Concept Studio
See the idea. Put it to the test.
956 activities from Kindergarten to Year 12, in 13 subject areas. A practical gives the idea, what you need, the steps, what you should see and a safety card. A teacher-led demonstration gives the idea and its hazards; its method is for tutors on the learning platform.
Review
Reviewed before publication (owner’s confirmation, 24 September 2026). That covers every activity here, and a practical’s page lists the sources its author read.
A safety card on every page
The risk, who supervises and the hazards. The 38 teacher-led demonstrations show their idea and hazards here; their materials, steps and sources, and any result, control or note that states a number or an amount, are for tutors and administrators on the learning platform.
School laboratory, not for home
207 activities are medium or high risk. Each says so on its page: In a school laboratory, with a teacher supervising, under the school's risk assessment. Not for home.
Curriculum references
Each activity lists the NSW syllabus outcomes and Australian Curriculum v9 codes it supports. They are references, not a verified or complete curriculum alignment.
Find an activity
41 activities
Investigating Science 11–12
A draining bottle: testing a mathematical model and refining it
A mathematical model predicts how the water level in a draining container falls; testing it against a real bottle shows where the ideal model fails and how one measured constant repairs it.
Practical, model not builtLow riskA model house: testing wall and window materials for heat loss
A small model house and a heat-loss equation both predict how building materials change indoor temperature, and the model’s limits (scale, no sun, no draughts) are as instructive as its results.
Practical, model not builtLow riskA scale model of the solar system on the school oval
Built to one scale for both sizes and distances, the solar system is specks separated by enormous gaps, something no page diagram can show.
Practical, model not builtLow riskAcids and bases: indicator colour against pH readings in a dilution series
Indicator colour is a qualitative observation and a pH meter reading is a quantitative one; a tenfold dilution raises the pH of a strong acid by one unit but of a weak acid by only about half a unit.
Practical, model not builtLow riskCollecting primary data for the law of conservation of mass
Mass is conserved in every reaction; an apparent gain or loss means matter crossed the boundary of the system, and even a sealed balloon can mislead a balance through buoyancy.
Practical, model not builtLow riskConservation of energy in a pendulum measured with a light gate
The law of conservation of energy predicts a pendulum bob’s speed at the bottom of its swing from the height it was released, whatever its mass.
Practical, model not builtLow riskCounting background radiation: the irregular pattern of random decay
Radioactive decay is random, so counts in equal time intervals scatter irregularly, yet the scatter follows a Poisson distribution in which the variance equals the mean.
Practical, model not builtLow riskDistance and exposure: the inverse square law measured with light
Intensity from a small source falls with the square of the distance, a law whose predictions a measurement can test, and the reason distance from a radiation source is a basic way to reduce exposure.
Practical, model not builtLow riskEnergy transfer in a bounce: tennis ball rebound against the ITF standard
Energy is conserved but not kept as motion: a bouncing ball returns only part of its gravitational energy, and a regulation tennis ball must lose between about 42 and 47 per cent of it in each bounce.
Practical, model not builtLow riskFinding a hidden target with rolling marbles: a model of Rutherford scattering
Like Rutherford’s team, the learner infers the size and shape of something unseen from how often and how projectiles are deflected, and the inference sharpens as the number of trials grows.
Practical, model not builtLow riskFive white solids: identifying chemicals from observations and clues
Colour changes, fizzing and warming are observations; the name of each powder is an inference, and some pairs cannot be told apart with the tests available.
PracticalMedium riskGrowth of radish seedlings watered with saline water
Plant growth falls once the salinity of the water a crop receives passes a threshold, and a class trial infers that threshold from its own data before checking it against published secondary data.
Practical, model not builtLow riskLogging the strata in a rock cutting
A rock face gives qualitative observations (colour, structures) and quantitative ones (bed thickness, grain size, dip), and the principle of superposition turns the measured log into an inferred sequence of events.
PracticalMedium riskMeasuring cells with a calibrated eyepiece graticule
A drawing or description of a cell stays qualitative until the microscope is calibrated; a graticule checked against a stage micrometer turns each observation into a measurement with a stated uncertainty.
PracticalMedium riskMendel’s law of dominance: his counts, a bead model and a chi-square test
Mendel’s law of dominance predicts a 3 : 1 ratio in the second generation; his own counts and a bead model test whether departures from 3 : 1 are within chance.
Practical, model not builtLow riskOhm’s law as a prediction: resistor against filament lamp
A law earns acceptance by predicting results before they are measured: Ohm’s law predicts the current in a resistor, and a filament lamp shows the conditions under which it stops applying.
Practical, model not builtLow riskPredicting the return of Halley’s comet with Kepler’s third law
A pattern of returns about every 76 years led Halley to infer one comet on one orbit; Kepler’s third law turns the orbit’s size into a period, and the spread in real intervals shows what the simple law omits.
Calculation and dataLow riskReactions of calcium carbonate: limewater, carbon dioxide and cave formation
Carbon dioxide first precipitates calcium carbonate from limewater and then redissolves it as calcium hydrogencarbonate; heating drives the gas off and the solid returns, the chemistry that dissolves caves and builds stalactites.
PracticalMedium riskSlaters in a damp and dry choice chamber, with a chi-square test
Counting where animals settle turns a behaviour into quantitative data, and a statistical test shows whether a small sample is enough to support the pattern the eye sees.
PracticalLow riskTesting the assumption that the Sun rises in the east
The Sun rises due east only at the equinoxes; on other dates the rising point swings north or south by an amount set by latitude and the Sun’s declination.
Practical, model not builtLow riskTesting trend models for carbon dioxide on Cape Grim data
Different mathematical models can fit the same past data almost equally well yet predict differently, so a model is judged by how well it predicts data it was not fitted to.
Calculation and dataLow riskThe Bernoulli effect measured in a Venturi tube
Where a flowing fluid speeds up through a narrow section its pressure falls, which a pair of water columns shows as a measurable height difference that Bernoulli’s equation predicts.
Practical, model not builtLow riskThe candle in a closed container: separating observation from inference
What is seen (the flame goes out, the water rises) is an observation; “the oxygen was used up” is an inference, and a designed test shows the rise is caused mainly by heating and cooling of the trapped gas.
PracticalMedium riskThe length of the NSW coast: measuring a fractal pattern with dividers
The measured length of a coastline grows as the measuring step shrinks, an irregular pattern Richardson found and Mandelbrot described with a fractal dimension, so a coastline has no single true length.
Practical, model not builtLow riskThe ‘life’ of an alkaline cell under different loads
A battery’s life is not one number: an alkaline cell delivers less charge at higher current, so an inference drawn from one test holds only for the load and schedule tested.
Practical, model not builtLow riskTiming a falling object: what quantitative data reveal that the eye cannot
A falling object gains speed at a constant rate, and only timed measurements show that a dense ball falls at about 9.8 m/s² while a light ball is slowed measurably by air.
Practical, model not builtLow riskCalibrating analogue and digital thermometers at two fixed points
Every thermometer carries systematic error; checking it at two fixed points, melting ice and boiling water corrected for the day’s air pressure, exposes that error and separates accuracy from resolution.
Practical, model not builtMedium riskControl groups and blinding: a double-blind taste test with a binomial test
Blinding, random order and a probability calculation separate a real ability to tell two products apart from chance and from expectation.
Practical, model not builtLow riskDiffraction as a measuring tool: a CD, a laser and an X-ray diffraction analogue
A regular pattern too fine to see spreads light into a predictable pattern, so measuring the pattern reveals the spacing; this is the principle that let X-ray diffraction reveal the helical structure of DNA.
Practical, model not builtMedium riskEarthquake-resistant design on a shake table: resonance, bracing and base isolation
A building sways most when the ground shakes near its natural frequency; bracing raises that frequency and base isolation lowers it, both moving the structure away from resonance as Newton’s second law predicts for a mass on a spring.
Practical, model not builtLow riskMeasuring the Earth with two shadows, the Eratosthenes method
Two shadow angles measured at local noon at places a known north–south distance apart give the Earth’s circumference by geometry, the method Eratosthenes chose because the Earth could not be measured directly.
Practical, model not builtLow riskPressure and volume of a gas: Boyle’s law and a hidden systematic error
For a fixed amount of gas at constant temperature, pressure times volume stays constant; air hidden in the connecting tube is a systematic error that bends the data until it is counted.
Practical, model not builtLow riskPriestley’s restored air made measurable: algae and hydrogencarbonate indicator
Priestley inferred that plants restore air that burning or breathing had spoiled; hydrogencarbonate indicator lets a learner measure the same effect as carbon dioxide removed in light and added in darkness.
PracticalLow riskReaction time from a falling ruler: reliability, sample size and outliers
One measurement of reaction time is unreliable; repeated trials and larger samples shrink the uncertainty of the mean, and each outlier has to be judged against a stated rule rather than deleted.
Practical, model not builtLow riskRefraction to telescope: building a two-lens telescope
The laws of refraction that shape a lens also fix a telescope’s design: two converging lenses separated by the sum of their focal lengths magnify by the ratio of those lengths.
Practical, model not builtMedium riskSample size and sample selection: drawing beads from a bag
Small samples scatter widely around the true proportion and the spread shrinks with the square root of sample size, while a biased way of choosing the sample misleads at every size.
Practical, model not builtLow riskTemperature and gas volume: Charles’s law and an estimate of absolute zero
At constant pressure a gas’s volume is proportional to its absolute temperature, so extrapolating measured volumes to zero estimates absolute zero, and the equipment limits how good that estimate is.
Practical, model not builtMedium riskTemperature and reaction rate: stopwatch against light sensor
The rate of the thiosulfate–acid reaction rises steeply with temperature, and the technology used to detect the end point, the eye with a stopwatch or a light sensor, decides how reliable the result is.
PracticalMedium riskTesting a label claim: how much acid does an antacid tablet neutralise?
An efficacy claim on a medicine label can be tested quantitatively: a back-titration measures how much acid one tablet neutralises and compares it with the stated active ingredient.
Practical, model not builtMedium riskTesting a sunscreen claim: how much UV gets through a thin layer?
An SPF is measured at a set thickness of sunscreen; a thinner layer lets through far more UV, so testing the claim shows why the amount applied matters as much as the number on the tube.
Practical, model not builtLow riskTesting the Doppler prediction with a swinging buzzer and a phone
A source moving toward a listener is heard at a higher frequency and moving away at a lower one; a swinging buzzer and a spectrum app turn Doppler’s prediction into a measured test.
Practical, model not builtMedium risk
For tutors and administrators
The materials and steps of every teacher-led demonstration are on the learning platform, with the safety card first. Sign in with a tutor or administrator account to read them.