A prediction written down before the test starts turns an ordinary experiment into a project a student actually owns — this is the single change that separates Predict-Observe-Explain from a worksheet. Every idea below follows the same three-step shape: guess what will happen and say why, build or run the test, then explain the gap between the guess and the result. That gap is usually where the physics lives.
Why predict first
Skipping straight to "let's see what happens" feels efficient, but it removes the one moment where a student has to commit to a model of how the world works. A wrong prediction is not a failure — it is the exact point where a misconception becomes visible and correctable. Right predictions are just as useful, because they confirm the student's mental model is doing real work. Either way, write the prediction down before starting; a prediction made after seeing the result is not a prediction.
The 10 project ideas
1. Spaghetti and paper bridge under load
Hook: how much weight can a bridge made from twenty strands of dry spaghetti actually hold?
What they do: design and build a small bridge spanning a fixed gap (20-30cm) using only spaghetti, paper, and tape, predict its breaking load in grams, then load it gradually with coins or weights until it fails.
Physics skills: forces in tension and compression, structural design, fair testing (one design change at a time).
Scope: 2 weeks, one afternoon a week — one week to design and build, one to test, break, and redesign.
2. Pendulum period vs length
Hook: does a longer pendulum swing slower, and by how much?
What they do: build pendulums of five different lengths, predict how period changes with length, time 10 swings for each length, then plot period against length and compare the curve to the prediction.
Physics skills: measurement and timing, graphing, relating a variable to a non-linear pattern.
Scope: 1 week, one afternoon.
3. Marble-run energy transfer
Hook: where does the energy go when a marble rolls from the top of a run to the bottom?
What they do: build a marble run from cardboard tubes and track, predict how fast the marble will be moving at three points, then measure using video slow-motion or a timing gate, and account for the energy "lost" to friction and sound.
Physics skills: gravitational potential energy, kinetic energy, energy dissipation.
Scope: 2 weeks, one afternoon a week.
4. Insulation materials vs ice melting
Hook: which household material actually slows down melting ice — bubble wrap, wool, or foil?
What they do: wrap identical ice cubes in different materials, predict a melting-time order, then time how long each takes to fully melt under the same conditions.
Physics skills: thermal conduction, fair testing, controlling variables.
Scope: 1 week, one afternoon (plus overnight freezing prep).
5. Simple circuit conductivity tester
Hook: which everyday materials complete a circuit and light a bulb?
What they do: build a simple battery-bulb-wire circuit with a gap for a test material, predict which of ten household items (coin, pencil lead, rubber, foil, water) will conduct, then test each and sort the results.
Physics skills: circuits, conductors vs insulators, classifying evidence.
Scope: 1 week, one afternoon.
6. Catapult range, one variable at a time
Hook: does a longer arm or a stiffer elastic band launch a projectile further?
What they do: build a small catapult, predict how range changes when arm length changes (band held constant), test and record, then repeat varying band tension with arm length held constant.
Physics skills: projectile motion, fair testing, isolating variables.
Scope: 2-3 weeks, one afternoon a week.
7. Parachute descent time vs canopy size
Hook: does doubling a parachute's canopy area halve its fall speed, or is the relationship different?
What they do: build parachutes with three or four different canopy areas from the same material, predict descent time from a fixed height, then drop and time each.
Physics skills: air resistance, terminal velocity, proportional reasoning.
Scope: 1-2 weeks, one afternoon a week.
8. Build and test a solar oven
Hook: can a shoebox lined with foil actually melt chocolate using only sunlight?
What they do: build a solar oven from a box, foil, and cling film, predict the maximum temperature it will reach, then test on a sunny day with a thermometer and log temperature over time.
Physics skills: reflection, absorption, energy transfer, data logging.
Scope: 2 weeks, one afternoon a week (weather-dependent).
9. Friction on different ramp surfaces
Hook: why does a toy car glide down a plastic ramp but stop halfway down a carpeted one?
What they do: build an adjustable ramp, predict which of four surface materials will produce the least friction, then measure the angle needed to start the car moving on each surface.
Physics skills: friction, forces on an incline, fair testing.
Scope: 1 week, one afternoon.
10. Egg-drop protection design
Hook: can you design a container that lets a raw egg survive a two-metre drop?
What they do: design and build a protective casing using limited materials, predict whether it will survive drops from 1m, 2m, and 3m, then test at each height and redesign after failure.
Physics skills: impact forces, energy absorption, iterative design.
Scope: 2 weeks, one afternoon a week.
Turning any of these into a proposal
Before building anything, it's worth writing down the question, the prediction, and how the test will be fair — see how to write a project proposal for KS3 for a template that works for any of the ten ideas above. If your child is choosing between several of these, helping your child choose a project topic covers how to pick one that will actually hold their interest for the full scope.
How an AI tutor fits in
None of these projects need a tutor standing over them, but Professor Newton is built for exactly the moment before the test — asking "what do you think will happen, and why?" before the pendulum swings or the catapult fires, the same Predict-Observe-Explain shape used throughout this list. See how AI tutors support project-based learning for how that works across subjects, not just physics.
FAQ
What makes a physics project different from a physics experiment?
An experiment usually follows a set method to reach a known result. A project starts with a genuine prediction, gives the student choices about design and variables, and ends in something built or measured that they can defend — the process is assessed alongside the answer.
Do these projects need special equipment?
Most use household materials — spaghetti, string, cardboard, ice, elastic bands. A few (circuits, solar ovens) benefit from a cheap multimeter or a sunny day, but none need a school lab.
How long should a KS3 physics project take?
Most of the ideas below fit into 2-3 weeks of one afternoon a week, which is long enough to predict, build, test, and redesign once — the redesign is where most of the physics understanding actually lands.