What Is the Purpose of a Control in an Experiment?

An experiment without a control is not an experiment; that’s how essential the control is. When a scientist speaks of a control, however, she might mean one of two things: a group of subjects not submitted to a treatment or the management of a nuisance factor in an experiment. Either way, without a control, a scientist cannot make a conclusion about the relationship between independent and dependent variables.

Though a natural inquiry of science students and even graduate students who should know better, according to the University of Colorado, the question “What is the purpose of a control in an experiment?” is equivalent to the question “What is the purpose of vegetables in a salad?” With the exception of fruit salads and pasta salads, the answer to both of these questions is that you can’t have the latter without the former. A study with a control has the basic requirement of being an experiment. A study without a control is a non-experimental study.

Remember that an experiment contains two variables of interest: the independent variable and the dependent variable. A scientist running an experiment will vary the independent variable and observe changes in the dependent variable. But according to Roger Kirk, distinguished professor of statistics and author of “Experimental Design,” such an experiment is not complete without considering other factors that might have roles in affecting the dependent variable. For example, an experimenter who suspects that talking to plants helps them grow might assign to his experiment the dependent variable “growth” and the independent variable “amount of words spoken to plant.” He could then talk to the plants and witness growth, attributing his words spoken to the growth of the plants. But it’s what he didn’t consider that might be the main reason for growth: aspects such as sunlight, temperature, and water given. His experiment lacked control.

When scientists refer to a “control,” they could be referring to one of two things. The first is the control group. The control group is a group of subjects in an experiment that lack an independent variable or have a “standard” value for the experiment’s independent variable, such as zero. A control group allows a scientist to compare it to the other group or groups in an experiment. If a scientist notices a significant difference between the control group and one or more of the other groups, he can logically lead to the conclusion that the independent variable has an impact on the dependent variable. For example, a scientist talking to plants will need another group of plants, the control, to undergo the same experiment with the independent variable, “amount of words spoken to plant,” being equal to zero. Then, all things being equal, if he witnesses the “listening” plants growing more quickly than the “deaf” plants, he can justify his hypothesis that plants grow faster when spoken to. In a sense, when the scientist doubles his subjects and does nothing special to half, he has a control.

A scientist using a control group must ensure that the control group is equal to the experimental group, or treatment group, in every way except the independent variable. This is the other meaning of “control.” A scientist working with plants might control the amount of sunlight all the plants get. He would also control their water intake, temperature and placement in a room. When he does this to both the control group and the treatment group, he can be certain that it is only the variation in the independent variable that leads to a variation in the dependent variable between the control group and treatment group. Without controlling these “nuisance factors,” a scientist cannot conclude the existence of a relationship between the dependent variable and the independent variable, which is often the main goal of an experiment.