In Science, a child may see a graph, experiment or picture and write what they think is happening. The answer may sound reasonable, but it can still miss the task. A question asking for an observation requires evidence from what is shown; an inference asks the student to interpret that evidence; a prediction asks what may happen under a stated condition.
For Punggol parents, these distinctions are worth teaching explicitly. Children can know the Science topic yet lose marks because they use the wrong kind of response. Process skills are not vocabulary to memorise in isolation. They are ways of thinking carefully about evidence and explaining what can—and cannot—be concluded.
- Teach the difference between what is directly seen, what is interpreted and what is expected to happen.
- Practise using evidence from diagrams, tables, experiments and everyday situations.
- Help the child avoid adding explanations when the task asks only for an observation.
- Check that predictions are linked to the stated condition and relevant concept.
- Use fresh examples so the student learns the thinking process, not a memorised sentence.
Why these Science verbs matter
Science questions often ask students to do different things with information. A child might be shown two plants, a results table or an experimental setup. The student must understand whether the task is to record what is visible, explain why it may have happened or state what is likely to happen next.
When students treat every question as “explain”, they may write too much, introduce unsupported ideas or overlook the precise evidence requested. When they treat every question as “state what you see”, they may fail to use scientific knowledge where interpretation is required.
Observation: describe the evidence
An observation is based on what can be directly seen, heard, measured or recorded. It should be specific and factual. For example, “the liquid became cloudy” describes a visible change. “A new substance formed” may be an inference unless the observation and context provide enough evidence to support that conclusion.
In a graph, an observation might be that the measured value increased as time increased over the interval shown. A good observation refers accurately to the information. It should not add a cause that the graph alone does not establish.
Inference: interpret the evidence
An inference is an interpretation based on observations and relevant knowledge. It explains what the evidence may mean. The student needs to connect the evidence to a suitable scientific idea, while avoiding claims that go beyond what is supported.
For example, if an object becomes warmer after receiving energy, a student may infer that energy has been transferred to it. The answer should be appropriate to the topic and the evidence provided. A plausible idea is not automatically a justified inference.
Prediction: say what may happen next
A prediction is a statement about an expected outcome, usually based on a pattern, prior observation or scientific principle. It should relate to the condition described in the question. If a graph shows a consistent increase across a range, the student may predict a further increase over a nearby interval, provided the context supports that expectation.
Students should avoid treating a prediction as a certainty when the evidence does not justify certainty. They should also state the predicted outcome clearly rather than simply repeating the trend without saying what is expected to happen.
Hypothesis: propose a testable explanation
A hypothesis is a proposed explanation or relationship that can be investigated. It is more than a guess. A useful hypothesis connects the factor being changed with an outcome that can be observed or measured. The exact wording expected depends on the level and task, but the student should understand that a hypothesis must be capable of being tested.
For example, a student might propose that increasing the amount of light available to a plant within the investigated range will affect its rate of growth. A good investigation then needs a suitable method and measurements to test the proposed relationship. Students should avoid writing a hypothesis that cannot be investigated with the evidence or apparatus available.
Use one simple situation to practise all four
Consider a classroom investigation where two identical containers are placed in different light conditions and plant growth is recorded over time. The exact scientific details should follow what the class has learnt.
- Observation: describe the recorded difference in the plant measurements.
- Inference: suggest what the observed difference may indicate, using relevant knowledge.
- Prediction: state what may happen if the same conditions continue, and explain the basis for the expectation.
- Hypothesis: propose a testable relationship between the light condition and a measurable growth outcome.
The student should be able to explain why each response is different. This is more useful than memorising four definitions and hoping they recognise the right word in an examination.
How parents can practise process skills at home
Use safe, ordinary observations such as changes in shadows during the day, melting ice in a suitable container or the pattern of water droplets on a cool surface. There is no need to turn the home into a laboratory or use chemicals, heat sources or electrical equipment. Ask the child to distinguish the visible evidence from what they think caused it.
For examination preparation, use school-provided diagrams, data tables and experiments. Ask the student to answer one question using only the evidence shown. Then ask what additional information would be needed to make a stronger inference or prediction.
Common mistakes to look for
- Writing a cause when the question asks only for an observation.
- Giving an opinion without pointing to evidence.
- Making a prediction that ignores the condition in the question.
- Calling an untestable guess a hypothesis.
- Claiming certainty when the evidence supports only a possible explanation.
- Repeating a definition without applying it to the situation.
Primary and Secondary expectations
Primary students can begin by making careful observations, comparing results and giving simple predictions supported by what they have noticed. Parents should use the language and depth taught at the child’s level, rather than introduce unnecessary technical detail.
Secondary students may be expected to work with variables, experimental design, evidence and more formal explanations depending on their G1, G2 or G3 subject. Use the current syllabus and teacher feedback for the student’s cohort. The official starting points are SEAB’s PSLE page and SEAB’s SEC information.
Frequently asked questions
Is an inference the same as an observation?
No. An observation describes evidence that can be seen or measured. An inference interprets that evidence using relevant knowledge.
Does every prediction need a reason?
The question determines how much explanation is required. The student should at least understand what evidence or principle supports the prediction and follow the task’s instructions.
Can a hypothesis be wrong?
Yes. A hypothesis is a testable proposal, not a guaranteed conclusion. An investigation may support or fail to support it.
Where can I read more?
See our guide to Science processes, observations and data and helping children explain answers.
Science answers should match the evidence task
When a student can distinguish observation, inference, prediction and hypothesis, they become more precise about what the evidence shows and what they are being asked to do with it. This builds exam craft, but more importantly, it develops honest scientific thinking: describe first, interpret carefully and do not claim more than the evidence allows.

