job does each source perform?
Do all pieces refer to the same object, condition and time?
Did I distinguish observation from inference?
Did I select the relevant concept rather than the most familiar topic?
Did I explain the mechanism once, clearly?
Did I connect the mechanism to the question’s condition?
Did I state the required outcome?
Did I double-count the same evidence?
Did I ignore any contradiction or method limit?
Does every sentence earn its place?

Evidence and Model Limits

Combining sources cannot make weak evidence strong by itself. Three representations may all originate from the same measurement. A graph may smooth or compress details. A diagram may be schematic. A table may contain too few observations to establish a trend. Written conditions may omit a possible confounding variable.

The learner should use the model appropriate to the Primary Science question while recognising what the sources can and cannot show. “The evidence supports” is different from “the evidence proves in every situation”.

Useful Internal Routes

Parent and Tutor Teaching Guide

When a learner gives a fragmented answer, do not begin by demanding “more points”. Ask the learner to place a finger on each source and say what job it performs.

“What does the text establish that the diagram does not?”
“What does the diagram reveal that the table does not?”
“What measured outcome does the data add?”
“Which one scientific relationship connects them?”
“Which sentence in your answer carries that connection?”

If the child extracts evidence well but cannot connect it, teach the mechanism. If the child knows the mechanism but mixes objects or time points, teach alignment. If the child writes everything, teach selection. These are different weak links.

A useful exercise is to cover one source and ask how the answer becomes weaker. If removing the diagram changes nothing, perhaps the learner never used its arrangement information. If removing the text changes nothing, the controlled conditions may have been ignored. If removing the data changes nothing, the answer may be a memorised science paragraph rather than a response to the evidence.

Return after a delay with a different combination of representations. Independence means the learner can assign evidence jobs in a new context, not merely reproduce one paragraph.

Authoritative and Research References

The Quiet Ending

The text tells you the conditions.

The diagram lets you see the arrangement.

The data show what changed.

Your scientific knowledge explains why those pieces belong together.

A complete answer is not a pile of information. It is a carefully built relationship in which every piece of evidence carries weight—and none is asked to carry more than it can.

Small Group Tutorials