Primary 6 Science Learning Guide | Testable Questions, Hypotheses & Investigation Design for PSLE

Primary 6 Science becomes more powerful when pupils can ask a scientific question, not only answer one. A good investigation begins before the apparatus is touched. The learner must identify a relationship worth testing, turn it into a testable question, make a reasoned hypothesis, choose the correct variables and decide what evidence would support or challenge the idea.

This guide develops testable questions, hypotheses, predictions and investigation design for Primary 6 and PSLE Science. It complements the existing fair-test and practical-planning guides by beginning one step earlier: with the scientific question itself.

Return to the Primary 6 Science Learning Hub.

The investigation-question rule

OBSERVATION → CURIOSITY → TESTABLE QUESTION → HYPOTHESIS → VARIABLES → METHOD → EVIDENCE → CONCLUSION.

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

Part I — Scientific curiosity is not yet a testable question

Curiosity: “Why do some surfaces make a toy car stop faster?”

Testable question: “How does surface type affect the distance travelled by the same toy car when the release condition is kept the same?”

The second version names a changed condition and a measurable outcome.

Part II — A testable question needs measurable evidence

Weak question: “Which plant likes light more?”

Stronger: “How does lamp distance affect the number of bubbles produced by an aquatic plant in five minutes?”

The stronger question defines what changes and what will be measured.

Part III — Question structure

A useful structure is:

How does [changed variable] affect [measured variable] when [important controlled conditions] are kept comparable?

This is not the only valid wording, but it makes the scientific relationship visible.

Part IV — Hypothesis versus prediction

A hypothesis proposes a relationship that can be tested.

A prediction states what result is expected if the hypothesis is correct.

Hypothesis: “Rougher surfaces produce a greater frictional effect on the moving car.”

Prediction: “The car will travel a shorter distance on the rougher surface under the same release conditions.”

Part V — A hypothesis should be scientifically motivated

Do not write a hypothesis as a random guess.

Use prior scientific knowledge:

“If the lamp is moved farther from the plant, fewer bubbles are expected because less light reaches the plant and less light energy is available for photosynthesis under otherwise suitable conditions.”

Part VI — Hypotheses can be wrong and still be useful

A scientific hypothesis does not need to be correct before testing. It needs to be testable.

If the result does not match the prediction, the pupil should not alter the data. The hypothesis may need revision or the method may need checking.

Part VII — Questions that are too broad

“How does the environment affect animals?” is too broad for one simple investigation.

Better: “How does water availability affect the number of pill bugs observed in two otherwise comparable habitats?”

Narrowing the question makes the evidence manageable.

Part VIII — Questions that change too many variables

Weak: “How do light, water and temperature affect plant growth?”

That asks about three variables at once.

Better: test one factor while holding other relevant conditions comparable.

Part IX — Observational questions versus experimental questions

Observational question: “Are more insects found in shaded or open areas?”

Experimental question: “How does light level affect the activity of organisms in a controlled setup?”

Observational studies can reveal associations. Controlled experiments can provide stronger causal evidence when the design isolates a variable.

Part X — Turn a statement into a question

Statement: “Plants seem to lose water faster on hot days.”

Question: “How does temperature affect the amount of water lost from otherwise comparable plants over the same time?”

Then decide whether the method can actually measure water loss validly.

Part XI — Turn a graph pattern into a new question

Suppose bubble count rises with light level and then plateaus.

New question: “What other condition may be limiting the measured output once additional light no longer increases bubble count?”

This may require a new investigation, but the question emerges from the evidence boundary.

Part XII — Turn an anomaly into a question

Three spring trials are similar; one is very different.

Possible question: “Did the spring fail to return to its original length before the anomalous trial?”

An anomaly can create a new investigation rather than being discarded.

Part XIII — Original investigation design: friction

Observation: car stops sooner on some surfaces.

Question: How does surface type affect travel distance?

Hypothesis: rougher surfaces produce greater frictional effect.

Prediction: travel distance will be shorter on rougher surfaces.

Changed variable: surface type.

Measured variable: distance travelled before stopping.

Controls: same car, same release, same starting point.

Original investigation design: photosynthesis

Question: How does lamp distance affect bubble count in five minutes?

Hypothesis: greater lamp distance reduces light reaching the plant.

Prediction: bubble count will decrease as lamp distance increases, within the tested range.

Limit: bubble count is an indirect measure of gas production.

Original investigation design: cooling

Question: How does insulating material affect the temperature decrease of equal water volumes over ten minutes?

Hypothesis: better insulation reduces thermal-energy transfer to surroundings.

Prediction: the best insulator will produce the smallest temperature decrease.

Part XIV — Variables should come from the question

Do not memorise variable labels first and force them onto a setup.

Question defines the relationship.

Relationship defines the changed and measured variables.

Other relevant conditions become controls.

Part XV — The evidence plan

Before testing, ask:

• What result would support the hypothesis?
• What result would challenge it?
• How many values are needed to see a pattern?
• What measurements should be repeated?
• What anomaly would require checking?

Part XVI — A hypothesis is not proven by one matching result

One result can support a hypothesis, but strong scientific confidence comes from repeated, well-controlled evidence.

Use “supports” rather than “proves forever”.

Part XVII — Revising the hypothesis

Suppose increasing light no longer increases bubble count beyond a point.

The original simple hypothesis “more light always gives more bubbles” is too broad.

Revised: “Increasing light increases the measured output only within part of the tested range; another factor may become limiting at higher light levels.”

Part XVIII — Testable versus untestable in the school context

Some questions are scientifically meaningful but impractical for a school investigation because they require dangerous equipment, very long time periods or measurements beyond available instruments.

A good school investigation is both scientifically testable and practically measurable.

Part XIX — The ASK test

1. A — Affect: what factor may affect what outcome?
2. S — See: what measurable evidence will show the effect?
3. K — Keep: what relevant conditions must be kept comparable?

This is an eduKate teaching mnemonic.

Part XX — Common question-design errors

• Question is too broad.
• No measurable outcome.
• Several variables change at once.
• Hypothesis simply repeats the question.
• Prediction is written as certain fact.
• Method does not measure the stated outcome.
• Conclusion exceeds the question’s scope.

Where to connect

• Investigations, Variables, Fair Tests & Method
• Practical Planning, Data Recording & Conclusions
• Primary 4 Testable Questions, Hypotheses & Scientific Curiosity

Retrieval checklist

• I can turn curiosity into a testable question.
• I can identify a measurable outcome.
• I distinguish hypothesis from prediction.
• I can use scientific knowledge to justify a hypothesis.
• I know a hypothesis can be revised.
• I can distinguish observational from experimental questions.
• I can design variables from the question.
• I can state what evidence would support or challenge an idea.
• I can turn anomalies into new questions.
• I can keep a school investigation practical and measurable.

The quiet return

A strong scientist does not begin with apparatus. The work begins with a question precise enough to test and open enough to learn from.

Observe. Ask. Make the relationship testable. Predict carefully. Design the evidence. Let the result revise the model.

Return to the Primary 6 Science Learning Hub.