Skip to main content
SWStudyWalks Search Sign in

Natural Sciences / Unit 1   NSC 0011 worked example, 60 to 90 seconds

Designing and Judging a Controlled Experiment

Captions available

The idea

To design a fair experiment, work in five steps. First, state the hypothesis and the prediction. Second, name the independent and dependent variables. Third, set up a treatment group and a control group that differ only in the independent variable; when the treatment means adding something, give the control a harmless substitute, so the act of adding is controlled too. Fourth, use several identical subjects in each group. Fifth, compare the groups: a difference in the predicted direction supports the hypothesis, and no difference rejects it.

In practice

Worked example: does a new plant food speed the growth of bean seedlings? Twenty seedlings of one variety are planted in identical pots of the same soil. Ten get 50 milliliters of plant-food solution each week; ten get 50 milliliters of plain water. Light and temperature are the same for all twenty. The independent variable is the plant food; the dependent variable is the growth in height over three weeks. Results: the treated seedlings grow an average of 9.0 centimeters and the controls 6.0 centimeters, so the treated plants grew 3.0 centimeters more, half again as much. The result supports the hypothesis. Had both groups grown 6.0 centimeters, the hypothesis would be rejected.

Where people slip

Judge the design before the result. A test without a matched control group cannot show what caused a difference, however large the difference looks.

Your place on the trail

Every lesson rests on earlier ones and opens later ones. Tap any to jump.

Something wrong in this lesson? Flag it  ·  Sources for Natural Sciences