How to Choose a Science Fair Question That You Can Actually Test

The hardest part of a science fair project is the first part: deciding what to do. Many students start with a topic (“volcanoes”, “plants”, “magnets”) and get stuck there. A topic isn’t a project. A testable question is.

What makes a question testable

A testable question has four features:

You can change one thing. The type of paper towel, the height of a ramp, the amount of light.
You can measure something. Millilitres, centimetres, seconds, grams, a count.
You can do it at home or school with materials you can get and in the time you have.
The answer isn’t already obvious and can’t be looked up in one search.

“How do paper towels work?” fails: nothing to change, nothing to measure. “Which brand of paper towel soaks up the most water?” passes.

From topic to question in four steps

Step 1: Pick a topic you’re curious about. Ice, plants, paper planes, sound, bouncing balls.

Step 2: List things that could change. For paper planes: wing size, paper weight, nose weight, fold type, throw angle.

Step 3: Pick one and decide what to measure. Wing size, measured by how far the plane flies in centimetres.

Step 4: Write it as “Does ___ change ___?” Does the length of a paper plane’s wings change how far it flies?

Quick test: the five-second check

Ask yourself:

What will I change? (one thing)
What will I measure, and in what units?
What will I keep the same?
Can I repeat it at least three times?
Is it safe?

If you can answer all five, you have a project.

Examples by grade

K to Grade 2: Which ball bounces highest? Does salt make an ice cube melt faster?
Grades 3 to 5: Do longer wings make a paper helicopter fall more slowly? Does a bean plant grow taller in sun or shade?
Grades 6 to 8: Which material keeps ice frozen longest? Does a bigger bottle make a better thermometer?

Demos aren’t experiments

A baking soda volcano is a demo: it shows something already known. Ask a question about it and it becomes an experiment: “Does warmer vinegar make the volcano fizz higher?” Judges nearly always prefer the experiment, and it doesn’t have to be fancy.

Don’t forget: a result you didn’t expect is still a result

If the data disagree with the prediction, that isn’t failure. It’s what science is for. “I predicted heavier would swing faster, and the swing time barely changed” is an excellent finding, as long as you can explain it.

Free planner

Our free Science Fair Planner has 28 home-friendly project ideas sorted by grade, a fair-test planner, a data table and graph grid, a claim-evidence-reasoning page, a display board layout and a judging rubric. You can also plan online with the fair-test planner.

Want practice before choosing? Try the lesson Why do scientists repeat their tests?, where students plan and run a real fair test in one class. Sign up for three full lessons free.