In an experiment, a control is a standard for comparison that remains constant and unchanged. It helps to isolate the effects of the variable being tested by providing a baseline measurement against which results can be evaluated.