Develop and evaluate questions and hypotheses for scientific investigation

What this dot point is asking

NESA wants you to construct a testable hypothesis from an inquiry question, distinguish a hypothesis from a prediction, and evaluate whether a proposed hypothesis is scientifically valid. Hypothesis construction is examined every year in Investigating Science.

The answer

A scientific investigation starts with an inquiry question that frames what we want to know. A good inquiry question is specific, answerable through observation or experiment, and connects to existing scientific knowledge.

From inquiry question to hypothesis

An inquiry question becomes a testable investigation through a hypothesis.

Hypothesis

A proposed explanation for an observation, written as a statement that names the variables and predicts a relationship. A scientific hypothesis must be falsifiable: there must be possible observations that would disprove it.

Example inquiry question

Does temperature affect the rate of an enzyme-catalysed reaction?

Hypothesis

As temperature increases from 10 to 40 degrees Celsius, the rate of catalase-mediated decomposition of hydrogen peroxide will increase, reaching a maximum near 37 degrees Celsius and then decreasing as the enzyme denatures.

Hypothesis versus prediction

These are related but distinct.

• A hypothesis is a general statement about the relationship between variables.
• A prediction is a specific, measurable outcome derived from the hypothesis under defined experimental conditions, usually phrased "if... then...".

Example. Hypothesis: temperature affects enzyme activity. Prediction: if catalase is exposed to 70 degrees Celsius for 5 minutes, then the rate of oxygen production will fall to less than 10 per cent of the rate at 37 degrees Celsius.

Falsifiability (Popper)

Philosopher Karl Popper argued that the defining feature of a scientific claim is that it can be proven wrong. A hypothesis like "consciousness influences quantum events" cannot be falsified by any conceivable experiment and is therefore not scientific. "Aspirin reduces fever in adults at 500 mg doses" can be falsified by a controlled trial and is therefore scientific.

What makes a good hypothesis

A scientifically valid hypothesis:

1. Names the variables. Independent variable (what you change) and dependent variable (what you measure).
2. Predicts a direction. Increase, decrease, or no change.
3. Is measurable. "Better" or "more" are not measurements. Specify units.
4. Is falsifiable. There must be possible outcomes that would disprove it.
5. Is based on existing knowledge. Not a wild guess, but a reasoned proposal grounded in prior science.

Examples in context

Example 1. Cape Grim baseline air monitoring. CSIRO and the Bureau of Meteorology operate the Cape Grim baseline air-pollution station on Tasmania's north-west tip because the wind there crosses thousands of kilometres of Southern Ocean before reaching the inlet. A researcher could begin with the inquiry question "Has atmospheric methane concentration over the Southern Ocean increased since 2010?" and pose the falsifiable hypothesis "Mean annual methane concentration measured at Cape Grim in 2024 will exceed the mean for 2010 by more than 50 parts per billion." The prediction is numerical, the dataset already exists, and the hypothesis is rejected if the difference is smaller than measurement uncertainty.

Example 2. Yiyili cultural burn comparison. A two-way science project with the Yiyili Aboriginal community in Western Australia compared traditional cool-burn fire regimes with conventional hazard-reduction burns. An inquiry question might be "Does cultural burning produce a different post-fire grass species composition than conventional burns?" A testable hypothesis: "Sites burned under Aboriginal cultural fire regimes will have at least 20 per cent greater perennial grass cover one year post-fire than adjacent sites burned by conventional ignition." This is falsifiable, names the variables, predicts a direction, and respects two-way knowledge through co-designed protocols under AIATSIS guidelines.

Try this

Q1. Define falsifiability and state why it is essential to a scientific hypothesis. [3 marks]

• Cue. Define as "capable of being disproved by observation," link to Popper, give one example of a falsifiable and one unfalsifiable claim.

Q2. A student writes the hypothesis: "Drinking green tea makes people healthier." Identify two flaws and rewrite the hypothesis so it is testable. [2+2 marks]

• Cue. Flaws: "healthier" is unmeasurable; "green tea" lacks dose or duration. Rewrite naming a measurable dependent variable, e.g. resting systolic blood pressure after 12 weeks.

Q3. A researcher proposes investigating whether traditional Bundjalung bush-medicine plants reduce wound healing time. (a) Construct an inquiry question. (b) Write a falsifiable hypothesis. (c) Identify one ethical consideration before any data is collected. [2+3+2 marks]

• Cue. (a) Frame as comparative question. (b) Name plant, application, measurable outcome, time frame. (c) AIATSIS-compliant community consent and benefit-sharing before any sample collection.