When you set up an experiment, you have to have one set up where you introduce the independent variable (the experimental set up) and one where you do not (the control). That way, you can compare the two and see the impact of the independent variable.
As the website I have linked to says:
The scientist must contrast an “experimental group” with a “control group”. The two groups are treated EXACTLY alike except for the ONE variable being tested. Sometimes several experimental groups may be used.
For example, if you want to determine the impact of adding a kind of fertilizer to your crops, you have to do it like this:
You need one plot of plants that get the fertilizer, but you also need one plot that does not get fertilized. Then you see how much they each produce and compare. By having the control group, you can see how much of an effect the fertilizer (the independent variable) had. If you had no control, you would not know how much impact the fertilizer had.
When using the scientific method to solve a problem or question, one formulates an educated guess or hypothesis which is a possible answer to the problem or question. In order to test your hypothesis, you perform an experiment. An experiment is set up to test the effect of a variable on an outcome. This variable that the experimenter manipulates is known as the independent variable. If some change occurs as a result of the independent variable, this change is the dependent variable.
If you set up an experiment, you need a standard of comparison to see if any change occurs as a result of your independent variable. Therefore, you divide your experiment into two groups. The experimental group and the control group. The experimental group or the manipulated group, gets the variable you are testing. However, the control group, does not get the variable. At the end of the experiment, you compare the data from both groups to see if your hypothesis was correct or not. For example, if I want to test the hypothesis that plants grow best if given a new plant food, then, if I have 500 plants for instance, I would divide my plants into two groups of 250 plants. The first 250 plants, or the experimental group would be given the plant food. The presence of this plant food is the independent variable I am manipulating. The other 250 plants are the control group. They will not receive any plant food.
I will collect data for several days or weeks. This data would be measurements of the height of my plants in both groups. All other variables must be the same in both groups--amount of light, water, type of pots, soil, etc. The only variable that is different would be the plant food given to the experimental group. At the end of the study, I will analyze the data, and see if I will accept or reject my hypothesis. A good experiment should be re-tested several times to make sure the results are accurate!