How Do You Choose a Manageable HSC Science Investigation Question?

You want to investigate why some toy cars travel farther than others. But “What makes a car fast?” is a huge question. You’ve got one ramp, one car, and a few science lessons. Where do you begin?

A manageable HSC science investigation question identifies one factor to change or compare, a response you can measure, and conditions you can keep reasonably consistent. It must also fit the equipment, time, and safety approval your school can provide.

For example, “What makes a car fast?” could become:

“How does the release height of a toy car on a fixed ramp affect its stopping distance on a level surface?”

That’s a question you can investigate. You know what to change, what to measure, and what equipment you’ll need. The next challenge is checking whether the experiment can produce evidence good enough to answer it.

01Start with something you can change and measure

Suppose you release a toy car from halfway up a ramp, then from near the top.

Which release would probably make the car travel farther across the floor?

You might predict the higher release point because the car starts with more gravitational potential energy. More of that energy can become kinetic energy as it rolls down.

That’s a reasonable prediction. But an investigation needs more than a reasonable prediction. It needs a way to test it.

Turn your interest into a testable question

A useful question structure is:

“How does [the factor you change] affect [the response you measure] under [specified conditions]?”

These parts have scientific names:

• Independent variable: The factor you deliberately change, such as release height.
• Dependent variable: The response you measure, such as stopping distance.
• Controlled variables: Important factors you try to keep constant, such as the car, ramp angle, and floor surface.

You also need boundaries: which values you’ll test, how you’ll measure the response, and when you’ll stop collecting data.

Here’s how the toy-car investigation develops.

The final question could be:

“How does release height (10 cm, 20 cm, and 30 cm) on a fixed ramp affect the stopping distance of the same toy car on a level floor?”

Notice that changing the release position along a fixed ramp is better than raising the entire ramp for each test.

Raising the ramp could change its angle or the connection to the floor. You would then be changing more than release height.

This leads to an important distinction: a question can sound precise without producing a fair comparison.

02Check whether the question can actually be answered

Before choosing your investigation, put it through four checks.

1. Can you measure the response clearly?

“How does temperature affect a reaction?” sounds scientific, but what response will you measure?

Depending on the reaction, you might measure gas volume, mass change, colour intensity, or the time taken to reach a defined endpoint.

These are not interchangeable. Each measures something different.

For example:

Choose the measurement before collecting anything.

Also consider your instrument’s resolution. If a ruler has 1 mm divisions, it cannot reliably distinguish two lengths differing by only 0.1 mm.

Repeating measurements is useful, but repeating an unsuitable measurement doesn’t automatically make it suitable.

A manageable question needs a response that your available instrument can detect with useful precision.

2. Can you make a fair comparison?

Imagine investigating whether warmer water dissolves sugar faster.

You prepare three cups at different temperatures. But you stir the hottest cup quickly, the middle cup slowly, and the coldest cup not at all.

Suppose the hottest cup finishes first. Was that because of temperature, stirring, or both?

You cannot tell.

Stirring is a confounding variable here: another factor that changes alongside your intended independent variable and could explain the result.

To isolate the effect of temperature, you would need a consistent stirring procedure, the same sugar mass and form, comparable water volumes, and a defined endpoint.

You also need to monitor water temperature because it can change during a trial.

The aim isn’t to control every imaginable influence. That’s usually impossible. It’s to identify the influences most likely to distort your comparison and manage them well enough for the question you’re asking.

3. Do you have the equipment and time?

A perfectly designed experiment on paper can fail when you discover that your school has one sensor shared between several groups.

Before committing, identify:

• The equipment you need, including measurement range and resolution.
• The number of conditions and repeated trials.
• The time required for preparation, measurements, resetting, and cleaning up.
• Whether the equipment and materials will actually be available when needed.
• A simpler measurement or approved secondary-data approach if the original method cannot work.

Think about time as a budget.

If each trial takes four minutes, nine trials require 36 minutes before preparation, cleaning, or unexpected delays.

And don’t confuse a 50-minute lesson with 50 minutes of uninterrupted data collection.

4. Will the school approve the method?

You should resolve safety and permission questions before becoming attached to a particular investigation.

Depending on the activity, concerns might include hot liquids, chemicals, electrical equipment, falling objects, fieldwork, biological material, or participation by other people.

The fact that an experiment appears online doesn’t mean your school can approve it.

Discuss the proposed materials, procedure, risks, supervision, and any ethical or privacy issues with your teacher. A risk assessment helps identify and control hazards, but it doesn’t replace school approval.

For NSW public schools, the NSW Department of Education’s science safety and compliance guidance is an official starting point. Other school sectors and individual schools may have additional or different procedures.

If the method isn’t approved, redesign it. Don’t treat safety requirements as obstacles to work around.

03Use a decision process before committing

You don’t need an elaborate scoring system. Work through these decisions in order.

These decisions have an order for a reason. There’s little point calculating how many trials you can complete if the school cannot approve the experiment.

Likewise, having enough time doesn’t help if your measuring instrument cannot detect the expected differences.

A secondary-source investigation may be an appropriate alternative if the task permits it, but the existing data must actually match the question you want to answer.

06The trap: a complicated question isn’t necessarily a better question

Consider these two investigations:

Question A: “How do temperature, humidity, light intensity, and plant species affect photosynthesis?”

Question B: “How does light intensity affect the rate of oxygen production in one aquatic plant species under specified conditions?”

Question A sounds ambitious. But investigating four independent factors together creates a much larger design and makes it harder to identify which factor explains a result.

Question B has a narrower comparison, although it still needs appropriate apparatus, a valid oxygen measurement, environmental controls, sufficient time, and safety approval.

Neither is automatically a good practical investigation. The method determines whether the question can be answered.

A strong question doesn’t need to produce a surprising result or prove your original hypothesis.

Sometimes the most useful finding is that the predicted effect wasn’t resolved by your measurements. That’s scientifically different from proving that no effect exists.

Your question should make that distinction possible.

07Which HSC requirements apply?

A science investigation may be part of a depth study, another school-based task, or a teacher-directed practical activity. The permission to choose your own question depends on the task.

NESA determines syllabus and HSC requirements. Your school determines the assessment instructions and the approved practical arrangements within those requirements.

As at 10 November 2026, the Physics and Chemistry Stage 6 syllabuses from 2017 remain in use for the current cohorts. The transition to the newer syllabuses is different for each subject:

Sources: NESA’s Physics 11-12 Syllabus (2025) and Chemistry 11-12 Syllabus (2025).

The 2017 Physics and 2017 Chemistry courses provide 15 hours of course time for depth studies in each year. Their 2025 replacements specify 10 hours in each year.

Those are course requirements, not guaranteed allocations of laboratory time for one individual investigation.

For your particular project, check your school’s assessment notice for the allowed investigation types, assessment criteria, submission requirements, equipment arrangements, and deadlines.

The NESA ACE Rules are an official source for broader assessment and certification requirements. UAC and tutoring providers do not approve individual school science investigations.

If you need a broader explanation of the course activity, see What Is a Depth Study in HSC Science?. The questioning, planning, measuring, and evaluating involved here also connect to What Is Working Scientifically in HSC Physics and Chemistry?.

08Questions and solutions

Question 1

A student is interested in how hanging masses affect the extension of a spring.

The school has approved a spring stand, one spring, a ruler, and masses of 50 g, 100 g, and 150 g, provided the spring remains within its safe working range.

The student’s proposed question is:

“Does adding weight make a spring stretch more?”

Rewrite this as a bounded, measurable investigation question. Identify the independent and dependent variables, two important controls, and one equipment or safety check.

Solution 1

A suitable question is: “How does an attached mass of 50 g, 100 g, or 150 g affect the extension of the same vertically suspended spring, measured in millimetres?”

The independent variable is the attached mass. The dependent variable is the spring’s extension, calculated by subtracting its unloaded length from its loaded length.

Two controls are using the same spring and measuring from the same reference point with the ruler in the same position.

The student should check that all three masses are within the spring’s approved working range and that the ruler can measure the expected extension with adequate resolution.

The original question was reasonable but insufficiently specific. “Stretch more” had no defined measurement, and “adding weight” didn’t specify the conditions being compared.

Question 2

A student wants to investigate how temperature affects the time taken for sugar to dissolve.

They use identical masses of sugar and equal water volumes, but the treatments differ:

Each condition is repeated five times. The recorded dissolving times decrease as temperature increases.

The student argues that five repeats prove temperature caused the difference.

Explain why this conclusion is not justified. Identify two competing explanations and propose a revised method that would allow a more meaningful temperature comparison.

Solution 2

The results cannot isolate temperature as the cause because stirring rate increases alongside temperature.

Two explanations are consistent with the observations:

• Higher water temperature reduced the time taken for the sugar to dissolve.
• Faster stirring reduced the dissolving time, either partly or entirely explaining the observed trend.

Both effects could also operate together.

Five repeats under each condition make the results more repeatable for those combined conditions. They do not remove the confounding variable.

A better design would vary water temperature while keeping the stirring procedure constant. The sugar mass, sugar form, water volume, container, and definition of the dissolving endpoint should also remain consistent.

The student should measure the water temperature during each trial, because it may change while the sugar dissolves.

The revised experiment would be able to compare dissolving times at different temperatures under a common stirring procedure. A pilot would still be needed to check that the timing method gives useful measurements.

Question 3

A student wants to determine which of two mat materials produces a higher rebound from a dropped ball.

They use the same ball and release it with its lowest point 60 cm above the laboratory bench.

Mat A is 1 cm thick. Mat B is 4 cm thick.

In a hypothetical trial, the peak rebound height of the ball’s lowest point, measured above the bench, is 36 cm for Mat A and 39 cm for Mat B.

One student concludes that Mat B produces a rebound 3 cm higher.

Another subtracts the mat thicknesses, obtains equal rebound heights, and concludes that the materials have identical rebound properties.

Are either of these conclusions justified? Calculate the release and rebound heights relative to each mat’s upper surface, then propose a fairer investigation.

Solution 3

Neither conclusion is justified. Mat B’s measured rebound peak is higher above the bench, but the mats have different thicknesses, and the ball is dropped from different heights relative to their upper surfaces.

First, calculate the heights relative to the mat surfaces.

For Mat A:

\[ \begin{aligned} h_{\text{drop,A}}&=60-1\\ &=59\text{ cm}\\ h_{\text{reb,A}}&=36-1\\ &=35\text{ cm} \end{aligned} \]

For Mat B:

\[ \begin{aligned} h_{\text{drop,B}}&=60-4\\ &=56\text{ cm}\\ h_{\text{reb,B}}&=39-4\\ &=35\text{ cm} \end{aligned} \]

Here, \(h_{\text{drop}}\) is the release height above the mat surface, and \(h_{\text{reb}}\) is the rebound height above that surface.

The rebound heights are both 35 cm when measured from the appropriate surfaces.

However, the release heights differ by 3 cm. The ball has less distance to fall before striking Mat B, so the impact conditions are not the same.

There is another problem: the mats differ in both material and thickness. Thickness can affect how a mat deforms and returns energy to the ball.

A fairer question would be:

“How does mat material affect the rebound height of the same ball when dropped from 50 cm above the surface of mats of equal thickness?”

The revised design should use comparable mat thicknesses, a fixed release height above each mat, the same ball, a consistent release method, and repeated measurements of rebound height above the mat surface.

The surprising point is that subtracting the mat thickness fixes the measurement reference but doesn’t fix the experiment.

Correct arithmetic cannot compensate for changing the physical conditions you intended to compare.

Question 4

A student wants to investigate whether rainfall makes a local creek more polluted.

The school cannot approve creek sampling during the assessment period but allows an investigation using suitable secondary data.

The student considers two hypothetical datasets:

• Dataset A: Eight weekly turbidity measurements from one creek, plus daily rainfall measurements from a weather station 40 km away. The sampling dates do not match.
• Dataset B: Daily turbidity measurements at 9 am over 28 days, plus rainfall totals for the preceding 24 hours, recorded at the same site on matching dates.

Which dataset offers the stronger basis for a manageable question?

Rewrite the original question, explain one important limitation of the chosen dataset, and consider whether a high turbidity reading after a day with no recorded rainfall would automatically disprove the proposed relationship.

Solution 4

Dataset B is the stronger option because the measurements refer to the same location and have a consistent daily relationship in time.

A defensible question is:

“What relationship exists between the previous 24 hours’ rainfall total and the 9 am turbidity reading at one creek monitoring site over the available 28-day period?”

This question is narrower than the original. It defines two measured quantities, a location, and a time period.

Importantly, turbidity measures the cloudiness of water caused by suspended material. It isn’t a complete measure of pollution.

The student could compare rainfall and turbidity records to investigate an association. They could not automatically conclude that rainfall caused each change in turbidity.

Other influences could include upstream conditions, sediment disturbance, construction activity, streamflow, or rainfall that occurred before the chosen 24-hour interval.

A high turbidity reading after zero recorded rainfall would not automatically disprove an association. Turbidity might respond to earlier upstream rainfall or another influence.

The student should retain such observations, investigate plausible explanations, and report the limits of the data.

Dataset A would be less useful for this particular question because its differences in location and timing make the proposed relationship harder to interpret.

The central decision is not simply to choose the dataset with more measurements. It is to choose data capable of addressing the exact question being asked.

Question 5

A student wants to investigate:

“How does light colour affect seedling growth when the amount of light is kept constant?”

The school approves a seven-day investigation with three light-tight growing boxes. Each box contains 10 seedlings of the same species.

One box uses a red LED, another a blue LED, and the third a white LED.

The student controls the water supplied and the air temperature. They adjust each lamp until a lux meter reads 500 lux at the seedlings, and they rotate the positions of the three boxes around the classroom each day.

For this question, assume the lux meter reports brightness weighted according to human visual sensitivity. Equal lux readings do not necessarily mean equal numbers of photons reach the plants.

At the end of the hypothetical investigation, seedlings in the blue-light box show the greatest average increase in height.

The student makes three claims:

1. The experiment has 10 independent repeats for each light colour.
2. The 500-lux readings mean that light quantity has been controlled fairly.
3. Swapping the coloured lamps between boxes halfway through the experiment would eliminate any remaining box-related effects.

Evaluate all three claims.

Can the available experiment establish that blue light causes greater seedling growth? Identify the essential changes needed for a stronger causal investigation, and propose the most defensible question if the school cannot provide more equipment or time.

Solution 5

The experiment cannot establish that blue light itself causes greater seedling growth. All three claims are flawed, and changing only one of the proposed features would not resolve the full problem.

Claim 1 is incorrect: ten seedlings do not provide ten independent experimental replicates of light colour.

All ten seedlings in each box share the same lamp and growing enclosure.

The treatment is applied at the box level. There is only one independent box for each colour.

The seedlings provide measurements of variation within each box, but they do not tell us how much the outcomes would vary between independently established blue-light boxes.

Suppose the blue-light box happens to have better ventilation. All ten seedlings could benefit from that condition.

The student could then observe consistently greater growth in the blue-light group without knowing whether light colour or the box environment caused it.

Rotating the positions of the boxes can help reduce differences caused by classroom location, but it does not separate the assigned lamp colour from the particular enclosure.

Claim 2 is incorrect: equal lux readings do not establish equal biologically relevant light quantities.

Lux measures illuminance using a weighting based on human vision.

Plants respond to photons across different wavelengths, and the photon supply associated with 500 lux can differ between red, blue, and white sources.

Therefore, two lamps with identical lux readings may deliver different numbers of photosynthetically relevant photons to the seedlings.

The experiment cannot separate a possible wavelength effect from a possible difference in photon supply.

A stronger investigation would need an appropriate measurement or calibration of the light delivered to the plants, with a clearly specified basis for matching light quantity.

Claim 3 is also incorrect: swapping lamps halfway through introduces a treatment-history problem.

Seedling growth accumulates over time.

A plant exposed to red light for part of the week and blue light for the remainder has experienced both treatments. Its final height does not represent an independent seven-day test of either colour.

The response to the second treatment could also depend on the plant’s earlier development.

Swapping lamps might seem to balance the boxes, but without a carefully designed experiment that accounts for treatment history, it creates another interpretation problem.

What would a stronger investigation require?

At minimum, it would need multiple independently assigned growing enclosures for each light colour, a valid method of comparing light supply across the colours, consistent environmental conditions, and an appropriate period of growth for each independent treatment.

With only three boxes and no suitable light-quantity measurement, those requirements cannot be met within the stated constraints.

The student should not claim to have isolated the causal effect of light colour.

A more defensible question using the available arrangement is:

“How do changes in mean seedling height over seven days compare among the three specified red-, blue-, and white-lamp growing arrangements?”

That is a descriptive comparison of the available systems. It does not establish that wavelength alone explains any observed difference.

The hardest part of choosing an investigation question is sometimes recognising that you can collect plenty of data without being able to answer the question you really wanted to ask.

The scientifically sound decision is to narrow the claim rather than overstate what the equipment and design can establish.

09What should you do after choosing your question?

Before committing to full data collection, show your teacher the proposed question, the variables, the measurement method, the planned repeats, the time budget, and the relevant safety considerations.

Once the method is approved, a small pilot can reveal whether the apparatus works, the measurements are sufficiently sensitive, and the planned conditions are practical.

If that pilot exposes a problem, adjust the question or method before collecting your main dataset.

A good investigation question is not the broadest or most impressive-sounding one. It’s the one your evidence can genuinely answer.