Consider a chemical test, such as a test for the presence of lead in drinking water. You would use both experimental and control samples. Whenever you run the test on experimental samples, you also need to run it on both positive and negative control samples--positive samples are prepared deliberately with a certain known concentration of lead, and negative samples are known not to contain any lead (you have to be careful here!). If you get the incorrect results for the control samples, then you know that the results for the experimental samples may be off, so you have to repeat the test. You also might want to periodically have the analyst run a proficiency test with the standard positive and negative controls as well as samples containing various known amounts of lead.
Experimental and control groups might more typically refer to situations such as drug trials, where you are comparing a new drug to another drug that is already approved and used to treat an illness. The experimental group would be the group of subjects having the illness that are given the new drug and the control group would also have the illness but would be given the approved drug (preferably in a double-blind study design--where neither the subjects nor the people administering the drug know which treatment is which). Note that both the experimental group and control group are "experimented on" - otherwise they could tell whether they received the treatment or not, and the results could be erroneous due to the placebo effect.