Designing a fair biological experiment — Biology, 14–17
A biological result is useful only when its test separates the factor you changed from everything else. Controls, repeated trials and clear measurements help you decide whether a pattern is real or just chance.
Change one thing, measure another
In a fair experiment, you deliberately change one factor, called the independent variable, and measure a result, called the dependent variable. Other important conditions stay the same. A comparison group shows what happens without the treatment, so the difference has a possible explanation.
Why one result can mislead
Plants given more fertiliser may grow taller, but perhaps they also received more water or more light. A fair design asks which change caused the difference instead of accepting the first story. Repeated trials reveal variation, and random assignment helps stop pre-existing differences from choosing the result for you.
Testing light and seedling growth
Place 20 similar seedlings in two groups of 10. Give both groups the same soil, water and temperature, but give one group 8 hours of light and the other 12. Measure height after seven days, using the same ruler. Compare the group means and report the spread, not just the tallest plant, then repeat the experiment.
“A bigger difference must be real”
A large-looking difference can still come from chance, especially with few samples or noisy measurements. It is reasonable to trust your eyes because a clear pattern feels convincing. But you should inspect all data, repeat the test, estimate variation and ask whether the groups were treated equally before claiming an effect.
Clinical trials and everyday claims
The same logic is used to test medicines: volunteers are assigned to treatment or comparison groups, outcomes are measured and side effects are recorded. You can use the habit when judging claims about supplements, diets or study methods. Ask what changed, what was compared, how many people took part and whether the measurement fits the claim.
Keep exploring
Other languages
Loading MyLeoNes™…