In a controlled experiment, the scientist keeps all the conditions of an experiment, such as temperature, humidity and amount of light, the same. This allows the scientist to change a single variable, also known as an independent variable, in order to see what resulting changes occur. For example, in an experiment involving the effect of gamma rays on marigolds, the scientist would change the amount of gamma rays from one batch of marigolds to another, but keep all other conditions the same.
A scientific experiment usually has a specific structure. It is based on a hypothesis, a statement predicting the results of the experiment. It uses an equipment list and setup instructions. It records the results of each time the experiment is performed. Once the experiment has been performed an appropriate number of times, the results can be written up.
Say a researcher wants to test the effect of watching cartoons on children. To make the experiment meaningful, everything else must be equal, such as the age and living conditions of the children, the kind of television they're watching and the conditions under which they're watching it. All of these conditions are difficult to control, and any time unpredictable human subjects are used in an experiment, it's hard to predict their effect on the experiment.
On the other hand, the same experiment with rats might be more effective because more things can be controlled about the rats, down to the genetic makeup of the rats themselves.
To make sure your experiment is structured, follow these basic steps of the scientific method: Make observations and study the problem you want to solve until you come up with a hypothesis, which is a prediction based on your knowledge and research. The experiment will prove or disprove your hypothesis, so you must create one that tests the hypothesis and controls all other conditions.