The scientific method
| English | Chinese | Pinyin |
|---|---|---|
| hypothesis | 假设 | jiǎ shè |
| independent variable | 自变量 | zì biàn liàng |
| dependent variable | 因变量 | yīn biàn liàng |
| control variables | 控制变量 | kòng zhì biàn liàng |
| repeats | 重复 | chóng fù |
| control group | 对照组 | duì zhào zǔ |
A hypothesis you cannot disprove is not a hypothesis
- A hypothesis 假设 is a testable statement. If no possible result would contradict it, it is not testable.
- "Plants grow better with music" is testable. "Plants have feelings we cannot measure" is not.
- The whole method is built to make it possible to find out you were wrong, efficiently.
Which is a testable hypothesis?
A result could contradict it. If nothing could, it is not a hypothesis.
The three variables
- The independent variable 自变量 is what you change.
- The dependent variable 因变量 is what you measure.
- The control variables 控制变量 are everything you hold constant so the change is the only explanation.
- Naming all three before starting is the planning mark, and it is where most students lose it.
In the fertiliser experiment, match each element to its role.
Naming all three before starting is the planning mark, awarded before any data exists.
Name one control variable for an experiment on how temperature affects how fast sugar dissolves.
Example: "The same mass of sugar with the same grain size in every trial."
Controls and repeats
- A control group 对照组 shows what happens without the treatment, so you can tell an effect from a trend.
- Repeats 重复 show the spread. One reading is an anecdote; five let you see whether a difference is bigger than the noise.
- ⚠ Report the spread, not just the average. Two means that differ by less than the variation within each group have not shown anything.
Why include plants that get no fertiliser at all?
Without a control, plants growing over four weeks proves nothing about the fertiliser.
Does fertiliser increase plant growth?
Independent: fertiliser amount. Dependent: height after four weeks. Controls: same species, soil, water, light, pot size. Control group: no fertiliser. Repeats: five plants per condition.
Writing those five lines before touching a plant is the plan, and a marker can award it before any data exists. Students who start planting first almost never recover the marks, because a control they did not think of cannot be added afterwards.
An experiment that fails is a result. Report it honestly with an explanation and it scores above one that quietly reports the textbook answer, because the method is what is being assessed.
Changing the fertiliser and the light together saves time and gives the same information.
No result can then tell you which one mattered, and no later analysis can rescue it.
Do not change two things at once. If you increase the fertiliser and also move the plants to a brighter window, no result can tell you which one mattered — and there is no analysis that can rescue it afterwards.