NSW Science (Years 7 to 10)
From Observations to Inferences and Predictions: What NSW Science Students Need to Know
Many students lose marks in their school assessment by writing an inference when the question asks for an observation, or by confusing a prediction with a guess. These three ideas, observations, inferences, and predictions, look similar on the surface but mean very different things in science. Getting them straight is one of the highest-value skills you can develop for the Observing the Universe module.
What the syllabus asks
The NSW Science 7–10 curriculum asks you to understand how scientists move from gathering information about the world to explaining it and predicting what will happen next. That journey runs through three connected steps: making observations, drawing inferences from those observations, and forming testable predictions that can be checked. The curriculum also asks you to recognise that scientists use two different investigation approaches, systematic observation and controlled experiments, and to understand when each one is appropriate.
The idea, explained
Observations
An observation is information gathered directly, using your senses or instruments. "The grass is wet" is an observation. "The temperature is 18°C" is an observation. Notice that neither statement includes a reason or an explanation. You are simply recording what you detect.
Inferences
An inference is a possible explanation of an observation, built from the evidence you have collected plus prior knowledge you already hold. If you observe that the grass is wet, you might infer that it rained overnight. That inference is reasonable, but it is not the only one. The same observation can lead to different inferences: rain, overnight dew, or a sprinkler left running could all explain wet grass. This is a crucial point. Because one observation can support more than one inference, scientists gather more observations to work out which inference is best supported by the evidence.
Testable predictions
Once you have an inference, you can use it to make a prediction. A prediction states what should happen in a particular situation if your inference is correct. Scientists often write predictions in an "If... then..." format. For example: if it rained overnight, then the road outside will also be wet. That prediction is testable because you can go outside and check the road. If the prediction comes true, your inference is supported. If the road is dry, your inference must be revised or rejected. This back-and-forth between predictions and observations is exactly how scientific knowledge moves forward.
Two ways scientists investigate
Not all scientific questions can be answered the same way, and the curriculum asks you to understand two legitimate approaches.
Systematic observation means making careful, planned, and repeated observations of the natural world without interfering with it. Tracking weather patterns over months, observing animal behaviour in the wild, and mapping the positions of stars are all examples. Scientists use this approach when the subject cannot be brought into a laboratory or manipulated. You cannot run a controlled experiment on a galaxy or a volcano, so systematic observation is the appropriate method.
Controlled experiments are used when the scientist can manipulate the situation. The scientist deliberately changes one factor, keeps all other factors constant, and measures what happens. Both approaches are genuine scientific research. The nature of the question being asked is what determines which approach fits best.
What the exam asks
School assessment questions on this concept tend to fall into a few predictable patterns.
- Classify a statement: You may be given a sentence and asked whether it is an observation or an inference. Look for whether the statement records a direct measurement or sense experience (observation) or offers an explanation (inference).
- Write a testable prediction: You may be given an inference and asked to write a prediction. Use the "If... then..." structure and make sure the prediction can actually be checked by observation or experiment.
- Choose an investigation method: A question might describe a scientific scenario and ask whether systematic observation or a controlled experiment is more appropriate. Ask yourself: can the scientist control and manipulate the variables? If yes, a controlled experiment fits. If the subject is in the natural world and cannot be interfered with, systematic observation is the answer.
- Evaluate an inference: You might be told that a prediction failed and asked what the scientist should do. The answer is that the inference must be revised or rejected.
Common mistakes
- Mixing up observations and inferences. Writing "it rained last night" when asked for an observation is a very common error. That is an inference. The observation is simply "the grass is wet."
- Writing untestable predictions. A prediction must be checkable. "The universe is interesting" cannot be tested, so it is not a scientific prediction.
- Thinking systematic observation is less scientific than a controlled experiment. Both are valid scientific methods. Systematic observation is not a shortcut; it is the correct approach when variables cannot be controlled.
- Assuming one observation proves an inference. A single observation supports an inference but does not prove it. More observations are needed to decide between competing inferences.
- Forgetting to revise the inference when a prediction fails. If a prediction does not come true, the inference behind it must be reconsidered, not ignored.
If you want to work through practice questions and get instant feedback on exactly this kind of reasoning, Avocado's AI-native tutoring service has you covered. Try the From Observations to Inferences and Predictions lesson to build confidence before your next school assessment.