Science Education Systems · Article 10. Maya, Jia Jun, Hana and Ethan remain fictional Punggol learners. Their different explanation habits show why knowing Science and communicating Science are connected but not identical capabilities.
The 50-second parent route
A Science explanation is not a pile of keywords.
It is a visible chain of reasoning.
The learner must identify what the question is asking, retrieve the relevant scientific relationship, use evidence from the situation and communicate the mechanism clearly enough that another person can follow it.
The route is:
command → target → condition → concept → evidence → relationship → causal chain → precise language → check → transfer
The common mistake is to begin with vocabulary.
“Photosynthesis.”
“Waterproof.”
“Friction.”
“Conductor.”
“Energy.”
The word may be correct and the explanation may still be incomplete.
A stronger question is:
What relationship must this word help me explain?
This article extends How Science Assessment Works, How Scientific Thinking Is Built and How Science Knowledge Networks Work.
1. Understanding can exist without reaching the page
Jia Jun looks at the question.
“I know this.”
He does.
He can explain it aloud when the tutor asks follow-up questions.
On paper he writes one word.
The examiner cannot mark the unexpressed chain in his head.
This is the central problem of scientific explanation:
knowledge has to become observable in language.
2. Scientific writing is part of scientific thinking
Writing does not simply report thought after the thinking is finished.
The act of constructing a precise explanation often reveals whether the relationship is truly understood.
Where the sentence breaks, the model may also be weak.
If a student cannot explain what causes the change, perhaps the causal edge is missing.
If the student cannot identify what “it” refers to, perhaps the system has not been parsed clearly.
If the learner uses four keywords but no verb, perhaps the relationship remains unstated.
3. Start with the command
State.
Describe.
Explain.
Compare.
Predict.
Suggest.
Infer.
Evaluate.
These do not request identical responses.
A learner who writes an explanation when asked to state may waste time.
A learner who states a fact when asked to explain may omit the mechanism.
The command helps define the answer shape.
4. Then find the target
“Explain why the bulb is dimmer.”
“Explain why the plant grew less.”
“Explain why less water was collected.”
All three ask for explanation.
The target differs.
The learner must identify what outcome is being explained.
Otherwise a scientifically true paragraph can still answer the wrong question.
5. Conditions determine which concept is relevant
A familiar topic is not enough.
Which condition changed?
Which condition stayed the same?
Which object is being compared?
Which stage?
Which direction?
Which setup?
Maya’s common error is recognising the topic before reading the condition.
Her repair is simple:
command + target + condition before concept.
6. Concepts are tools, not decorations
The word “conductor” is useful if the question is about transfer through a material.
It may be irrelevant if the question asks about flexibility.
The word “photosynthesis” is useful if the process explains the observed plant outcome.
It may be irrelevant if the question asks about water loss through leaves.
Scientific vocabulary earns its place through relevance.
7. Explanation requires a relationship
“Waterproof” names a property.
“The material is waterproof, so water does not pass through it easily and the contents remain dry” carries the relationship.
The second answer connects:
property → effect → function
That is an explanation architecture.
8. Primary Science often uses a small number of recurring explanation structures
Property → function.
Structure → function.
Condition → process → outcome.
Evidence → inference.
Change → effect.
Comparison → conclusion.
Interaction → observed result.
Learners become stronger when they recognise these relationships beneath the topic surface.
9. Structure-function explanations are a major Science family
Why is this material suitable?
Why does this plant part help?
Why is this organ structured in this way?
Why does an adaptation improve survival under particular conditions?
The explanation links a feature to what that feature enables.
At younger levels the chain may be short.
Later Biology makes it more detailed.
10. Condition-process-outcome explanations are another major family
Light decreases.
A process is affected.
An outcome changes.
Resistance changes.
Current changes.
Brightness changes.
Temperature changes.
Particle behaviour changes.
An observed property changes.
This family becomes increasingly important as causal chains lengthen.
11. Evidence-inference explanations require discipline
The learner observes X.
What can be concluded?
Not everything plausible.
Only what the evidence supports together with relevant scientific knowledge.
The explanation should make that connection visible.
12. A keyword can be correct and still earn little
Jia Jun writes:
“Friction.”
The word identifies the concept.
But what did friction do?
Where did it act?
How did it affect the motion?
The marker needs the scientific relationship.
13. A long answer can also be weak
Ethan knows a great deal.
He writes six sentences.
Three are correct.
Two are irrelevant.
One contains the actual explanation.
Length is not depth.
Depth is the completeness and accuracy of the relevant relationship.
14. Concision should come after completeness
Jia Jun wants the shortest possible answer.
That can become a strength later.
First he must know what cannot be removed.
Scientific concision preserves:
the subject;
the relevant scientific relationship;
the direction of change where needed;
and the required evidence.
Everything else can be tested for necessity.
15. Causal chains need direction
A learner may know A and B but reverse the arrow.
Does higher temperature cause faster particle movement in the model, or is the answer being written backward?
Does reduced light affect a process, or does the process reduce the light?
Direction is part of scientific accuracy.
Useful verbs help:
increases;
decreases;
causes;
allows;
prevents;
transfers;
produces;
absorbs;
reflects;
reduces.
16. “Because” does not guarantee causality
A student can write:
“The plant grew less because it was shorter.”
Grammatically causal.
Scientifically empty.
Another writes:
“The bulb is dimmer because less brightness is produced.”
Restatement, not mechanism.
The word “because” cannot do the reasoning by itself.
17. Circular explanations should be exposed
“It dissolved because it is soluble.”
Depending on context, this may identify a property but may not explain the mechanism being asked.
“It reacted because it is reactive.”
“It grew because it was growing.”
Circularity often appears when vocabulary replaces relationship.
Ask:
What new information did the second half of the sentence add?
18. Scientific explanation often needs intermediate steps
At Primary 3, one link may be enough.
At Primary 5, perhaps two or three.
At Secondary level, several intermediate processes may be needed.
The student must know how much mechanism the question and mark allocation require.
The goal is not to write every fact known.
It is to include enough causal structure.
19. Intermediate steps are where hidden misconceptions live
A student begins with the correct condition and ends with the correct outcome.
Ask for the middle.
Suddenly the misconception appears.
This is why explanation is diagnostically powerful.
The middle of the chain reveals the model.
20. Evidence should be woven into explanation when the question supplies it
A table shows one setup produced fewer bubbles.
The learner should not ignore the table and recite a generic fact.
Use the evidence.
“Setup B produced fewer bubbles under the changed condition, indicating a lower rate of the process…”
The exact wording depends on level and question.
The principle is stable:
given evidence should do work in the answer.
21. Quoting data is not the same as interpreting it
“A = 12, B = 7.”
That reports.
“A produced a greater value than B under the stated conditions.”
That compares.
“The difference supports the prediction that…”
That interprets.
Strong explanations move beyond copying numbers.
22. Graph explanations require variable language
“The graph goes up” is visual description.
Scientific interpretation names the variables.
“As X increases, Y increases over this range.”
Then the learner may need to explain the mechanism behind the pattern.
This translation from shape to relationship is central in Secondary Science.
23. Diagrams distribute information across space
Labels.
Arrows.
relative positions.
connections.
stages.
symbols.
The learner has to convert spatial information into a verbal relationship.
That is why diagram-to-sentence practice is valuable.
24. Maya’s explanation problem begins before writing
She reads quickly.
Chooses the right topic.
Misses the changed condition.
Her explanation is scientifically valid for the wrong scenario.
The repair is not better prose.
It is better question reception.
This is why explanation diagnosis must move upstream.
25. Jia Jun’s explanation problem is under-expression
His mental chain is longer than his written chain.
The tutor asks:
“What happens because it is waterproof?”
He answers aloud.
“Write that.”
One repeated intervention can change many future answers.
26. Hana’s explanation problem is over-qualification
“Maybe.”
“Perhaps.”
“I think.”
Scientific uncertainty matters, but when the evidence and school model support a clear conclusion, excessive hedging weakens communication.
Her repair is confidence calibrated to evidence.
27. Ethan’s explanation problem is relevance control
He knows alternative explanations.
Excellent in inquiry.
Dangerous in a short-answer question if every possibility is written.
His repair is:
answer the requested relationship first; add qualification only if the question requires it.
28. Scientific vocabulary needs verbs
Nouns name entities and concepts.
Verbs carry change and interaction.
plant;
water;
light;
food.
Those nouns do not yet form a model.
absorbs;
transports;
uses;
produces.
Now relationships appear.
Good Science writing depends heavily on accurate verbs.
29. Connectives carry logic
because;
therefore;
so;
as a result;
compared with;
whereas;
when;
if;
under these conditions;
Scientific language uses connectives to expose the structure of reasoning.
But the concepts must still be correct.
30. Pronouns can hide unclear scientific reference
“It moves there because it causes it to increase.”
What is “it”?
Which quantity increases?
Scientific writing benefits from explicit nouns when multiple entities are present.
This is where English precision becomes part of Science performance.
31. Comparison answers need a common basis
“Plant A is taller and Plant B has more leaves.”
Two facts.
Not a clean comparison on one basis.
Strong comparison aligns the dimension:
height with height;
leaf number with leaf number;
temperature with temperature;
rate with rate.
Then the learner can state similarity or difference precisely.
32. Prediction answers should include direction when possible
“It will change.”
How?
Increase?
Decrease?
Move left?
Become brighter?
Take longer?
A prediction should be specific enough to be testable.
33. Inference answers should not exceed the evidence
The learner sees one result.
Then writes “all.”
Or “always.”
Or a broad causal conclusion that the setup did not establish.
Scientific explanation requires scope control.
Say what the evidence earns.
34. Evaluation answers require a three-part structure
limitation → effect → improvement
What is the problem?
Why does it matter?
What change would address it?
This is stronger than generic phrases such as “human error.”
See How Science Experiment Design Works.
35. Explanation should match developmental level
A Primary 3 learner does not need university mechanism.
A Secondary student may need more than a Primary explanation.
Good teaching chooses the level of resolution required by the curriculum while avoiding statements that create future misconceptions.
This is model calibration.
36. More detail is not always more correct
Adding advanced facts can make an answer less relevant.
It can introduce inaccuracies.
It can waste time.
It can obscure the relationship expected at the current level.
The best explanation is sufficient, accurate and appropriately resolved.
37. Less detail is not always elegant
“Because of energy.”
Too little.
“Because of force.”
Too little.
Concision becomes elegant only after the mechanism is preserved.
38. Model answers should reveal structure, not become scripts
A model answer is useful for comparison.
What concept did it use?
Where is the causal link?
How did it use evidence?
Which words were essential?
Then close it.
Answer a changed-context question.
If the learner can transfer the structure, the model answer taught something useful.
39. Sentence frames are scaffolds
“Because ___, therefore ___.”
“Compared with ___, ___.”
“When ___ increases, ___ decreases because ___.”
These frames can help a learner who cannot yet organise the relationship.
But scaffolds should fade.
The learner must eventually choose the answer architecture independently.
40. Oral explanation is a useful bridge to written explanation
Ask the learner to talk through the answer first.
Listen for the correct relationship.
Then convert the oral reasoning into a concise written response.
This is especially useful when understanding is stronger than written expression.
Jia Jun often needs exactly this bridge.
41. But oral fluency can also hide weak precision
Ethan can speak brilliantly for two minutes.
Ask for one sentence.
Now the central relationship must be selected.
Compression can reveal whether the learner knows what matters most.
42. Drawing can reveal explanation gaps
If words are tangled, ask for a simple diagram.
What changes?
What direction?
What flows?
What interacts?
Then translate the diagram back into language.
Representation switching exposes missing edges.
43. Primary 3 explanation: complete the first relationship
The main job is to move beyond one-word answers.
Property → use.
Observation → simple inference.
Classification rule → placement.
Magnetic interaction → result.
Short, complete, age-appropriate.
44. Primary 4 explanation: connect processes
The learner increasingly needs to describe how one part or condition affects another.
Diagrams and sequences support explanation.
The child should learn to write relationships rather than only list facts.
45. Primary 5 explanation: causal chains lengthen
Systems create multi-step answers.
The child may need two or three connected causal moves.
Working memory becomes a bottleneck.
Drawing the chain before writing can help.
46. Primary 6 explanation: integration under examination constraints
The learner must identify relevant concepts quickly, use evidence, avoid answer-template overreach and communicate enough mechanism without wasting time.
This is where explanation becomes part of PSLE performance control.
See Primary 6 Science and PSLE Science in Punggol.
47. Secondary explanation: invisible models enter the chain
Now the learner may explain visible events using particles, cells, forces, energy or electrical quantities.
The chain crosses scales.
Observation → model → mechanism → outcome.
Representational fluency becomes essential.
See How Scientific Models Grow With the Learner.
48. Physics explanations often combine words and equations
An equation can establish a quantitative relationship.
The written explanation should interpret what that relationship means in the situation.
Substitution alone is not always understanding.
Words and Mathematics should reinforce each other.
49. Chemistry explanations often cross three representational layers
Macroscopic observation.
Particle explanation.
Symbolic representation.
Students need to know which layer the question asks for and how the layers connect.
Confusing them produces weak explanations.
50. Biology explanations often require structure, process and system
What structure is present?
What function does it support?
Which process occurs?
How does that affect the organism or system?
Long Biology answers become easier when this architecture is visible.
51. Scientific uncertainty should appear when the evidence requires it
Not every conclusion is absolute.
A result may suggest.
Support.
Be consistent with.
Fail to establish.
A mature learner calibrates language to evidence strength.
But uncertainty language should not be pasted everywhere.
52. “Proves” is a high-risk word
School students often write “this proves…” after limited evidence.
More careful language may be appropriate depending on context:
supports;
shows under these conditions;
is consistent with;
suggests.
The exact wording should suit the level and question.
The habit is scope control.
53. Explanations should survive paraphrase
If the learner can only produce one memorised sentence, the concept may be brittle.
Ask the student to explain the same relationship in a different accurate way.
Then use it in another context.
Meaning should survive wording changes.
54. Explanations should survive representation change
Same concept.
New diagram.
New table.
New graph.
Can the learner still reconstruct the explanation?
This is a transfer test.
55. Explanation error logs should record the missing relationship
Not only:
“Forgot keyword.”
Better:
“Named waterproof property but did not connect it to water not passing through and the object staying dry.”
The second note tells the learner what to repair.
56. A three-student explanation clinic
The tutor places three answers on the table.
One has the correct keyword only.
One has the correct relationship but vague vocabulary.
One has precise vocabulary but reversed causality.
The students diagnose each.
Which is closest?
What is missing?
Which error is conceptual?
Which is linguistic?
This makes quality visible.
57. Peer explanation is valuable because listeners test coherence
If Hana explains to Maya and Maya asks, “Why does that cause this?” a missing link becomes visible.
Peer questions can expose assumptions the speaker did not notice.
Small-group tuition can use this deliberately while preserving independent attempt first.
58. Parents can help with one question
“Finish the thought.”
Child: “Because it is waterproof.”
Parent: “Finish the thought.”
Child: “So water does not pass through easily and the contents stay dry.”
No forty-minute lecture required.
One prompt can build a habit.
59. Another useful parent question is “What in the question makes you say that?”
This pulls evidence into the explanation.
It also helps separate general knowledge from information supplied in the setup.
Again, use lightly.
Home should not become a continuous oral examination.
60. Retrieval supports explanation because the relationship must be available
If the learner cannot retrieve the concept, no amount of writing structure will save the answer.
Knowledge and language must meet.
The next article, How Science Retrieval and Memory Work, develops this layer.
61. Transfer supports explanation because unfamiliar contexts remove scripts
A learner may memorise one perfect model answer.
Change the object.
Change the diagram.
Change the condition.
Can the learner rebuild the relationship?
See How Science Transfer Works.
62. A compact Science explanation checklist
- What is the command?
- What exactly is the target?
- Which condition matters?
- Which scientific concept is relevant?
- What evidence is given?
- What direction does the relationship run?
- Which intermediate steps are necessary?
- Have I used precise nouns and verbs?
- Did I answer the question rather than display everything I know?
- Does the conclusion match the evidence?
63. Frequently asked questions
Why can my child know Science but lose open-ended marks?
The weak link may be translating understanding into a complete, relevant and precise causal explanation.
Are keywords important?
Yes. Scientific vocabulary supports precision, but keywords must be used inside the correct relationship.
Should my child memorise model answers?
Use model answers to study structure and precision. Then close them and answer a changed-context question to test transfer.
How long should a Science answer be?
Long enough to make the required relationship visible, but no longer than necessary. The needed length depends on the level, question and marks.
Why does my child write too much?
The learner may have weak relevance control or may not know which relationship the command is targeting.
Why does my child write too little?
The learner may compress the answer before completing the causal chain, or may not know how to express the relationship.
How can parents help?
Ask occasional prompts such as “Finish the thought” or “What in the question makes you say that?” rather than re-teaching the entire topic.
How does explanation change in Secondary Science?
Explanations become more abstract, quantitative and discipline-specific, often requiring invisible models, longer causal chains and precise technical vocabulary.
64. Continue the Science Education Systems series
- Science Education Systems
- How Scientific Thinking Is Built
- How Science Assessment Works
- How Science Misconception Repair Works
- How Science Knowledge Networks Work
- How Science Experiment Design Works
- How Science Retrieval and Memory Work
- How Science Transfer Works
Conclusion: An explanation is a bridge another mind can cross
Maya has the observation.
Jia Jun has the keyword.
Hana has the precision.
Ethan has the possibilities.
Science asks each of them to build a bridge.
From question to concept.
From concept to evidence.
From evidence to mechanism.
From mechanism to language.
The best answer is not the answer with the most impressive vocabulary.
It is the one that allows another person to follow the scientific relationship without having to guess what the learner meant.
Read the command.
Find the target.
Respect the condition.
Select the concept.
Use the evidence.
Carry the causal chain.
Write precisely.
Check the direction.
Stop when the explanation is complete.
Then try it again when the question changes.
