A film canister, a fizz of gas, and a launch. The real experiment isn't the bang — it's designing a fair test to find what makes it fly highest.
Everyone wants the highest launch. But "make the best rocket" is really a question about experimental design: which one thing will you change, what will you measure, and what must you keep the same so the comparison is fair?
Change the amount of vinegar and the amount of bicarb and the canister at once, and a good launch tells you nothing — you can't say which change caused it. The whole skill is isolating a single cause.
Vinegar (an acid) meets bicarbonate of soda (a base). They react and release carbon dioxide gas. Sealed in the canister, the gas has nowhere to go — pressure climbs until it forces the lid off. Gas blasts down; by Newton's third law, the canister is pushed up. Play it:
Put what you're measuring in the centre. Around the outside go all the factors that could affect it. Tap one factor to make it your independent variable — instantly, every other factor becomes a control you must hold steady.
The competitive instinct is to tweak everything at once and see what wins. Commit to an answer before reading on.
Change one factor; keep the rest the same. If you change three things and the rocket flies higher, you can't tell which change did it — maybe two helped and one hurt. Isolating one independent variable is the only way the result actually means something.
The misconception this corrects: that a "better" experiment tests more changes at once. The opposite is true — a fair test is disciplined about changing exactly one thing.
Here the independent variable is the volume of vinegar; everything else is held constant. Drag the slider and read the launch height. Notice there's a sweet spot — and that a clean trend only appears because nothing else is changing.
| Vinegar / ml | 2 | 4 | 6 | 8 | 10 |
|---|---|---|---|---|---|
| Height / m | 0.9 | 1.8 | 2.6 | 2.5 | 2.1 |
Accept: "peaks at 6 ml then falls" for the trend; any sensible reliability improvement. Do not accept: "change the vinegar and bicarb both" (breaks the fair test).
Fill the blanks so the sentence reads as a model answer.
In a fair test you change one variable and measure the variable. Every other factor is a variable, kept the same so the test stays . Repeating trials and taking a mean improves .
The one factor you deliberately vary. Pick it, and you've defined your whole experiment.
The outcome you record — here, launch height. It responds to the independent variable.
Every other factor held constant, so the comparison is fair and the result is valid.
Repeat trials and take a mean. Controlling variables gives validity; repeating gives reliability.