Primary 6 Science Learning Guide | Multi-Step Problem Solving, Constraints & Decision Paths for PSLE

The hardest Primary 6 Science questions often require several correct decisions in sequence. The pupil must read a diagram, identify the changed condition, choose the relevant concept, interpret evidence, eliminate an alternative and then explain the outcome. One early mistake can send the entire answer down the wrong path.

This guide develops multi-step problem solving, constraint propagation, decision paths and structured scientific reasoning for PSLE Science.

Return to the Primary 6 Science Learning Hub.

The multi-step rule

RECONSTRUCT → IDENTIFY CONSTRAINTS → CHOOSE FIRST DECISION → PROPAGATE CONSEQUENCES → CHECK EVIDENCE → REVISE IF NECESSARY → ANSWER.

This is an eduKate reasoning routine, not an official SEAB marking formula.

Part I — Multi-step does not mean long

A question can require several reasoning operations even if the final answer is one sentence.

Example:

1. Identify that surface type changed.
2. Notice car and release stayed the same.
3. Compare travel distances.
4. Select friction as the relevant interaction.
5. Explain the shorter distance.

The final answer may still be concise.

Part II — Constraints remove branches

Constraints are facts that rule out possibilities.

If the same car and same release are used, differences in car mass or release energy are removed as explanations.

If water volume and starting temperature are equal, those conditions cannot explain a difference in cooling.

Part III — Use a decision tree mentally

Question: Why did Plant B produce fewer bubbles?

Decision path:

• Did light differ? Yes.
• Did water differ? No.
• Did time differ? No.
• Is bubble count the measured outcome? Yes.
• Relevant relationship: light availability and photosynthesis-related output.

This reduces random concept selection.

Part IV — Solve the first uncertain step first

If you do not know what the arrows in a food web mean, do not predict population changes yet.

If you do not know which variable is measured, do not evaluate the conclusion yet.

Repair the earliest uncertain step before continuing.

Part V — Constraint propagation

When a condition is fixed, its consequences travel through later reasoning.

“All other conditions remain the same” means later steps cannot introduce a new temperature, mass or time difference unless the question states one.

Part VI — Avoid branching explosion

Some pupils imagine every possible scientific cause.

Use question conditions to eliminate branches quickly.

Only retain alternatives consistent with the actual setup.

Part VII — Multi-step graph questions

A graph question may require:

1. read axes;
2. identify trend;
3. locate plateau;
4. compare two regions;
5. infer a boundary;
6. predict cautiously outside the measured point.

Skipping the axis step can invalidate all later reasoning.

Part VIII — Multi-step experiment questions

An investigation question may require:

1. identify the scientific question;
2. identify changed variable;
3. identify measured variable;
4. identify a confound;
5. suggest an improvement;
6. judge the conclusion.

These are connected but distinct jobs.

Part IX — Multi-step system questions

Example: drought affects a food web.

Possible chain:

less water → fewer plants → less food for herbivores → herbivore decline → possible predator effects.

Check whether predators have alternative prey before continuing.

Part X — Multi-step energy questions

Solar fan:

light energy → electrical energy → motor operation → kinetic energy of blades.

If fan speed falls, identify whether the changed condition affects input, pathway, converter or output.

Part XI — Multi-step force questions

Spring car:

• spring compressed;
• elastic force acts;
• car begins moving;
• surface friction slows car;
• distance measured.

Different sub-questions may target different stages.

Part XII — Multi-step biological systems

Meal to movement:

food eaten → digestion → nutrients absorbed → blood transports nutrients → respiration releases energy → muscles perform work.

Do not collapse all systems into “food gives energy”.

Part XIII — Decision checkpoints

After each major step ask:

• Does this match the question?
• Does this contradict any given condition?
• Is the direction correct?
• Do I have evidence?
• Am I still answering the target?

Part XIV — Backtracking

If a later result looks impossible, return to the earliest decision.

Example: conclusion says Plant B had more photosynthesis although it produced fewer bubbles under otherwise comparable conditions.

Check whether the trend was read backwards or the target plant was confused.

Part XV — Contradiction as an alarm

A contradiction is useful because it tells the pupil something upstream may be wrong.

Do not patch the final sentence while leaving the faulty earlier decision untouched.

Part XVI — External working reduces load

Use small notes:

Changed: lamp distance

Measured: bubbles / 5 min

Same: plant, water, time

Trend: farther → fewer

Then write the answer.

This reduces working-memory demand.

Part XVII — One representation at a time

If a question combines text, diagram and table:

1. read text for conditions;
2. read diagram for structure;
3. read table for evidence;
4. combine only after each is understood.

Part XVIII — Original workshop 1 — mixed circuit question

A circuit diagram has two branches and a table of bulb brightness.

Steps:

1. trace branches;
2. identify which switch changes;
3. determine which branch opens;
4. read brightness table;
5. explain only the affected bulbs.

Original workshop 2 — cooling graph

Two cooling curves are shown.

Steps:

1. confirm same starting temperature;
2. compare temperatures at same times;
3. identify which cools more slowly;
4. infer better insulation;
5. avoid claiming behaviour beyond the graph.

Original workshop 3 — plant experiment

Two plants differ in light and water.

Steps:

1. identify intended variable: light;
2. notice water also changed;
3. recognise confound;
4. reject strong causal conclusion;
5. suggest keeping water comparable.

Original workshop 4 — food web change

Insect X decreases. Bird A eats X and Y. Hawk eats Bird A.

Steps:

1. identify X decrease;
2. check Bird A’s alternative food Y;
3. avoid assuming Bird A must fall;
4. therefore avoid assuming hawk must fall;
5. write a conditional prediction.

Part XIX — The PATH test

1. P — Problem: what exact job is required?
2. A — Allowed facts: what conditions and evidence are given?
3. T — Track decisions: what sequence of reasoning is needed?
4. H — Halt and check: does each step still fit?

This is an eduKate teaching mnemonic.

Part XX — Common multi-step errors

• Skipping the first variable-identification step.
• Introducing a condition ruled out by the question.
• Reading a graph correctly but choosing the wrong mechanism.
• Using a true concept at the wrong stage.
• Jumping over an essential intermediate step.
• Failing to backtrack after contradiction.
• Overloading working memory instead of writing small notes.

Part XXI — Timed multi-step strategy

Under time pressure:

1. identify command;
2. mark changed and measured quantities;
3. write one-word concept candidate;
4. verify with evidence;
5. answer directly;
6. flag if contradiction remains.

Where to connect

• Question Deconstruction, Conditions & Constraints
• Integrated Multi-Concept Reasoning & Unfamiliar PSLE Setups
• Primary 4 Multi-Step Problem Solving

Retrieval checklist

• I can break a complex question into decisions.
• I use constraints to eliminate impossible branches.
• I solve the earliest uncertain step first.
• I propagate fixed conditions through later reasoning.
• I can backtrack when a contradiction appears.
• I separate text, diagram and table before combining them.
• I can use external notes to reduce working-memory load.
• I recognise when different concepts belong to different stages.
• I check each step against evidence.
• I can produce a concise final answer after a complex reasoning path.

The quiet return

Complex Science questions become manageable when the pupil stops treating them as one giant problem. Each correct decision narrows the next one.

Reconstruct. Constrain. Decide. Propagate. Check. Backtrack when needed. Then answer cleanly.

Return to the Primary 6 Science Learning Hub.