What it is. The scientific method is the standard way to test an idea. You don't need a lab coat to use it; you need to follow the 6 steps honestly. Skip step 3 (hypothesis) and you're guessing. Skip step 5 (analyze) and you can't claim anything.
1Observe
Notice something in the world that makes you curious. Strong scientists observe first, ask questions second.
Example: You water two plants the same way for a week, but one is in the window and the other is in a closet. The one in the closet looks pale and droopy.
Pitfall: jumping straight to "I bet sunlight matters" without first describing what you noticed. Observations come BEFORE explanations.
2Question
Turn the observation into a specific, testable question. Vague questions get vague answers.
Example: "Does the amount of light a plant gets affect how green it is?" Specific. Testable. Has a fair-test design lurking inside it.
Pitfall: a question that can't be tested (like "Why do plants exist?"). If you can't design an experiment for it, narrow it down.
3Hypothesize
Write down what you think will happen AND why, before you do the experiment. A hypothesis is an educated guess, not a wish.
Example: "I think the plant with more light will be greener because plants use sunlight to make chlorophyll, which is what makes them green."
Pitfall: writing the hypothesis after you see the result so it always "matches". That's not science, that's cheating yourself.
4Experiment
Design and run a fair test. Change ONE thing (the independent variable), measure the effect (dependent variable), and keep everything else the same (controlled variables). See the Variables drill for practice.
Example: Two plants, same species, same soil, same water, same temperature. ONE difference: hours of sunlight per day (2h vs 8h). Run the test for 14 days.
Pitfall: changing two things at once. If you give one plant more light AND more water, you can't tell which mattered.
5Analyze
Look at the data. Make a chart or graph. What's the pattern? What's surprising? Are there any outliers? A blob of numbers with no analysis is not science.
Example: Plant A (2h light): leaf colour score 3/10 at day 14. Plant B (8h light): leaf colour score 8/10. The difference is large and consistent.
Pitfall: ignoring data that doesn't match your hypothesis. Include everything. Surprising results are often the most interesting.
6Conclude
Answer your original question, supported by the data. Say what the data shows, what it doesn't show, and what you'd test next.
Example: "More light produced greener leaves (8/10 vs 3/10) over 14 days. This supports my hypothesis. To test further, I'd vary water and temperature too, and run with more plants per group."
Pitfall: over-claiming. Two plants is not strong evidence about all plants everywhere. Be honest about what your experiment proves and doesn't.
📋 Worked example: bouncing balls
A short end-to-end example you can imitate for a school project.
| Step | What I wrote |
|---|---|
| 1. Observe | Some balls bounce higher than others when I drop them. |
| 2. Question | Does the type of surface (tile, carpet, wood) affect how high a ball bounces? |
| 3. Hypothesis | The harder the surface, the higher the bounce, because hard surfaces don't absorb the energy. |
| 4. Experiment | Same tennis ball, dropped from 1.0 m, onto three surfaces (tile, carpet, wood). 5 drops each, measure bounce height with a ruler. |
| 5. Analyze | Tile: 65 cm avg. Wood: 60 cm. Carpet: 35 cm. Tile is hardest and bounced highest. Carpet bounced lowest by a lot. |
| 6. Conclude | The data supports my hypothesis. Harder surfaces give higher bounces. Next test: vary drop height, or try different ball types. |